Search

This thesis details research undertaken to develop novel techniques to deliver an automated methodology for optimising hybrid electric powertrains. Uniquely the target for optimisation is high performance road and track cars. Novelty in the research described includes the use of existing components; evaluating multiple topologies, and optimising for multiple, often conflicting, objectives. The problem statement is based on the need to optimise complex powertrain systems in a more virtual development cycle to reduce the use of physical prototype vehicles and deliver more optimised powertrains in a shorter timeframe. A genetic algorithm (GA) approach was used as a baseline process. The thesis then describes a series of improvements and best practices for implementing this GA approach in the real world. The research concluded that the optimisation algorithms can be better used to solve hybrid electric vehicle development as an automated search tool, rather than an optimisation process. Based on this, multiple techniques were implemented to reduce the search space and improve the efficiency of this search, it resulted in 62% in time savings and over 39% less evaluations. Among these changes a graphic user interface to define components and topologies of interest, was developed, prioritising objectives, adding restrictive penalties (to reduce wasted simulation effort), and building a test database that can be used by the search algorithm to avoid evaluating repeated topologies. Additionally, this database serves as the outcome for the user to improve the confidence in his selection, gain insights on vehicle sensitivities, and as an historic database that can be further grown with new studies.

Data stream management systems exist to support dynamic analysis of streaming data, often to inform decision-making. Decision Support Systems (DSSs) exist to enable decisions to be made that take into account user priorities. However, although these categories of system are now quite mature, there has been little work investigating their use together. Bringing these technologies together in a way that enables trustable decision support for dynamic applications is a difficult problem with particular impact in the military and medical domains. A framework has been proposed, comprising eight desiderata for trusted dynamic decision support. These desiderata aim to inform architects of dynamic DSSs on the implications of different capacities for decision support. An approach to dynamic decision support employing Genetic Algorithms (GAs) has been proposed. Two case studies have been utilised to show how this approach can be leveraged to provision DSSs with our desiderata. Weighted Product Model (WPM), Analytic Hierarchy Process (AHP), Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) and Preference Ranking Organization METHod for Enrichment Evaluation (PROMETHEE) have been assessed on the stability of results and the consistency of trade-offs, two of our desiderata. This assessment determined TOPSIS to be the method that is the most suitable for dynamic decision support. The problem of evaluating the effect of DSS features on trust has also been addressed and a theoretical framework modelling trust and its antecedents in a real-time DSS has been proposed. This model has then been used to carry out an assessment of the impact of explanation, preferences and dynamic updates as components of dynamic decision support, giving designers of DSSs an indication of which of these features are likely to have a positive impact on decision making in a dynamic environment. Finally, the research has concluded with the identification and discussion of potential areas for future investigation.

Data synthesis is a statistical disclosure control technique that prevents the leakage of personal information from survey data. Rubin, who originally proposed this technique, treated the confidential data within a dataset as missing and then replaced those data using multiple imputation [103]. Most methods in data synthesis were then developed based on this principle. However, data synthesis is a multi-objective problem that aims to maximise information utility as well as minimising disclosure risks, and these methods have no explicit mechanism for balancing the objectives. This issue is the basis for the line of enquiry embodied in this thesis. The need to optimise competing objectives suggests the possible use of iterative machine learning techniques for data synthesis, but - to date - investigations of this possibility have been limited. In the thesis, a new synthesis method using Genetic Algorithms (GAs) is introduced. GAs are evolutionary computational methods that simulate natural evolution. They allow candidates (which in this thesis are datasets) to compete, reproduce and mate in a pre-determined environment until one or more of them perfectly fits the environment (which is defined by a set of objectives). GAs were firstly used on binary strings and now they have variants that deal with different problems and data forms. In this thesis, a GA data synthesiser whose candidates are matrix and real-coded data is designed, and most of its parameters and hyper-parameters tested. A new information utility function to measure the overall divergence from synthetic data to the original data is used. The results of running the synthesiser on a real dataset are presented, which show that the GA approach successfully produced plausible synthetic data using a single utility objective and they were proved to be able to seek for a trade-off between information utility and disclosure risks during the process of synthesising. The overall conclusion is that GAs represent a significant opportunity for the practice of data synthesis.

This critical synopsis of prior work by Luca Parisi is submitted in support of a PhD by Published Work. The work focuses on deriving accurate, reliable and explainable clinical decision support systems as minimum clinically viable solutions leveraging Machine Learning (ML) and evolutionary algorithms, for the first time, to facilitate early diagnostic predictions of Parkinson's Disease and hypothermia in hospitals, as well as prognostic predictions of optimal postoperative recovery area and of chronic hepatitis. Despite the various pathological aetiologies, the underlying capability of ML-based algorithms to serve as a minimum clinically viable solution for predicting early diagnosis and prognosis has been thoroughly demonstrated. Feature selection (FS) is a proven method for increasing the performance of ML-based classifiers for several applications. Although advances in ML, such as Deep Learning (DL), have denied the usefulness of any extrinsic FS by incorporating it in their architectures, e.g., convolutional filters in convolutional neural networks, DL algorithms often lack the required explainability to be understood and interpreted by clinicians within the context of the diagnostic and prognostic tasks of interest. Their relatively complicated architectures, the hardware required for running them and the limited explainability or interpretability of their architectures, the decision-making process - although as assistive tools - driven by the algorithms' training and predictive outcomes have hindered their application in a clinical setting. Luca Parisi's work fills this translational research gap by harnessing the explainability of using traditional ML- and evolutionary algorithms-based FS methods for improving the performance of ML-based algorithms and devise minimum viable solutions for diagnostic and prognostic purposes. The work submitted here involves independent research work, including collaborative studies with Marianne Lyne Manaog (MedIntellego®) and Narrendar RaviChandran (University of Auckland). In particular, conciliating his work as a Senior Artificial Intelligence Engineer and volunteering commitment as the President and Research Committee Leader of a student-led association named the "University of Auckland Rehabilitative Technologies Association", Luca Parisi decided to embark on most research works included in this synopsis to add value to society via accurate, reliable and explainable, hence clinically viable applications of AI. The key findings of these studies are: (i) ML-based FS algorithms are sufficient for devising accurate, reliable and explainable ML-based classifiers for aiding prediction of early diagnosis for Parkinson's Disease and chronic hepatitis; (ii) evolutionary algorithms-based optimisation is a preferred method for improving the accuracy and reliability of decision support systems aimed at aiding early diagnosis of hypothermia; (iii) evolutionary algorithms-based optimisation methods enable to devise optimised ML-based classifiers for improving postoperative discharge; (iv) whilst ML-based algorithms coupled with ML based FS methods are the minimum clinically viable solution for binary classification problems, ML-based classifiers leveraging evolutionary algorithms for FS yield more accurate and reliable predictions, as reducing the search space and overlapping regions for tackling multi-class classification problems more effectively, which involve a higher number of degrees of freedom. Collectively, these findings suggest that, despite advances in ML, state-of-the-art ML algorithms, coupled with ML-based or evolutionary algorithms for FS, are enough to devise accurate, reliable and explainable decision support systems for performing both an early diagnosis and a prediction of prognosis of various pathologies.

Procedural animation systems are capable of synthesising life-like organic motion automatically. However due to extensive parameterisation, tuning these systems can be very difficult. Not only are there potentially hundreds of interlinked parameters, the resultant animation can be very subjective, and the process is difficult to automate effectively. The research presented in this thesis is divided into three stages. Our first motivation is to examine whether artificially intelligent characters appear more or less human-like in virtual reality (VR). Our results indicate that there is a clear split in how we perceive an artificial character depending on viewing method and game type. Our second motivation is to assess whether anonymous individuals can anneal a procedurally animated creature towards a desired outcome. To do this we present an online system which used crowd-sourcing to direct a genetic algorithm. This methodology is further tested by asking users to interactively rate a population of virtual dolphins to a prescribed behavioural criterion. Our results show that within a few generations a group of users can successfully tune an animation system toward a desired behaviour. Our final motivation is to investigate if there are differences in animation and behavioural preference between observations made across 2D screens and VR. We describe a study where users tuned two sets of dolphin animation systems in parallel, one using a normal monitor and another using an Oculus Rift. Our results indicate that being immersed in VR leads to some key differences in preferred behaviour.

Enterprise efficiency is limited by data exchange. A product designer might specify the geometry of a product with a Computer Aided Design program, an engineer might re-use that geometry data to calculate physical properties of the product using a Finite Element Analysis program. These different domains place different requirements on the product representation. Representations of product data required for different tasks is dependent on the vendor software associated with those tasks, sharing data between different vendor programs is limited by incompatibility of the vendor formats used. In the case of Computer Aided Design where the virtual form of an object is modelled, no standard data format captures complete model data. Common data standards transfer model surface geometry without capturing the topological elements from which these geometries are constructed. There are prescriptive data representations to allow these features to be specified in a neutral format, but little incentive for vendors to adopt these schemes. Recent efforts instead focus on identifying similar feature elements between different vendor CAD programs, however this approach relies on onerous manual identification requiring frequent revision. This research develops methods to automate the task of mapping relationships between different data format representations. Two independent matching techniques identify similar CAD feature functions between heterogeneous programs. Text similarity and object geometry matching techniques are combined to match the data formats associated with CAD programs. An efficient search for matching function parameters is performed using a genetic algorithm that incorporates semantic data matching and geometry data matching. A greedy semantic matching algorithm is developed that compares with the Doc2vec short text matching technique over the API dataset tested. A SVD geometric surface registration technique is developed that requires fewer calculations than an equivalent Iterative Closest Point method.

