Search

Automated container terminals (ACTs) represent state-of-the-art facilities for container handling and are a current development trend. However, enhancing their operational efficiency while minimizing energy consumption remains a challenge. While metaheuristics are helpful in addressing container terminal problems, their capabilities are limited when used alone to tackle complex or integrated problems. Hybrid models can be more effective in facing these challenges. This research proposes a hybrid model, termed WOA + PSO, which combines the Whale Optimization Algorithm (WOA) and Particle Swarm Optimization (PSO) as a novel approach to address the integrated scheduling problem of automated quay cranes (AQCs), automated lift vehicles (ALVs), and automated stacking cranes (ASCs) in an ACT. Additionally, the WOA + PSO collaborates with a simulation model in a framework to become a simulation-based optimization approach. The performance of WOA + PSO is evaluated by comparing it with its base models, WOA and PSO, as well as a genetic algorithm (GA), through extensive experiments. The results show that WOA + PSO outperforms the others in achieving the objective of balancing operational efficiency and energy consumption. [ABSTRACT FROM AUTHOR]

Traditional heuristic optimization algorithms are no longer applicable to the multiple satellites scheduling with large-scale tasks, as they are unable to provide satisfactory performance in terms of convergence and speed when addressing complex constraint conflicts. Therefore, we propose a multi-satellite cooperative scheduling method for large-scale tasks based on a hybrid graph neural network (GNN) and metaheuristic algorithm. By using the representation and extraction capability of GNN for relations in graph, the features of large-scale tasks and their constraint relations were expressed, and the generalized knowledge was extracted. And combined with metaheuristic algorithm, the optimization framework for large-scale satellite task scheduling based knowledge and data was implemented. The method consists of a GNN pre-solver module for static constraint conflicts and a metaheuristic optimization module for dynamic constraint conflicts of satellite missions. In the GNN pre-solver module, the graph sample and aggregate network was used to learn and extract the common knowledge in the static task-conflict graph, and provided effective prior information for the metaheuristic optimization module. The quadratic unconstrained binary optimization loss function was designed with multiple influencing factors, to convert the discrete optimization function into a continuous function; a greedy threshold was also added to improve the task completion rate. Finally, numerical experimental results showed that the proposed method can achieve an efficient solution to the multi-satellite scheduling problem with tens of thousands tasks. Compared with the commonly used multi-satellite scheduling algorithms (the genetic algorithm (GA) and greedy-GA), the proposed method can obtain higher-quality solutions under the same conditions and greatly improve the computational efficiency of large-scale mission planning. [ABSTRACT FROM AUTHOR]

Eyeglasses are critical for correcting and protecting the vision of children with myopia/hyperopia. Children's heads grow as their age increases, which can easily cause improperly fitting and uncomfortable eyeglasses. Hence, better sizing for children's eyeglasses is needed, but to date this topic is underexplored. To solve this, we proposed a comfort-oriented and assembly-guided approach based on a statistical ophthalmic model to establish an eyeglasses sizing system. First, we developed an assembly-guided design scheme in which eyeglasses could be separated into three components – rims, temples, and nose pads – and assembled in different combinations. Then, we modeled the geometric variances of children's heads' ophthalmic regions. After that, we used descriptive statistics to size nose pads and temples and employed a statistical model with a genetic algorithm to search for representative shapes and customize rim sizes. Here, the coverage rate and overall comfort were computed as objective functions. Experimental results showed that both an assembly-guided design scheme and comfort-oriented searching strategy can effectively produce more comfortable eyeglasses sizing systems. Our method can also be extended to other head and body wearables. • This study pioneers the establishment of an eyeglasses sizing system for children. • We developed a novel comfort-oriented framework to size eyeglasses. • We presented a novel assembled-eyeglasses design scheme to size eyeglasses. • We created a 3D statistical ophthalmic model for children's eyeglasses design. [ABSTRACT FROM AUTHOR]

Modular design is a common idea in modern system design. In a modular system, since most components are modular, components with the same function but different qualities can swap positions with each other. Therefore, from the perspective of reliability engineering, component allocation of modular systems (that is, placing high-quality components in key positions and lower-quality components in other positions) has become an efficient quality control method. Component allocation problem, also termed CAP, has become a specialized class of problems and is widely used in multiple industries such as hydraulics and communications. However, it is noticed that when CAPs are applied to some complex systems in reality, some new situations such as shared positions will arise, making the modeling and solving of CAP more complicated. For example, in the installation of communication base stations, operators generally only allow base station facilities within their own sphere of influence to install their own base stations. Also, for economic reasons, communication operators in nearby areas are willing to share the positions of boundaries between different regions. The introduction of shared positions has become a difficult problem in the current field of CAPs. To the best of our knowledge, existing works do not address the multi-type component allocation problem with shared positions. Therefore, this paper establishes a new component allocation model for multi-type component allocation problem with shared positions, namely SP-MCAP, based on the scenario of base station installation in communication systems. Then, an integrated genetic algorithm, termed the BI-based two-stage genetic algorithm (BITSGA), is designed to address it. To verify the efficiency of BITSGA, some numerical experiments are performed, respectively. Finally, a case on a wireless transmission system is given to show its application in real-world problems. [ABSTRACT FROM AUTHOR]