Brain Computer Interfaces (BCI) are a domain of hardware/software in which a user can interact with a machine without the need for motor activity, communicating instead via signals generated by the nervous system. These interfaces provide life-altering benefits to users, and refinement will both allow their application to a much wider variety of disabilities, and increase their practicality. The primary method of acquiring these signals is Electroencephalography (EEG). This technique is susceptible to a variety of different sources of noise, which compounds the inherent problems in BCI training data: large dimensionality, low numbers of samples, and non-stationarity between users and recording sessions. Feature Selection and Transfer Learning have been used to overcome these problems, but they fail to account for several characteristics of BCI. This thesis extends both of these approaches by the use of Search-based algorithms. Feature Selection techniques, known as Wrappers use 'black box' evaluation of feature subsets, leading to higher classification accuracies than ranking methods known as Filters. However, Wrappers are more computationally expensive, and are prone to over-fitting to training data. In this thesis, we applied Iterated Local Search (ILS) to the BCI field for the first time in literature, and demonstrated competitive results with state-of-the-art methods such as Least Absolute Shrinkage and Selection Operator and Genetic Algorithms. We then developed ILS variants with guided perturbation operators. Linkage was used to develop a multivariate metric, Intrasolution Linkage. This takes into account pair-wise dependencies of features with the label, in the context of the solution. Intrasolution Linkage was then integrated into two ILS variants. The Intrasolution Linkage Score was discovered to have a stronger correlation with the solutions predictive accuracy on unseen data than Cross Validation Error (CVE) on the training set, the typical approach to feature subset evaluation. Mutual Information was used to create Minimum Redundancy Maximum Relevance Iterated Local Search (MRMR-ILS). In this algorithm, the perturbation operator was guided using an existing Mutual Information measure, and compared with current Filter and Wrapper methods. It was found to achieve generally lower CVE rates and higher predictive accuracy on unseen data than existing algorithms. It was also noted that solutions found by the MRMR-ILS provided CVE rates that had a stronger correlation with the accuracy on unseen data than solutions found by other algorithms. We suggest that this may be due to the guided perturbation leading to solutions that are richer in Mutual Information. Feature Selection reduces computational demands and can increase the accuracy of our desired models, as evidenced in this thesis. However, limited quantities of training samples restricts these models, and greatly reduces their generalisability. For this reason, utilisation of data from a wide range of users is an ideal solution. Due to the differences in neural structures between users, creating adequate models is difficult. We adopted an existing state-of-the-art ensemble technique Ensemble Learning Generic Information (ELGI), and developed an initial optimisation phase. This involved using search to transplant instances between user subsets to increase the generalisability of each subset, before combination in the ELGI. We termed this Evolved Ensemble Learning Generic Information (eELGI). The eELGI achieved higher accuracy than user-specific BCI models, across all eight users. Optimisation of the training dataset allowed smaller training sets to be used, offered protection against neural drift, and created models that performed similarly across participants, regardless of neural impairment. Through the introduction and hybridisation of search based algorithms to several problems in BCI we have been able to show improvements in modelling accuracy and efficiency. Ultimately, this represents a step towards more practical BCI systems that will provide life altering benefits to users.

The failure of oil well drillstrings due to torsional and longitudinal stresses caused by stick-slip phenomena during the drilling operation causes great expense to industry. Due to the complicated and harsh drilling environment, modelling of the drillstring becomes an essential requirement in studies. Currently, this is achieved by modelling the drillstring as a torsional lumped model (which ignores the length of the drillstring) for real-time measurement and control. In this thesis, a distributed-lumped model including the effects of drillstring length was developed to represent the drillstring, and was used to simulate stick-slip vibration. The model was developed with increasing levels of detail and the resultant models were validated against typical measured signals from the published literature. The stick-slip model describes the friction model that exists between the cutting tool and the rock. Based on theoretical analysis and mathematical formulation an efficient and adaptable model was created which was then used in the application of a method of species conserving genetic algorithm (SCGA) to optimise the drilling parameters. In conclusion, it was shown that the distributed-lumped model showed improved detail in predicting the transient response and demonstrated the importance of including the drillstring length. Predicting the response of different parameters along the drillstring is now possible and this showed the significant effect of modelling the drillcollar. The model was shown to better represent real system and was therefore far more suited to use with real time measurements.

The prevalence of digital manufacturing in creating increasingly complex products with small batch sizes, requires effective methods for production process planning. Toolpath generation is one of the challenges for manufacturing technologies that function based on the controlled movement of an end effector against a workpiece. The current approaches for determining suitable tool paths are highly dependent on machine structure, manufacturing technology and product geometry. This dependence can be very expensive in a volatile production environment where the products and the resources change quickly. In this research, a novel approach for the flexible generation of toolpaths using a mathematical formulation of the desired objective is proposed. The approach, based on optimisation techniques, is developed by discretising the product space into a number of grid points and determining the optimal sequence of the tool tip visiting these points. To demonstrate the effectiveness of the approach, the context of milling machining has been chosen and a genetic algorithm has been developed to solve the optimisation problem. The results show that with meta-heuristic methods, flexible tool paths can indeed be generated for industrially relevant parts using existing computational power. Future computing platforms, including quantum computers, could extend the applicability of the proposed approach to much more complex domains for instantaneous optimisation of the detailed manufacturing process plan.

Global climate change is one of the most significant challenges faced by humanity in the 21st century. The adoption of the Paris agreement was the latest action by the international community aimed to limit its impact, by employing a range of strategies for greenhouse gas emissions mitigation. biofuels have a high potential to reduce net CO2 emissions, which are the primary contributor to climate change. Miscanthus is a rhizomatous C4 grass, which has been identified as an ideal candidate for a biofuel crop. The development of simulation models of Miscanthus plays a crucial role in improving the crop and making it a viable product. Crop models contribute to our understanding of the biology of the plant and its interaction with the environment, and assist in decision-making in a wide range of processes, including breeding, agriculture, and environmental planning. Crop models are unified in their reliance upon phenotypic data for parameterisation. However, the collection of phenotypic data presents a bottleneck to developing models, as it is a costly and time-consuming process. This research aims to use machine learning and other computational methods to build models and reduce the cost of data collection, without sacrificing modelling accuracy. A range of machine learning models of Miscanthus were developed and their precision was validated against an established crop model and real data. An optimisation method was applied to the previously developed models, which assessed the importance of commonly measured phenotypic variables and identified crucial periods of the growing season for data collection. This research has shown that machine learning could be a powerful tool in crop modelling, for building accurate predictive models quickly and easily. It has demonstrated a methodology for minimising data collection costs, which informs experiment design. Thus, it has the potential to play an important role in Miscanthus breeding and agriculture.

Sewer and oil pipeline spillage issues have become major causes of pollution in urban and rural areas usually caused by blockages in the water storage and drainage system, and oil spillage of underground oil pipelines. An effective way of avoiding this problem will be by deploying some mechanism to monitor these installations at each point in time and reporting unusual liquid activity to the relevant authorities for prompt action to avoid a flooding or spillage occurrence. This research work presents a low cost energy efficient liquid level monitoring technique using Radio Frequency Identification Technology. Passive UHF RFID tags have been designed, modelled and optimized. A simple rectangular tag, the P-shaped tag and S-shaped tag with UHF band frequency of operation (850-950 MHz) has been designed and modelled. Detailed parametric analysis of the rectangular tag is made and the optimised design results analysed and presented in HFSS and Matlab. The optimised rectangular tag designs are then deployed as level sensors in a gully pot. Identical tags were deployed to detect 4 distinct levels in alternate positions and a few inches in seperation distance within the gully pot height (Low, Mid, High and Ultra high). The radiation characteristic of tag sensors in deployment as modelled on HFSS is observed to show consistent performance with application requirements. An in-manhole chamber antenna for an underground communication system is analysed, designed, deployed and measured. The antenna covers dual-band impedance bandwidths (i.e. 824 to 960 MHz, and 1710 to 2170 MHz). The results show that the antenna prototype exhibits sufficient impedance bandwidth, suitable radiation characteristics, and adequate gains for the required underground wireless sensor applications. Finally, a Linearly Shifted Quadrifilar Helical Antenna (LSQHA) designed using Genetic Algorithm optimisation technique for adoption as an RFID reader antenna is proposed and investigated. The new antenna confirms coverage of the RFID bandwidth 860-960 MHz with acceptable power gain of 13.1 dBi.

This doctoral thesis investigates the application of simulation-based optimization (SBO) as an alternative to conventional optimization techniques when the inherent uncertainty and complex features of real manufacturing systems need to be considered. Inspired by a real-world production planning setting, we provide a general formulation of the situation as an extended knapsack problem. We proceed by proposing a solution approach based on single and multi-objective SBO models, which use simulation to capture the uncertainty and complexity of the manufacturing system and employ meta-heuristic optimizers to search for near-optimal solutions. Moreover, we consider the design of matheuristic approaches that combine the advantages of population-based meta-heuristics with mathematical programming techniques. More specifically, we consider the integration of mathematical programming techniques during the initialization stage of the single and multi-objective approaches as well as during the actual search process. Using data collected from a manufacturing company, we provide evidence for the advantages of our approaches over conventional methods (integer linear programming and chance-constrained programming) and highlight the synergies resulting from the combination of simulation, meta-heuristics and mathematical programming methods. In the context of the same real-world problem, we also analyse different single and multi-objective SBO models for robust optimization. We demonstrate that the choice of robustness measure and the sample size used during fitness evaluation are crucial considerations in designing an effective multi-objective model.

For many decisions validation of their outcomes is invariably problematic to objectively assess. Therefore to aid analysis and validation of decision outcomes, approaches which provide improved traceability and more semantically meaningful measurements of the decision process are required. Hence, this research investigates traceability, transparency, interactivity and auditability to improve the decision making process. Approaches and evaluation measures are proposed to facilitate a richer decision making experience. Multi-Criteria Decision Analysis (MCDA) seeks to determine the suitability of alternatives of a goal with respect to multiple criteria. A key component of prominent MCDA methods is the concept of pairwise comparison. For a set of elements, pairwise comparison enables an accurate and transparent extraction and codification of a decision maker’s preferences, though facilitating a separation of concerns. From a set of pairwise comparisons, a ranking of the elements under consideration can be calculated. There are scenarios when a set of pairwise comparisons undergo alteration, both for individual and multiple decision makers. A set of measures of compromise are proposed to quantify the alteration that a set of pairwise comparisons undergo in such scenarios. The measures seek to provide a decision maker with meaningful knowledge regarding how their views have altered. A set of pairwise comparisons may be inconsistent. When inconsistency is present it adversely affects a ranking of the elements derived from the comparisons. Moreover inconsistency within pairwise comparisons used for consideration of more than a handful of elements is almost inevitable. Existing approaches that seek to alter a set of comparisons to reduce inconsistency lack traceability, flexibility, and specific consideration of alteration to the judgments in a way that is meaningful to a decision maker. An approach to inconsistency reduction is proposed that seeks to address these issues. For many decisions the opinions of multiple decision makers are utilized, either to avail of their combined expertise or to incorporate conflicting views. Aggregation of multiple decision makers’ pairwise companions seek to combine the views of the group into a single representation of views. An approach to group aggregation of pairwise comparisons is proposed that models compromise between the decision makers, facilitates decision maker constraints, considers inconsistency reduction during aggregation and dynamically incorporates decision maker weights of importance. With internet access becoming widespread being able to garner the views of a large group of decision makers’ views has become feasible. An approach to the aggregation of a large group of decision makers’ preferences is proposed. The approach facilitates understanding regarding both the agreement and conflict within the group during calculation of an overall group consensus. A Multi-Objective Optimisation Decision Software (MOODS) prototype tool has been developed that implements both the new measures of compromise and the proposed approaches to inconsistency reduction and group aggregation.

Methodologies designed to support new product development are receiving increasing interest in recent literature. A significant percentage of new product failure is attributed to a mismatch between designed product features and consumer liking. A variety of methodologies have been proposed and tested for consumer liking or preference prediction, ranging from statistical methodologies e.g. multiple linear regression (MLR) to non-statistical approaches e.g. artificial neural networks (ANN), support vector machines (SVM), and belief rule-based (BRB) systems. BRB has been previously tested for consumer preference prediction and target setting in case studies from the beverages industry. Results have indicated a number of technical and conceptual advantages which BRB holds over the aforementioned alternative approaches. This thesis focuses on presenting further advantages and applications of the BRB methodology for consumer liking prediction. The features and advantages are selected in response to challenges raised by three addressed case studies. The first case study addresses a novel industry for BRB application: the fast moving consumer goods industry, the personal care sector. A series of challenges are tackled. Firstly, stepwise linear regression, principal component analysis and AutoEncoder are tested for predictors’ selection and data reduction. Secondly, an investigation is carried out to analyse the impact of employing complete distributions, instead of averages, for sensory attributes. Moreover, the effect of modelling instrumental measurement error is assessed. The second case study addresses a different product from the personal care sector. A bi-objective prescriptive approach for BRB model structure selection and validation is proposed and tested. Genetic Algorithms and Simulated Annealing are benchmarked against complete enumeration for searching the model structures. A novel criterion based on an adjusted Akaike Information Criterion is designed for identifying the optimal model structure from the Pareto frontier based on two objectives: model complexity and model fit. The third case study introduces yet another novel industry for BRB application: the pastry and confectionary specialties industry. A new prescriptive framework, for rule validation and random training set allocation, is designed and tested. In all case studies, the BRB methodology is compared with the most popular alternative approaches: MLR, ANN, and SVM. The results indicate that BRB outperforms these methodologies both conceptually and in terms of prediction accuracy.

This thesis describes the approach and results of the research to create a TRIZ based computer aided innovation tools (AEGIS and Design for Wow). This research has mainly been based around two tools created under this research: called AEGIS (Accelerated Evolutionary Graphics Interface System), and Design for Wow. Both of these tools are discussed in this thesis in detail, along with the test data, design methodology, test cases, and research. Design for Wow (http://www.designforwow.com) is an attempt to summarize the successful inventions/ designs from all over the world on a web portal which has multiple capabilities. These designs/innovations are then linked to the TRIZ Principles in order to determine whether innovative aspects of these successful innovations are fully covered by the forty TRIZ principles. In Design for Wow, a framework is created which is implemented through a review tool. The Design for Wow website includes this tool which has been used by researcher and the users of the site and reviewers to analyse the uploaded data in terms of strength of TRIZ Principles linked to them. AEGIS (Accelerated Evolutionary Graphics Interface System) is a software tool developed under this research aimed to help the graphic designers to make innovative graphic designs. Again it uses the forty TRIZ Principles as a set of guiding rules in the software. AEGIS creates graphic design prototypes according to the user input and uses TRIZ Principles framework as a guide to generate innovative graphic design samples. The AEGIS tool created is based on TRIZ Principles discussed in Chapter 3 (a subset of them). In AEGIS, the TRIZ Principles are used to create innovative graphic design effects. The literature review on innovative graphic design (in chapter 3) has been analysed for links with TRIZ Principles and then the DNA of AEGIS has been built on the basis of this study. Results from various surveys/ questionnaires indicated were used to collect the innovative graphic design samples and then TRIZ was mapped to it (see section 3.2). The TRIZ effects were mapped to the basic graphic design elements and the anatomy of the graphic design letters was studied to analyse the TRIZ effects in the collected samples. This study was used to build the TRIZ based AEGIS tool. Hence, AEGIS tool applies the innovative effects using TRIZ to basic graphic design elements (as described in section 3.3). the working of AEGIS is designed based on Genetic Algorithms coded specifically to implement TRIZ Principles specialized for Graphic Design, chapter 4 discusses the process followed to apply TRIZ Principles to graphic design and coding them using Genetic Algorithms, hence resulting in AEGIS tool. Similarly, in Design for Wow, the content uploaded has been analysed for its link with TRIZ Principles (see section 3.1 for TRIZ Principles). The tool created in Design for Wow is based on the framework of analysing the TRIZ links in the uploaded content. The ‘Wow’ concept discussed in the section 5.1 and 5.2 is the basis of the concept of Design for Wow website, whereby the users upload the content they classify as ‘Wow’. This content then is further analysed for the ‘Wow factor’ and then mapped to TRIZ Principles as TRIZ tagging methodology is framed (section 5.5). From the results of the research, it appears that the TRIZ Principles are a comprehensive set of innovation basic building blocks. Some surveys suggest that amongst other tools, TRIZ Principles were the first choice and used most .They have thus the potential of being used in other innovation domains, to help in their analysis, understanding and potential development.

This thesis discusses the development and performance of a novel evolutionary algorithm, the Targeted Estimation of Distribution Algorithm (TEDA). TEDA takes the concept of targeting, an idea that has previously been shown to be effective as part of a Genetic Algorithm (GA) called Fitness Directed Crossover (FDC), and introduces it into a novel hybrid algorithm that transitions from a GA to an Estimation of Distribution Algorithm (EDA). Targeting is a process for solving optimisation problems where there is a concept of control points, genes that can be said to be active, and where the total number of control points found within a solution is as important as where they are located. When generating a new solution an algorithm that uses targeting must first of all choose the number of control points to set in the new solution before choosing which to set. The hybrid approach is designed to take advantage of the ability of EDAs to exploit patterns within the population to effectively locate the global optimum while avoiding the tendency of EDAs to prematurely converge. This is achieved by initially using a GA to effectively explore the search space before transitioning into an EDA as the population converges on the region of the global optimum. As targeting places an extra restriction on the solutions produced by specifying their size, combining it with the hybrid approach allows TEDA to produce solutions that are of an optimal size and of a higher quality than would be found using a GA alone without risking a loss of diversity. TEDA is tested on three different problem domains. These are optimal control of cancer chemotherapy, network routing and Feature Subset Selection (FSS). Of these problems, TEDA showed consistent advantage over standard EAs in the routing problem and demonstrated that it is able to find good solutions faster than untargeted EAs and non evolutionary approaches at the FSS problem. It did not demonstrate any advantage over other approaches when applied to chemotherapy. The FSS domain demonstrated that in large and noisy problems TEDA’s targeting derived ability to reduce the size of the search space significantly increased the speed with which good solutions could be found. The routing domain demonstrated that, where the ideal number of control points is deceptive, both targeting and the exploitative capabilities of an EDA are needed, making TEDA a more effective approach than both untargeted approaches and FDC. Additionally, in none of the problems was TEDA seen to perform significantly worse than any alternative approaches.

This thesis presents a comprehensive work on computerised analysis of fetal heart rate (FHR) features, including feature extraction, feature selection, analysis of influencing factors and setting up/validation of a computerised decision support system. Firstly, a novel feature – pattern readjustment – was extracted and tested. Clinical data were used to train a Support Vector Machine (SVM) to detect pattern readjustment. Then, the association of pattern readjustment and adverse labour outcome was investigated. The validation results with clinical experts show that the pattern readjustment can be accurately detected, while the study on labour outcome shows that the feature is related to fetal acidemia at birth. Secondly, Genetic Algorithms were employed as a feature selection method to select a best subset of FHR features and to use them to predict fetal acidemia with linear and nonlinear SVM. The diagnostic power of the classifier output using selected features was tested on the total set of 7,568 cases. As the classifier output increases, there is a consistent increase of the risk of fetal acidemia. Thirdly, an important influencing factor on FHR features - signal loss – was investigated. A bivariate model was built to estimate error based on signal loss. Validation results show that the bivariate model can accurately predict the error generated by signal loss. The influence of signal loss on labour outcome classification was also investigated. Finally, a computerised decision support system to estimate the risk of fetal acidemia was set up based on the above studies. The system was validated using new retrospective data. Validation results show that the system is capable of predicting adverse labour outcome and providing timely decision support. It is the first time an intrapartum computerised FHR decision support system has been built and validated on this size of dataset. With further improvements, such a system could be implemented clinically in the long term.

Financial contagion refers to a scenario in which small shocks, which initially affect only a few financial institutions or a particular region of the economy, spread to the rest of the financial sector and other countries whose economies were previously healthy. This resembles the “transmission” of a medical disease. Financial contagion happens both at domestic level and international level. At domestic level, usually the failure of a domestic bank or financial intermediary triggers transmission by defaulting on inter-bank liabilities, selling assets in a fire sale, and undermining confidence in similar banks. An example of this phenomenon is the failure of Lehman Brothers and the subsequent turmoil in the US financial markets. International financial contagion happens in both advanced economies and developing economies, and is the transmission of financial crises across financial markets. Within the current globalise financial system, with large volumes of cash flow and cross-regional operations of large banks and hedge funds, financial contagion usually happens simultaneously among both domestic institutions and across countries. There is no conclusive definition of financial contagion, most research papers study contagion by analyzing the change in the variance-covariance matrix during the period of market turmoil. King and Wadhwani (1990) first test the correlations between the US, UK and Japan, during the US stock market crash of 1987. Boyer (1997) finds significant increases in correlation during financial crises, and reinforces a definition of financial contagion as a correlation changing during the crash period. Forbes and Rigobon (2002) give a definition of financial contagion. In their work, the term interdependence is used as the alternative to contagion. They claim that for the period they study, there is no contagion but only interdependence. Interdependence leads to common price movements during periods both of stability and turmoil. In the past two decades, many studies (e.g. Kaminsky et at., 1998; Kaminsky 1999) have developed early warning systems focused on the origins of financial crises rather than on financial contagion. Further authors (e.g. Forbes and Rigobon, 2002; Caporale et al, 2005), on the other hand, have focused on studying contagion or interdependence. In this thesis, an overall mechanism is proposed that simulates characteristics of propagating crisis through contagion. Within that scope, a new co-evolutionary market model is developed, where some of the technical traders change their behaviour during crisis to transform into herd traders making their decisions based on market sentiment rather than underlying strategies or factors. The thesis focuses on the transformation of market interdependence into contagion and on the contagion effects. The author first build a multi-national platform to allow different type of players to trade implementing their own rules and considering information from the domestic and a foreign market. Traders’ strategies and the performance of the simulated domestic market are trained using historical prices on both markets, and optimizing artificial market’s parameters through immune - particle swarm optimization techniques (I-PSO). The author also introduces a mechanism contributing to the transformation of technical into herd traders. A generalized auto-regressive conditional heteroscedasticity - copula (GARCH-copula) is further applied to calculate the tail dependence between the affected market and the origin of the crisis, and that parameter is used in the fitness function for selecting the best solutions within the evolving population of possible model parameters, and therefore in the optimization criteria for contagion simulation. The overall model is also applied in predictive mode, where the author optimize in the pre-crisis period using data from the domestic market and the crisis-origin foreign market, and predict in the crisis period using data from the foreign market and predicting the affected domestic market.

Process Algebras are a Formal Modelling methodology which are an effective tool for defining models of complex systems, particularly those involving multiple interacting processes. However, describing such a model using Process Algebras requires expertise from both the modeller and the domain expert. Finding the correct model to describe a system can be difficult. Further more, even with the correct model, parameter tuning to allow model outputs to match experimental data can also be both difficult and time consuming. Evolutionary Algorithms provide effective methods for finding solutions to optimisation problems with large and noisy search spaces. Evolutionary Algorithms have been proven to be well suited to investigating parameter fitting problems in order to match known data or desired behaviour. It is proposed that Process Algebras and Evolutionary Algorithms have complementary strengths for developing models of complex systems. Evolutionary Algorithms require a precise and accurate fitness function to score and rank solutions. Process Algebras can be incorporated into the fitness function to provide this mathematical score. Presented in this work is the Evolving Process Algebra (EPA) framework, designed for the application of Evolutionary Algorithms (specifically Genetic Algorithms and Genetic Programming optimisation techniques) to models described in Process Algebra (specifically PEPA and Bio-PEPA) with the aim of evolving fitter models. The EPA framework is demonstrated using multiple complex systems. For PEPA this includes the dining philosophers resource allocation problem, the repressilator genetic circuit, the G-protein cellular signal regulators and two epidemiological problems: HIV and the measles virus. For Bio-PEPA the problems include a biochemical reactant-product system, a generic genetic network, a variant of the G-protein system and three epidemiological problems derived from the measles virus. Also presented is the EPA Utility Assistant program; a lightweight graphical user interface. This is designed to open the full functionality and parallelisation of the EPA framework to beginner or naive users. In addition, the assistant program aids in collating and graphing after experiments are completed.

The implementation of a low carbon electricity system within the GB requires a significant change to the generation mix with an increasing role for renewable generation. Much of this generation will be intermittent. To date system balancing has largely relied on predicting demand and ensuring provision. With substantial intermittency, continuation of this paradigm necessitates significant investment in peaking plant and/or storage. However, some of this investment can be avoided by harnessing the flexibility inherent in many electrical loads. Despite the attractiveness of such services, we do not see their large-scale implementation. The aim of this thesis is to consider why. A historical analysis reveals that both nationalisation and subsequent privatisation provide precedents for significant structural change as the integration of large-scale flexible demand might require. The need for political will is identified as a crucial enabling factor. Without an ideological driver, however, a perception of economic and/or technological risk can preclude the implementation of supportive policy. This perception is addressed through demonstration. An effective demonstration must show the ability to aggregate many small loads in a coordinated manner. A genetic algorithm that provides this core dispatch and optimisation capability is presented. This algorithm is shown to be effective in aggregating many small loads to provide a net effect that can be used as a balancing service and to do so in an optimal way considering both cost and reliability. Having demonstrated feasibility appropriate incentives must be created. An initial outline for a framework based on SysML is presented that can be used to identify where structural barriers to implementation are present to aid the design of appropriate policy incentives.

The work presented herein focuses on the design of distributed autonomous controllers for collective behaviour of Micro-unmanned Aerial Vehicles (MAVs). Two alternative approaches to this topic are introduced: one based upon the Evolutionary Robotics (ER) paradigm, the other one upon flocking principles. Three computer simulators have been developed in order to carry out the required experiments, all of them having their focus on the modelling of fixed-wing aircraft flight dynamics. The employment of fixed-wing aircraft rather than the omni-directional robots typically employed in collective robotics significantly increases the complexity of the challenges that an autonomous controller has to face. This is mostly due to the strict motion constraints associated with fixed-wing platforms, that require a high degree of accuracy by the controller. Concerning the ER approach, the experimental setups elaborated have resulted in controllers that have been evolved in simulation with the following capabilities: (1) navigation across unknown environments, (2) obstacle avoidance, (3) tracking of a moving target, and (4) execution of cooperative and coordinated behaviours based on implicit communication strategies. The design methodology based upon flocking principles has involved tests on computer simulations and subsequent experimentation on real-world robotic platforms. A customised implementation of Reynolds’ flocking algorithm has been developed and successfully validated through flight tests performed with the swinglet MAV. It has been notably demonstrated how the Evolutionary Robotics approach could be successfully extended to the domain of fixed-wing aerial robotics, which has never received a great deal of attention in the past. The investigations performed have also shown that complex and real physics-based computer simulators are not a compulsory requirement when approaching the domain of aerial robotics, as long as proper autopilot systems (taking care of the ”reality gap” issue) are used on the real robots.

A human thermoregulation model has been adapted for predicting the thermal response of Typical Older Persons. The model known as the Older Persons Model predicts the core body temperature and regulatory responses of the older people in environmental exposures of cold, warm and hot. The model was developed by modifying an existing dynamic human thermoregulation model using anthropometric and thermo-physical properties of older people. The Model defines the body as two interrelating systems of the body structure (passive system) and the control system of the central nervous system (active system). The Older person's passive system of the model was developed by meticulously extracting relevant experimental data from selected published research works relating to anthropometric and thermo-physical properties of older people. The resultant body structure (passive system) is a multi-segmented representation of a Typical Older Person. The active system (central nervous system) was developed by the application of a novel optimization method based on the working principles of Genetic Algorithms. The use of Genetic Algorithm enables the complex characteristics of the central nervous system of the older persons to be well represented and evaluated based on available data. Active system control signal coefficients for sweating, shivering, vasodilation and vasoconstriction were explicitly derived based on experimental data sourced from literature. The Older Persons Model has been validated using independent experimental data and its results show good agreement with measured data. Furthermore, the Older Persons Model has been applied to several test cases extracted from published literature and its results show good agreement with published findings on the thermal behaviour of older persons. An interview study conducted as part of this research revealed that, professionals (built environment specialists) found the Older Persons Model useful in assisting to further understand the thermal response of the older persons. In conclusion, the adaptation of an existing human thermoregulation model has resulted in a new model, which allows improved prediction of heat and cold strain of the older person although there exist limitations.

In the area of artificial intelligence, the development of Evolutionary Algorithms (EAs) has been very active, especially in the last decade. These algorithms started to evolve when scientists from various regions of the world applied the principles of evolution to algorithmic search and problem solving. EAs have been utilised successfully in diverse complex application areas. Their success in tackling hard problems has been the engine of the field of Evolutionary Computation (EC). Nowadays, EAs are considered to be the best solution to use when facing a hard search or optimisation problem. Various improvements are continually being made with the design of new operators, hybrid models, among others. A very important example of such improvements is the use of parallel models of GAs (PGAs). PGAs have received widespread attention from various researchers as they have proved to be more effective than panmictic GAs, especially in terms of efficacy and speedup. This thesis focuses on, and investigates, cellular Genetic Algorithms (cGAs)-a competitive variant of parallel GAs. In a cGA, the tentative solutions evolve in overlapped neighbourhoods, allowing smooth diffusion of the solutions. The benefits derived from using cGAs come not only from flexibility gains and their fitness to the objective target in combination with a robust behaviour but also from their high performance and amenability to implementation using advanced custom silicon chip technologies. Nowadays, cGAs are considered as adaptable concepts for solving problems, especially complex optimisation problems. Due to their structural characteristics, cGAs are able to promote an adequate exploration/exploitation trade-off and thus maintain genetic diversity. Moreover, cGAs are characterised as being massively parallel and easy to implement. The structural characteristics inherited in a cGA provide an active area for investigation. Because of the vital role grid structure plays in determining the effectiveness of the algorithm, cellular dimensionality is the main issue to be investigated here. The implementation of cGAs is commonly carried out on a one- or two-dimensional structure. Studies that investigate higher cellular dimensions are lacking. Accordingly, this research focuses on cGAs that are implemented on a three-dimensional structure. Having a structure with three dimensions, specifically a cubic structure, facilitates faster spreading of solutions due to the shorter radius and denser neighbourhood that result from the vertical expansion of cells. In this thesis, a comparative study of cellular dimensionality is conducted. Simulation results demonstrate higher performance achieved by 3D-cGAs over their 2D-cGAs counterparts. The direct implementation of 3D-cGAs on the new advanced 3D-IC technology will provide added benefits such as higher performance combined with a reduction in interconnection delays, routing length, and power consumption. The maintenance of system reliability and availability is a major concern that must be addressed. A system is likely to fail due to either hard or soft errors. Therefore, detecting a fault before it deteriorates system performance is a crucial issue. Single Event Upsets (SEUs), or soft errors, do not cause permanent damage to system functionality, and can be handled using fault-tolerant techniques. Existing fault-tolerant techniques include hardware or software fault tolerance, or a combination of both. In this thesis, fault-tolerant techniques that mitigate SEUs at the algorithmic level are explored and the inherent abilities of cGAs to deal with these errors are investigated. A fault-tolerant technique and several mitigation techniques are also proposed, and faulty critical data are evaluated critical fault scenarios (stuck at ‘1’ and stuck at ‘0’ faults) are taken into consideration. Chief among several test and real world problems is the problem of determining the attitude of a vehicle using a Global Positioning System (GPS), which is an example of hard real-time application. Results illustrate the ability of cGAs to maintain their functionality and give an adequate performance even with the existence of up to 40% errors in fitness score cells. The final aspect investigated in this thesis is the dynamic characteristic of cGAs. cGAs, and EAs in general, are known to be stochastic search techniques. Hence, adaptive systems are required to continue to perform effectively in a changing environment, particularly when tackling real-world problems. The adaptation in cellular engines is mainly achieved through dynamic balancing between exploration and exploitation. This area has received considerable attention from researchers who focus on improving the algorithmic performance without incurring additional computational effort. The structural properties and the genetic operations provide ways to control selection pressure and, as a result, the exploration/exploitation trade-off. In this thesis, the genetic operations of cGAs, particularly the selection aspect and their influence on the search process, are investigated in order to dynamically control the exploration/exploitation trade-off. Two adaptive-dynamic techniques that use genetic diversity and convergence speeds to guide the search are proposed. Results obtained by evaluating the proposed approaches on a test bench of diverse-characteristic real-world and test problems showed improvement in dynamic cGAs performance over their static counterparts and other dynamic cGAs. For example, the proposed Diversity-Guided 3D-cGA outperformed all the other dynamic cGAs evaluated by obtaining a higher search success rate that reached to 55%.

The objectives of this work were to investigate, design and implement beam steering antenna arrays for mobile and wireless applications using the genetic algorithm (GA) and particle swarm optimisation (PSO) techniques as optimisation design tools. Several antenna designs were implemented and tested: initially, a printed dipole antenna integrated with a duplex RF switch used for mobile base station antenna beam steering was investigated. A coplanar waveguide (CPW) to coplanar strip (CPS) transition was adopted to feed the printed dipole. A novel RF switch circuit, used to control the RF signal fed to the dipole antenna and placed directly before it, was proposed. The measured performance of the RF switch was tested and the results confirmed its viability. Then two hybrid coupled PIN diode phase shifters, using Branchline and Rat-Race ring coupler structures, were designed and tested. The generation of four distinct phase shifts was implemented and studied. The variations of the scattering parameters were found to be realistic, with an acceptable ±2 phase shift tolerance. Next, antenna beam steering was achieved by implementing RF switches with ON or OFF mode functions to excite the radiating elements of the antenna array. The switching control process was implemented using a genetic algorithm (GA) method, subject to scalar and binary genes. Anti-phase feeding of radiating elements was also investigated. A ring antenna array with reflectors was modelled and analysed. An antenna of this type for mobile base stations was designed and simulation results are presented. Following this, a novel concept for simple beam steering using a uniform antenna array operated at 2.4 GHz was designed using GA. The antenna is fed by a single RF input source and the steering elements are reactively tuned by varactor diodes in series with small inductors. The beam-control procedure was derived through the use of a genetic algorithm based on adjusting the required reactance values to obtain the optimum solution as indicated by the cost function. The GA was also initially used as an optimisation tool to derive the antenna design from its specification. Finally, reactive loading and time modulated switching techniques are applied to steer the beam of a circular uniformly spaced antenna array having a source element at its centre. Genetic algorithm (GA) and particle swarm optimisation (PSO) processes calculate the optimal values of reactances loading the parasitic elements, for which the gain can be optimised in a desired direction. For time modulated switching, GA and PSO also determine the optimal on and off times of the parasitic elements for which the difference in currents induced optimises the gain and steering of the beam in a desired direction. These methods were demonstrated by investigating a vertically polarised antenna configuration. A prototype antenna was constructed and experimental results compared with the simulations. Results showed that near optimal solutions for gain optimisation, sidelobe level reduction and beam steering are achievable by utilising these methods. In addition, a simple switching process is employed to steer the beam of a horizontally polarised circular antenna array. A time modulated switching process is applied through Genetic Algorithm optimisation. Several model examples illustrate the radiation beams and the switching time process of each element in the array.

The past two decades have witnessed tremendous advances in the field of high accuracy, high mass resolution data acquisition of complex samples such as crude oils and the human proteome. With the development of Fourier transform ion cyclotron resonance mass spectrometry, the rapidly growing field of petroleomics has emerged, whose goal is to process and analyse the large volumes of complex and often poorly understood data on crude oils generated by mass spectrometry. As global oil resources deplete, oil companies are increasingly moving towards the extraction and refining of the still plentiful reserves of heavy, carbon rich and highly contaminated crude oil. It is essential that the oil industry gather the maximum possible amount of information about the crude oil prior to setting up the drilling infrastructure, in order to reduce processing costs. This project describes how machine learning can be used as a novel way to extract critical information from complex mass spectra which will aid in the processing of crude oils. The thesis discusses the experimental methods involved in acquiring high accuracy mass spectral data for a large and key industry-standard set of crude oil samples. These data are subsequently analysed to identify possible links between the raw mass spectra and certain physical properties of the oils, such as pour point and sulphur content. Methods including artificial neural networks and self organising maps are described and the use of spectral clustering and pattern recognition to classify crude oils is investigated. The main focus of the research, the creation of an original simulated annealing genetic algorithm hybrid technique (SAGA), is discussed in detail and the successes of modelling a number of different datasets using all described methods are outlined. Despite the complexity of the underlying mass spectrometry data, which reflects the considerable chemical diversity of the samples themselves, the results show that physical properties can be modelled with varying degrees of success. When modelling pour point temperatures, the artificial neural network achieved an average prediction error of less than 10% while SAGA predicted the same values with an average accuracy of more than 85%. It did not prove possible to model any of the other properties with such statistical significance; however improvements to feature extraction and pre-processing of the spectral data as well as enhancement of the modelling techniques should yield more consistent and statistically reliable results. These should in due course lead to a comprehensive model which the oil industry can use to process crude oil data using rapid and cost effective analytical methods.

Fuzzy association rule mining discovers patterns in transactions, such as shopping baskets in a supermarket, or Web page accesses by a visitor to a Web site. Temporal patterns can be present in fuzzy association rules because the underlying process generating the data can be dynamic. However, existing solutions may not discover all interesting patterns because of a previously unrecognised problem that is revealed in this thesis. The contextual meaning of fuzzy association rules changes because of the dynamic feature of data. The static fuzzy representation and traditional search method are inadequate. The Genetic Iterative Temporal Fuzzy Association Rule Mining (GITFARM) framework solves the problem by utilising flexible fuzzy representations from a fuzzy rule-based system (FRBS). The combination of temporal, fuzzy and itemset space was simultaneously searched with a genetic algorithm (GA) to overcome the problem. The framework transforms the dataset to a graph for efficiently searching the dataset. A choice of model in fuzzy representation provides a trade-off in usage between an approximate and descriptive model. A method for verifying the solution to the hypothesised problem was presented. The proposed GA-based solution was compared with a traditional approach that uses an exhaustive search method. It was shown how the GA-based solution discovered rules that the traditional approach did not. This shows that simultaneously searching for rules and membership functions with a GA is a suitable solution for mining temporal fuzzy association rules. So, in practice, more knowledge can be discovered for making well-informed decisions that would otherwise be lost with a traditional approach.

The main objective of this research is to design, model and implement several antenna geometries using electromagnetic band gap (EBG) material and a defected ground plane. Several antenna applications are addressed with the aim of improving performance, particularly the mutual coupling between the elements. The EBG structures have the unique capability to prevent or assist the propagation of electromagnetic waves in a specific band of frequencies, and have been incorporated here in antenna structures to improve patterns and reduce mutual coupling in multielement arrays. A neutralization technique and defected ground plane structures have also been investigated as alternative approaches, and may be more practical in real applications. A new Uni-planar Compact EBG (UC-EBG) formed from a compact unit cell was presented, giving a stop band in the 2.4 GHz WLAN range. Dual band forms of the neutralization and defected ground plane techniques have also been developed and measured. The recorded results for all antenna configurations show good improvement in terms of the mutual coupling effect. The MIMO antenna performance with EBG, neutralization and defected ground of several wireless communication applications were analysed and evaluated. The correlation coefficient, total active reflection coefficient (TARC), channel capacity and capacity loss of the array antenna were computed and the results compared to measurements with good agreement. In addition, a computational method combining Genetic Algorithm (GA) with surface meshing code for the analysis of a 2×2 antenna arrays on EBG was developed. Here the impedance matrix resulting from the meshing analysis is manipulated by the GA process in order to find the optimal antenna and EBG operated at 2.4 GHz with the goal of targeting a specific fitness function. Furthermore, an investigation of GA on 2×2 printed slot on DGS was also done.

Typically, systems biology focuses on the form and function of networks of biochemical interactions. Questions inevitably arise as to the evolutionary origin of those networks' properties. Such questions are of interest to a growing number of systems biologists, and several groups have published studies shown how varying environments can affect network topology and lead to increased evolvability. For decades, evolutionary biologists have also investigated the evolution of evolvability and its relationship to the interactions between genotype and phenotype. While the perspectives of systems and evolutionary biologists sometimes differ, their interests in patterns of interactions and evolvability have much in common. This thesis attempts to bring together the perspectives of systems and evolutionary theory to investigate the evolutionary effects of fluctuating environments. Chapter 1 introduces the necessary themes, terminology and literature from these fields. Chapter 2 explores how rapid environmental fluctuations, or "noise", affects network size and robustness. In Chapter 3, we use the Avida platform to investigate the relationship between genetic architecture, fluctuating environments and population biology. Chapter 4 examines contingency loci as a physical basis for evolvability, while chapter 5 presents a 500-generation laboratory evolution experiment which exposes E. coli to varying environments. The final discussion, concludes that the evolution of generalism can lead to genetic architectures which confer evolvability, which may arise in rapidly fluctuating environments as a by-product of generalism rather than as a selected trait.

The purpose of this thesis is to present solutions to overcome several key difficulties that limit the application of numerical modelling in communication cable design and analysis. In particular, specific limiting factors are that simulations are time consuming, and the process of comparison requires skill and is poorly defined and understood. When much of the process of design consists of optimisation of performance within a well defined domain, the use of artificial intelligence techniques may reduce or remove the need for human interaction in the design process. The automation of human processes allows round-the-clock operation at a faster throughput. Achieving a speedup would permit greater exploration of the possible designs, improving understanding of the domain. This thesis presents work that relates to three facets of the efficiency of numerical modelling: minimizing simulation execution time, controlling optimization processes and quantifying comparisons of results. These topics are of interest because simulation times for most problems of interest run into tens of hours. The design process for most systems being modelled may be considered an optimisation process in so far as the design is improved based upon a comparison of the test results with a specification. Development of software to automate this process permits the improvements to continue outside working hours, and produces decisions unaffected by the psychological state of a human operator. Improved performance of simulation tools would facilitate exploration of more variations on a design, which would improve understanding of the problem domain, promoting a virtuous circle of design. The minimization of execution time was achieved through the development of a Parallel TLM Solver which did not use specialized hardware or a dedicated network. Its design was novel because it was intended to operate on a network of heterogeneous machines in a manner which was fault tolerant, and included a means to reduce vulnerability of simulated data without encryption. Optimisation processes were controlled by genetic algorithms and particle swarm optimisation which were novel applications in communication cable design. The work extended the range of cable parameters, reducing conductor diameters for twisted pair cables, and reducing optical coverage of screens for a given shielding effectiveness. Work on the comparison of results introduced ―Colour maps‖ as a way of displaying three scalar variables over a two-dimensional surface, and comparisons were quantified by extending 1D Feature Selective Validation (FSV) to two dimensions, using an ellipse shaped filter, in such a way that it could be extended to higher dimensions. In so doing, some problems with FSV were detected, and suggestions for overcoming these presented: such as the special case of zero valued DC signals. A re-description of Feature Selective Validation, using Jacobians and tensors is proposed, in order to facilitate its implementation in higher dimensional spaces.

There is a growing need to manage water resources in a sustainable way, particularly in semi arid areas, with dramatic social and economic development as well as rapid population growth. Optimising water allocation in a river basin is an important aspect ensuring equitable and efficient water use. This research develops an optimisation approach (the Integrated Water Resource Optimisation model, IWRO) to optimise the conjunctive use of surface water and groundwater resources in a sustainable manner. The IWRO model is comprised of a surface water optimisation model (SWO) and the Tsinghua groundwater optimisation (TGO) model. These models employ Genetic Algorithms (GAs) to optimise water allocation. Application of a surface water optimisation (SWO) model incorporating a GA is demonstrated initially for a simple test case, through which the GA approach was validated against known solutions. Sensitivity analysis of different operators and parameters related to GAs was also carried out. The validated SWO model was then applied to a more complex system, the Shiyang River Basin in Gansu Province in China, to maximise equitable surface water supplies. On the groundwater side, the GA approach was applied with the existing Tsinghua groundwater model to optimise groundwater supplies with sustainability considerations. The results were compared with those from an existing model (the WEAP model), indicating that the IWRO model is capable of satisfying the objectives of equitable water allocation and groundwater sustainability set for it. In the context of Integrated Water Resources Management (IWRM), account must be taken of a wide range of social and environmental issues. Different scenarios were therefore designed for the Shiyang River Basin management. Various criteria in terms of economic, social, environment and water security were also indentified for further multi-criterion decision making analysis.

In recent years, the potential benefits of high-throughput virtual screening to the drug discovery community have been recognized, bringing an increase in the number of tools developed for this purpose. These programs have to process large quantities of data, searching for an optimal solution in a vast combinatorial range. This is particularly the case for protein-ligand docking, since proteins are sophisticated structures with complicated interactions for which either molecule might reshape itself. Even the very limited flexibility model to be considered here, using ligand conformation ensembles, requires six dimensions of exploration - three translations and three rotations - per rigid conformation. The functions for evaluating pose suitability can also be complex to calculate. Consequently, the programs being written for these biochemical simulations are extremely resource-intensive. This work introduces a pure computer science approach to the field, developing techniques to improve the effectiveness of such tools. Their architecture is generalized to an abstract pattern of nested layers for discussion, covering scoring functions, search methods, and screening overall. Based on this, new stratagems for molecular docking software design are described, including lazy or partial evaluation, geometric analysis, and parallel processing implementation. In addition, a range of novel algorithms are presented for applications such as active site detection with linear complexity (PIES) and small molecule shape description (PASTRY) for pre-alignment of ligands. The various stratagems are assessed individually and in combination, using several modified versions of an existing docking program, to demonstrate their benefit to virtual screening in practical contexts. In particular, the importance of appropriate precision in calculations is highlighted.

Self-assembly --- by which ordered structures spontaneously emerge from disordered components --- and biological evolution --- by which Charles Darwin's "endless forms most beautiful" have emerged from the simple chemistry of prehistoric Earth --- may both be pictured as search processes on high-dimensional landscapes, defined respectively by the concepts of energy and fitness. The rich dynamics of these search processes will be studied in order to explain biologically observed features of self-assembly and evolution. This study will introduce and analyse the behaviour of a model for a paradigmatic example of self-assembly, the icosahedral virus capsid, a symmetric structure formed from interacting protein subunits. Results for the thermodynamics and kinetics of model virus assembly will be presented, and the model will be extended to include complicating effects such as different subunit types and cellular crowding. A more general formalism for analysing self-assembling systems will then be introduced and used to suggest a well-defined complexity measure of universal applicability to self-assembled structures. The biological evolution of simple self-assembling structures will be studied using genetic algorithms. The suitability of this modelling approach and its dependence on the many parameters involved will be investigated. Several active areas of enquiry in the field of evolution, including the evolution of complexity, the presence and effect of neutral networks, fluctuations in evolutionary time series and the emergence of symmetry will be investigated within this framework. Throughout this study, we will use the picture of "exploration and exploitation": different approaches to performing optimisation on an unknown landscape, essentially corresponding to a random search and a hill-climbing approach respectively . We will show that, both in self-assembly and evolution, finding an optimal combination of these two approaches gives rise to many of the observed features in this study.

The renewal of derelict inner-city urban districts suffering from high levels of socio-economic deprivation and sustainability problems is one of the key research areas in urban planning and regeneration. Subject to a wide range of social, economical and environmental factors, decision support for an optimal allocation of residential and service lots within such districts is regarded as a complex task. Pre-assessment of various neighbourhood factors before the commencement of actual location allocation of various public services is considered paramount to the sutainable outcome of regeneration projects. Spatial assessment in such derelict built-up areas requires planning of lot assignment for residential buildings in a way to maximize accessibility to public services while minimizing the deprivation of built neighbourhood areas. However, the prediction of socio-economic deprivation impact on the regeneration districts in order to optimize the location-allocation of public service infrastructure is a complex task. This is generally due to the highly conflicting nature of various service structures with various socio-economic and environmental factors. In regards to the problem given above, this thesis presents the development of an evolutionary AI-based decision support systemto assist planners with the assessment and optimization of regeneration districts. The work develops an Adaptive Network Based Fuzzy Inference System (ANFIS) based module to assess neighbourhood districts for various deprivation factors. Additionally an evolutionary genetic algorithms based solution is implemented to optimize various urban regeneration layouts based upon the prior deprivation assessment model. The two-tiered framework initially assesses socio-cultural deprivation levels of employment, health, crime and transport accessibility in neighbourhood areas and produces a deprivation impact matrix overthe regeneration layout lots based upon a trained, network-based fuzzy inference system. Based upon this impact matrix a genetic algorithm is developed to optimize the placement of various public services (shopping malls, primary schools, GPs and post offices) in a way that maximize the accessibility of all services to regenerated residential units as well as contribute to minimize the measure of deprivation of surrounding neighbourhood areas. The outcome of this research is evaluated over two real-world case studies presenting highly coherent results. The work ultimately produces a smart urban regeneration toolkit which provides designer and planner decision support in the form of a simulation toolkit.

Although traditionally Wireless Sensor Network (WSNs) have been regarded as static sensor arrays used mainly for environmental monitoring, recently, its applications have undergone a paradigm shift from static to more dynamic environments, where nodes are attached to moving objects, people or animals. Applications that use WSNs in motion are broad, ranging from transport and logistics to animal monitoring, health care and military. These application domains have a number of characteristics that challenge the algorithmic design of WSNs. Firstly, mobility has a negative effect on the quality of the wireless communication and the performance of networking protocols. Nevertheless, it has been shown that mobility can enhance the functionality of the network by exploiting the movement patterns of mobile objects. Secondly, the heterogeneity of devices in a WSN has to be taken into account for increasing the network performance and lifetime. Thirdly, the WSN services should ideally assist the user in an unobtrusive and transparent way. Fourthly, energy-efficiency and scalability are of primary importance to prevent the network performance degradation. This thesis contributes toward the design of a new hybrid optimization algorithm; ENAMS (Energy optimizatioN Algorithm for Mobile Sensor networks) which is based on the Evolutionary Computation and Swarm Intelligence to increase the life time of mobile wireless sensor networks. The presented algorithm is suitable for large scale mobile sensor networks and provides a robust and energy- efficient communication mechanism by dividing the sensor-nodes into clusters, where the number of clusters is not predefined and the sensors within each cluster are not necessary to be distributed in the same density. The presented algorithm enables the sensor nodes to move as swarms within the search space while keeping optimum distances between the sensors. To verify the objectives of the proposed algorithm, the LEGO-NXT MIND-STORMS robots are used to act as particles in a moving swarm keeping the optimum distances while tracking each other within the permitted distance range in the search space.

New technological advances and the requirements to increasingly abide by new safety laws in engineering design projects highly affects industrial products in areas such as automotive, aerospace and railway industries. The necessity arises to design reduced-cost hi-tech products with minimal complexity, optimal performance, effective parameter robustness properties, and high reliability with fault tolerance. In this context the control system design plays an important role and the impact is crucial relative to the level of cost efficiency of a product. Measurement of required information for the operation of the design control system in any product is a vital issue, and in such cases a number of sensors can be available to select from in order to achieve the desired system properties. However, for a complex engineering system a manual procedure to select the best sensor set subject to the desired system properties can be very complicated, time consuming or even impossible to achieve. This is more evident in the case of large number of sensors and the requirement to comply with optimum performance. The thesis describes a comprehensive study of sensor selection for control and fault tolerance with the particular application of an ElectroMagnetic Levitation system (being an unstable, nonlinear, safety-critical system with non-trivial control performance requirements). The particular aim of the presented work is to identify effective sensor selection frameworks subject to given system properties for controlling (with a level of fault tolerance) the MagLev suspension system. A particular objective of the work is to identify the minimum possible sensors that can be used to cover multiple sensor faults, while maintaining optimum performance with the remaining sensors. The tools employed combine modern control strategies and multiobjective constraint optimisation (for tuning purposes) methods. An important part of the work is the design and construction of a 25kg MagLev suspension to be used for experimental verification of the proposed sensor selection frameworks.

Wireless Sensor Networks(WSNs) have gained a lot of attention because of their potential to immerse deeper into people' lives. The applications of WSNs range from small home environment networks to large habitat monitoring. These highly diverse scenarios impose different requirements on WSNs and lead to distinct design and implementation decisions. This thesis presents an optimization framework for WSN design which selects a proper set of protocols and number of nodes before a practical network deployment. A Genetic Algorithm(GA)-based Sensor Network Design Tool(SNDT) is proposed in this work for wireless sensor network design in terms of performance, considering application-specific requirements, deployment constrains and energy characteristics. SNDT relies on offine simulation analysis to help resolve design decisions. A GA is used as the optimization tool of the proposed system and an appropriate fitness function is derived to incorporate many aspects of network performance. The configuration attributes optimized by SNDT comprise the communication protocol selection and the number of nodes deployed in a fixed area. Three specific cases : a periodic-measuring application, an event detection type of application and a tracking-based application are considered to demonstrate and assess how the proposed framework performs. Considering the initial requirements of each case, the solutions provided by SNDT were proven to be favourable in terms of energy consumption, end-to-end delay and loss. The user-defined application requirements were successfully achieved.

Safety systems are designed to prevent the occurrence of certain conditions and their future development into a hazardous situation. The consequence of the failure of a safety system of a potentially hazardous industrial system or process varies from minor inconvenience and cost to personal injury, significant economic loss and death. To minimise the likelihood of a hazardous situation, safety systems must be designed to maximise their availability. Therefore, the purpose of this thesis is to propose an effective safety system design optimization scheme. A multi-objective genetic algorithm has been adopted, where the criteria catered for includes unavailability, cost, spurious trip and maintenance down time. Analyses of individual system designs are carried out using the latest advantages of the fault tree analysis technique and the binary decision diagram approach (BDD). The improved strength Pareto evolutionary approach (SPEA2) is chosen to perform the system optimization resulting in the final design specifications. The practicality of the developed approach is demonstrated initially through application to a High Integrity Protection System (HIPS) and subsequently to test scalability using the more complex Firewater Deluge System (FDS). Computer code has been developed to carry out the analysis. The results for both systems are compared to those using a single objective optimization approach (GASSOP) and exhaustive search. The overall conclusions show a number of benefits of the SPEA2 based technique application to the safety system design optimization. It is common for safety systems to feature dependency relationships between its components. To enable the use of the fault tree analysis technique and the BDD approach for such systems, the Markov method is incorporated into the optimization process. The main types of dependency which can exist between the safety system component failures are identified. The Markov model generation algorithms are suggested for each type of dependency. The modified optimization tool is tested on the HIPS and FDS. Results comparison shows the benefit of using the modified technique for safety system optimization. Finally the effectiveness and application to general safety systems is discussed.

Genetic Algorithms (GAs) are a search heuristic modeled on the processes of evolution. They have been used to solve optimisation problems in a wide variety of fields. When applied to the optimisation of intervention schedules for optimal control problems, such as cancer chemotherapy treatment scheduling, GAs have been shown to require more fitness function evaluations than other search heuristics to find fit solutions. This thesis presents extensions to the GA crossover process, termed directed intervention crossover techniques, that greatly reduce the number of fitness function evaluations required to find fit solutions, thus increasing the effectiveness of GAs for problems of this type. The directed intervention crossover techniques use intervention scheduling information from parent solutions to direct the offspring produced in the GA crossover process towards more promising areas of a search space. By counting the number of interventions present in parents and adjusting the number of interventions for offspring schedules around it, this allows for highly fit solutions to be found in less fitness function evaluations. The validity of these novel approaches are illustrated through comparison with conventional GA crossover approaches for optimisation of intervention schedules of bio-control application in mushroom farming and cancer chemotherapy treatment. These involve optimally scheduling the application of a bio-control agent to combat pests in mushroom farming and optimising the timing and dosage strength of cancer chemotherapy treatments to maximise their effectiveness. This work demonstrates that significant advantages are gained in terms of both fitness function evaluations required and fitness scores found using the proposed approaches when compared with traditional GA crossover approaches for the production of optimal control schedules.

A novel real-time control method that minimises linear system vibrations when it is subjected to an arbitrary external excitation is proposed in this study. The work deals with a discrete differential dynamic programming type of problem, in which an external disturbance is controlled over a time horizon by a control force strategy constituted by the well-known convolution approach. The proposed method states that if a control strategy can be established to restore an impulse external disturbance, then the convolution concept can be used to generate an overall control strategy to control the system response when it is subjected to an arbitrary external disturbance. The arbitrary disturbance is divided into impulses and by simply scaling, shifting and summation of the obtained control strategy against the impulse input for each impulse of the arbitrary disturbance, the overall control strategy will be established. Genetic Algorithm was adopted to obtain an optimal control force plan to suppress the system vibrations when it is subjected to a shock disturbance, and then the Convolution concept was used to enable the system response to be controlled in real-time using the obtained control strategy. Numerical tests were carried out on a two-degree of freedom quarter-vehicle active suspension model and the results were compared with results generated using the Linear Quadratic Regulator (LQR) method. The method was also applied to control the vibration of a seven-degree of freedom full-vehicle active suspension model. In addition, the effect of a time delay on the performance of the proposed approach was also studied. To demonstrate the applicability of the proposed method in real-time control, experimental tests were performed on a quarter-vehicle test rig equipped with a pneumatic active suspension. Numerical and experimental results showed the effectiveness of the proposed method in reducing the vehicle vibrations. One of the main contributions of this work besides using the Convolution concept to provide a real time control strategy is the reduction in the number of sensors needed to construct the proposed method as the disturbance amplitude is the only parameter needed to be measured (known). Finally, having achieved what has been proposed above, a generic robust control method is accomplished, which not only can be applied for active suspension systems but also in many other fields.

The application of optimization to Water Distribution Systems encompasses the use of computer-based techniques to problems of many different areas of system design, maintenance and operational management. As well as laying out the configuration of new WDS networks, optimization is commonly needed to assist in the rehabilitation or reinforcement of existing network infrastructure in which alternative scenarios driven by investment constraints and hydraulic performance are used to demonstrate a cost-benefit relationship between different network intervention strategies. Moreover, the ongoing operation of a WDS is also subject to optimization, particularly with respect to the minimization of energy costs associated with pumping and storage and the calibration of hydraulic network models to match observed field data. Increasingly, Evolutionary Optimization techniques, of which Genetic Algorithms are the best-known examples, are applied to aid practitioners in these facets of design, management and operation of water distribution networks as part of Decision Support Systems (DSS). Evolutionary Optimization employs processes akin to those of natural selection and “survival of the fittest” to manipulate a population of individual solutions, which, over time, “evolve” towards optimal solutions. Such algorithms are characterized, however, by large numbers of function evaluations. This, coupled with the computational complexity associated with the hydraulic simulation of water networks incurs significant computational overheads, can limit the applicability and scalability of this technology in this domain. Accordingly, this thesis presents a methodology for applying Genetic Algorithms to Water Distribution Systems. A number of new procedures are presented for improving the performance of such algorithms when applied to complex engineering problems. These techniques approach the problem of minimising the impact of the inherent computational complexity of these problems from a number of angles. A novel genetic representation is presented which combines the algorithmic simplicity of the classical binary string of the Genetic Algorithm with the performance advantages inherent in an integer-based representation. Further algorithmic improvements are demonstrated with an intelligent mutation operator that “learns” which genes have the greatest impact on the quality of a solution and concentrates the mutation operations on those genes. A technique for implementing caching of solutions – recalling the results for solutions that have already been calculated - is demonstrated to reduce runtimes for Genetic Algorithms where applied to problems with significant computation complexity in their evaluation functions. A novel reformulation of the Genetic Algorithm for implementing robust stochastic optimizations is presented which employs the caching technology developed to produce an multiple-objective optimization methodology that demonstrates dramatically improved quality of solutions for given runtime of the algorithm. These extensions to the Genetic Algorithm techniques are coupled with a supporting software library that represents a standardized modelling architecture for the representation of connected networks. This library gives rise to a system for distributing the computational load of hydraulic simulations across a network of computers. This methodology is established to provide a viable, scalable technique for accelerating evolutionary optimization applications.

The one- and two-sided bipartite graph drawing problem alms to find a layout of a bipartite graph, with vertices of the two parts placed on parallel imaginary lines, that has the minimum number of edge-crossings. Vertices of one part are in fixed positions for the one-sided problem, whereas all vertices are free to move along their lines in the two-sided version. Many different heuristics exist for finding approximations to these problems, which are NP-hard. New sequential and parallel methods for producing drawings with low edgecrossings are investigated and compared to existing algorithms, notably Penalty Minimisation and Sifting, the current leaders. For the one-sided problem, new methods that include those based on simple stochastic hillclimbing, simulated annealing and genet.ic algorithms were tested. The new block-crossover genetic algorithm produced very good results with lower crossings than existing methods, although it tended to be slower. However, time was a secondary aim, the priority being to achieve low numbers of crossings. This algorithm can also be seeded with the output of an existing algorithm to improve results; combining with Penalty Minimisation in this way improved both the speed and number of crossings. Four parallel methods for the one-sided problem have been created, although two were abandoned because they gave bad results for even simple graphs. The other two methods, based on stochastic hill-climbing, produced acceptable results in faster times than similar sequential methods. PVM was used as the parallel communication system. Two new heuristics were studied for the two-sided problem, for which the only known existing method is to apply one-sided algorithms iteratively. The first is based on a heuristic for the linear arrangment problem; the second is a method of performing stochastic hill-climbing on two sides. A way of applying anyone-sided algorithm iteratively was also created. The linear arrangement method based on the Koren-Harel multi-scale algorithm achieved the best results, outperforming iterative Barycentre (previously the best method) and iterative Penalty Minimisation. Another area of this work created three new heuristics for the k-planar drawing problem where k > 1. These are the first known practical algorithms to solve this problem. A sequential genetic algorithm based on TimGA is devised to work on k-planar graphs. Two parallel algorithms, one island model and the other a 'mesh' model, are also given. Comparison of results for k = 2 indicate that the parallel island method is better than the other two methods. MPI was used for the parallel communication. Overall, 14 new methods are introduced, of which 10 were developed into working algorithms. For the one-sided bipartite graph drawing problem the new block-crossover genetic algorithm can produce drawings with lower crossings than the current best available algorithms. The parallel methods do not perform as well as the sequential ones, although they generally achieved the same results faster. All of the new two-sided methods worked well; the weighted two-sided swap stochastic hill-climbing method was comparable to the existing best method, iterative Barycentre, and generally produced drawings with lower crossings, although it suffered with needing a good termination condition. The new methods based on the linear arrangement problem consistently produced drawings with lower crossings than iterative Barycentre, although they were nearly always slower. A new parallel algorithm for the k-planar drawing problem, based on the island model, generally created drawings with the lowest edge-crossings, although no algorithms were known to exist to make comparisons.

In this thesis the control of ultrafast (10⁻¹³ s) optical pulses used for metrological applications has been investigated. Two different measurement set-ups have been considered, both based around the `pump-probe' technique, where an optical pulse is divided into two parts, one to `pump' or excite a physical system of interest, the other to `probe' or measure the outcome. In both cases the measurement uses electro-optic sampling (EOS), where an electric field is measured by detecting changes in the optical probe pulse polarisation after interaction with the field. In the first study, a method for wavelength metrology in the terahertz (THz) region has been demonstrated by producing an optical pulse shaper and genetic algorithm to control pump pulses and so indirectly influence the THz spectra they generate. In the second study an OPO (optical parametric oscillator) has been developed to provide ultrafast optical pulses for the generation of < 100 fs electrical pulses for metrology using quantum interference control (QUIC). QUIC electrical signals have been demonstrated successfully by charge accumulation measurements and the QUIC electrical pulse temporally measured using EOS, though the low signal levels due to power restrictions mean the QUIC electrical pulse is unsuitable for metrology at this time. Finally, a portable optical pulse measurement device based around frequency-resolved optical gating (FROG) has been designed, built and tested. This has been shown to be capable of retrieving amplitude and phase information in both the temporal and spectral domains for optical pulses as short as 20 fs duration. The ability to characterise shaped pulses also has been demonstrated successfully, with the requirements for full automation identified.

The objectives of this research work are to investigate, design and implement active integrated antennas comprising active devices connected directly to the patch radiators, for various applications in high efficiency RF front-ends, integrated oscillator antennas, design and optimization of harmonic suppression antennas using a genetic algorithm (GA). A computer-aided design approach to obtain a class F operation to optimizing the optimal fundamental load impedance and designing the input matching circuits for an active integrated antenna of the transmitting type is proposed and a case study of a design for 1.6 GHz is used to confirm the design principle. A study of active integrated oscillator antennas with a series feed back using a pseudomorphic high electronmobility transistor (PHEMT) confirms the design procedure in simulation and measurement for the oscillator circuit connected directly to the active antenna. Subsequently, another design of active oscillator antenna using bipolar junction transistor (BJT) improves the phase noise of the oscillation and in addition to achieve amplitude shift keying (ASK) and amplitude modulation (AM) modulation using the proposed design circuit. Moreover, the possibility of using a sensor patch technique to find the power accepted by the antenna at harmonic frequencies is studied. A novel numerical solution, for designing and optimizing active patch antennas for harmonic suppression using GA in collaboration with numerical electromagnetic computation (NEC), is presented. A new FORTRAN program is developed and used for adaptively meshing any planar antenna structure in terms of wire grid surface structures. The program is subsequently implemented in harmonic suppression antenna design and optimization using GA. The simulation and measurement results for several surface structures show a good agreement.

This thesis introduces Intelligent Fitness Functions and Partial Fitness Functions both of which can improve the performance of a genetic algorithm which is limited to a fixed run time. An Intelligent Fitness Function is defined as a fitness function with a memory. The memory is used to store information about individuals so that duplicate individuals do not need to have their fitness tested. Different types of memory (long and short term) and different storage strategies (fitness based, time base and frequency based) have been tested. The results show that an intelligent fitness function, with a time based long term memory improves the efficiency of a genetic algorithm the most. A Partial Fitness Function is defined as a fitness function that only partially tests the fitness of an individual at each generation. Thus only promising individuals get fully tested. Using a partial fitness function gives the genetic algorithm more evolutionary steps in the same length of time as a genetic algorithm using a normal fitness function. The results show that a genetic algorithm using a partial fitness function can achieve higher fitness levels than a genetic algorithm using a normal fitness function. Finally a genetic algorithm designed to solve a substitution cipher is compared to one equipped with an intelligent fitness function and another equipped with a partial fitness function. The genetic algorithm with the intelligent fitness function and the genetic algorithm with the partial fitness function both show a significant improvement over the genetic algorithm with a conventional fitness function.

This thesis investigates the functionality of the units used in connectionist Artificial Intelligence systems. Artificial Neural Networks form the foundation of the research and their units, Artificial Neurons, are first compared with alternative models. This initial work is mainly in the spatial-domain and introduces a new neural model, termed a Taylor Series neuron. This is designed to be flexible enough to assume most mathematical functions. The unit is based on Power Series theory and a specifically implemented Taylor Series neuron is demonstrated. These neurons are of particular usefulness in evolutionary networks as they allow the complexity to increase without adding units. Training is achieved via various traditiona and derived methods based on the Delta Rule, Backpropagation, Genetic Algorithms and associated evolutionary techniques. This new neural unit has been presented as a controllable and more highly functional alternative to previous models. The work on the Taylor Series neuron moved into time-domain behaviour and through the investigation of neural oscillators led to an examination of single-celled intelligence from which the later work developed. Connectionist approaches to Artificial Intelligence are almost always based on Artificial Neural Networks. However, another route towards Parallel Distributed Processing was introduced. This was inspired by the intelligence displayed by single-celled creatures called Protoctists (Protists). A new system based on networks of interacting proteins was introduced. These networks were tested in pattern-recognition and control tasks in the time-domain and proved more flexible than most neuron models. They were trained using a Genetic Algorithm and a derived Backpropagation Algorithm. Termed "Artificial BioChemical Networks" (ABN) they have been presented as an alternative approach to connectionist systems.

A computational study has been carried out to investigate smoldering ignition and propagation in polyurethane foam. The onedimensional, transient, governing equations for smoldering combustion in a porous fuel are solved accounting for improved solid-phase chemical kinetics. A systematic methodology for the determination of solid-phase kinetics suitable for numerical models has been developed and applied to the simulation of smoldering combustion. This methodology consists in the correlation of a mathematical representation of a reaction mechanism with data from previous thermogravimetric experiments. Geneticalgorithm and trail-and-error techniques are used as the optimization procedures. The corresponding kinetic parameters for two different mechanisms of polyurethane foam smoldering kinetics are quantified: a previously proposed 3-step mechanism and a new 5-step mechanism. These kinetic mechanisms are used to model one-dimensionalsmoldering combustion, numerically solving for the solid-phase and gasphase conservation equations in microgravity with a forced flow of oxidizer gas. The results from previously conducted microgravity experiments with flexible polyurethane foam are used for calibration and testing of the model predictive capabilities. Both forward and opposed smoldering configurations are examined. The model describes well both opposed and forward propagation. Specifically, the model predicts the reaction-front thermal and species structure, the onset of smoldering ignition, and the propagation rate. The model results reproduce the most important features of the smolder process and represent a significant step forward in smoldering combustion modeling.

This thesis describes a novel approach to the evolution of Modular Artificial Neural Networks. Standard Evolutionary Algorithms, used in this application include: Genetic Algorithms, Evolutionary Strategies, Evolutionary Programming and Genetic Programming; however, these often fail in the evolution of complex systems, particularly when such systems involve multi-domain sensory information which interacts in complex ways with system outputs. The aim in this work is to produce an evolutionary method that allows the structure of the network to evolve from simple to complex as it interacts with a dynamic environment. This new algorithm is therefore based on Incremental Evolution. A simulated model of a legged robot was used as a test-bed for the approach. The algorithm starts with a simple robotic body plan. This then grows incrementally in complexity along with its controlling neural network and the environment it reacts with. The network grows by adding modules to its structure - so the technique may also be termed a Growth Algorithm. Experiments are presented showing the successful evolution of multi-legged gaits and a simple vision system. These are then integrated together to form a complete robotic system. The possibility of the evolution of complex systems is one advantage of the algorithm and it is argued that it represents a possible path towards more advanced artificial intelligence. Applications in Electronics, Computer Science, Mechanical Engineering and Aerospace are also discussed.

A novel delamination identification technique based on a low-population genetic algorithm for the quantitative characterisation of a single delamination in composite laminated panels is developed, and validated experimentally The damage identification method is formulated as an inverse problem through which system parameters are identified. The input of the inverse problem, the central geometric moments (CGM), is calculated from the surface out-of-plane displacements measurements of a delaminated panel obtained from Digital Speckle Pattern Interferometry (DSPI). The output parameters, the planar location, size and depth of the flaw, are the solution to the inverse problem to characterise an idealised elliptical flaw. The inverse problem is then reduced to an optimisation problem where the objective function is defined as the L2 norm of the difference between the CGM obtained from a finite element (FE) model with a trial delamination and the moments computed from the DSPI measurements. The optimum crack parameters are found by minimising the objective function through the use of a low-population real-coded genetic algorithm (LARGA). DSPI measurements of ten delaminated T700/LTM-45EL carbon/epoxy laminate panels with embedded delaminations are used to validate the methodology presented in this thesis.

This thesis presents research into the possibility of developing a computerised system that can evaluate the aesthetics and engineering aspects of solid shapes. One of the research areas is also to include such an evaluation system into an existing evolutionary CAD system which utilizes the Genetic Algorithms (GAs) technology. An extensive literature survey has been carried out to better understand and clarify the vagueness and subjectivity of the concept of aesthetics, which leads to the work of defining and quantifying a set of aesthetic parameters. This research achieves its novelty in aiming to assist designers in evaluating the aesthetics and functional aspects of designs early in the conceptual design stage, and its inclusion into an evolutionary CAD system. The field of Computer Aided Design (CAD) lacks the aesthetics aspect of the design, which is very crucial in evaluating designs especially considering the trend towards virtual prototypes replacing physical prototypes. This research has managed to suggest, define and quantify a set of aesthetic and functional elements or parameters, which will be the basis of solid shape evaluation. This achievement will help designers in determining the fulfilment of design targets, where the designers will have a full control to determine the priority of each evaluation element in the developed system. In achieving this, computer software including a programming language package and CAD software are involved, which eventually led to the development of a prototype system called Computer Aided Aesthetics and Functions Evaluation (CAAFE). An evolutionary CAD system called Evolutionary Form Design (EFD), which utilizes GAs, has been available for few years now. It evolves shapes for quick and creative suggestions, however it lacks the automated evaluation and aesthetics aspects of the design. This research has worked into the integrating of CAAFE into EFD, which led to a system that could evolve objects based on a selected and weighed aesthetic and functional elements. Finally, surveys from users have also been presented in this thesis to offer improvement to the scoring system within the CAAFE system.

The planning of distribution systems consists in determining the optimum site and size of new substations and feeders in order to satisfy the future power demand with minimum investment and operational costs and an acceptable level of reliability. This problem is a combinatorial, non-linear and constrained optimization problem. Several solution methods based on genetic algorithms have been reported in the literature; however, some of these methods have been reported with applications to small systems while others have long solution time. In addition, the vast majority of the developed methods handle planning problems simplifying them as single-objective problems but, there are some planning aspects that can not be combined into a single scalar objective; therefore, they require to be treated separately. The cause of these shortcomings is the poor representation of the potential solutions and their genetic operators This thesis presents the design of a genetic algorithm using a direct representation technique and specialized genetic operators for power distribution system expansion planning problems. These operators effectively preserve and exploit critical configurations that contribute to the optimization of the objective function. The constraints of the problems are efficiently handle with new strategies. The genetic algorithm was tested on several theoretical and real large-scale power distribution systems. Problems of network reconfiguration for loss reduction were also included in order to show the potential of the algorithm to resolve operational problems. Both single-objective and multi-objective formulations were considered in the tests. The results were compared with results from other heuristic methods such as ant colony system algorithms, evolutionary programming, differential evolution and other genetic algorithms reported in the literature. From these comparisons it was concluded that the proposed genetic algorithm is suitable to resolve problems of largescale power distribution system planning. Moreover, the algorithm proved to be effective, efficient and robust with better performance than other previous methods.