Highlights • The extra loss caused by uncertainty is measured. • A MIP model under uncertainty is formulated considering a lot of uncertain scenarios. • A GA-based framework combined with three-stage algorithm is proposed. Abstract In a container terminal, the arriving times and handling volumes of the vessels are uncertain. The arriving times of the external trucks and the number of containers which are needed to be brought into or retrieved from a container terminal by external trucks within a period are also uncertain. Yard crane (YC) scheduling is under uncertainty. This paper addresses a YC scheduling problem with uncertainty of the task groups' arriving times and handling volumes. We do not only optimize the efficiency of YC operations, but also optimize the extra loss caused by uncertainty for reducing risk of adjusting schedule as the result of the task groups' arriving times and handling volumes deviating from their plan. A mathematical model is proposed for optimizing the total delay to the estimated ending time of all task groups without uncertainty and the extra loss under all uncertain scenarios. Furthermore, a GA-based framework combined with three-stage algorithm is proposed to solve the problem. Finally, the proposed mathematical model and approach are validated by numerical experiments. [ABSTRACT FROM AUTHOR]

Abstract The healthcare industry is confronted with the challenge of offering customized services while in the meantime to control increasing healthcare costs. Modularization is an important approach to reduce healthcare costs and improve patient-centered services via decreasing process complexity and enhancing flexibility through configuring pre-identified service modules. Recognizing the importance of modularity for healthcare services, this paper introduces Design Structure Matrix (DSM) as a technique for healthcare process modularization. A DSM-based modularization and sequencing algorithm is developed to allocate healthcare activities to service modules using Genetic Algorithm (GA) and arrange sequences of services both within and across service modules to support modular clinical pathway design. The proposed algorithm is implemented with a case study, the results of which have demonstrated the feasibility and applicability of the proposed DSM-based modularization method for healthcare process design. [ABSTRACT FROM AUTHOR]

Highlights: • A novel method was proposed for allocating incoming components in limited site space. • This study developed a model for the storage area planning and applied it. • A method was firstly presented to identify unoccupied locations in real time. • Intelligently acquired variables were considered in GA. Abstract Rehandling of construction components, such as pipes, structural steel elements, and curtain walls, may increase the handling cost and reduce the construction efficiency, which is a critical issue for storage area plans of a project. Moreover, on some construction sites where space is limited, there are not adequate storage areas for centralized stacking of components and frequent changes in spatial state. Existing studies have investigated site layout planning for temporary facilities including arranging a storage area for the same type of material, which still have limitations in solving the above problems. This study proposes a novel and flexible arrangement method for incoming components in limited site space. This method is non-centralized and adaptive to the dynamic change of the actual component requirements based on construction activities and the real-time storage area availability. Therefore, a construction components storage areas planning (CCSAP) model is developed for dynamic allocation of construction components storage areas. Building information modeling (BIM) can be used to generate the material requirements planning before construction according to the actual construction activities. Real-time spatial recognition is a critical step for dynamic allocation of construction components storage areas because no such research has been done. This paper firstly presents an imaging technology with a low-rank matrix to identify on-site unoccupied locations automatically in real time. In addition, genetic algorithms (GA) consider two types of decision variables: actual components supply and real-time space availability. Finally, a dynamic visualization platform is built for planning construction components storage areas. An implementation example is demonstrated to validate principles and this model and shows a 21.9% reduction in the handling cost and a 19.4% increase in the construction efficiency compared with conventional methods. [ABSTRACT FROM AUTHOR]

The hybrid flow shop scheduling problem (HFSP) is a research hotspot in the shop scheduling problem. However, few studies have investigated HFSP involving parallel machine scheduling and worker competencies. Therefore, this study investigates a complex product assembly line scheduling problem considering parallel team scheduling, worker competencies, and transport time. The makespan, transport time, imbalance degree of team workload, and imbalance degree of cycle time are taken as optimization objectives. First, a mathematical model considering transport time and worker competencies is developed, and then a mixed encoding and decoding method is proposed. Second, an algorithm that hybridizes the genetic algorithm and simulated annealing algorithm is proposed, involving an improved Nawaz-Enscore-Ham heuristic method and a strengthened elite retention strategy to improve the quality of the solution. Finally, based on nine examples generated from real production statuses, the proposed algorithm is compared with the other four algorithms. The results show that the proposed algorithm is superior in terms of the quality and efficiency of solutions. [ABSTRACT FROM AUTHOR]

In recent decades, various domains such as military, aerospace and supply chain are becoming increasingly complex, forming system-of-systems (SoSs). The effectiveness of SoSs is mainly determined by their architecture, thus selecting proper architecture for SoSs is crucial to maximize their effectiveness. While most studies focus on specific issues, such as mission planning and SoSs meta-architecture selection, few papers deal with the two issues together. This paper introduces the task oriented SoSs architecture selection problem (TSASP) and proposes an combination of reinforcement learning (RL) and evolutionary algorithm (EA) (CRE) to address this problem. CRE comprises two layers: the inner layer utilizes a DQN-based algorithm to solve the mission planning problem, and the outer layer with improved genetic algorithm optimizes the SoSs architecture selection based on the output of inner one. A synthetic air and missile defense (AMD) example is conducted to test the effectiveness of proposed method, and results show that CRE can improve the mission planning effect by select a feasible architecture. [ABSTRACT FROM AUTHOR]

This paper introduces a mixed-integer nonlinear programming (MINLP) model aimed at designing a dependable closed-loop supply chain that can function effectively despite disruptions. The model incorporates several strategies to reduce the risks associated with random facility disruptions. First, the concept of p -robustness criterion is employed, which sets upper bounds on the cost in various disruption scenarios to protect against disruptions. To accommodate for the varying levels of reliability of facilities, the network consists of two categories of infrastructure: reliable and defective hybrid infrastructure. Thirdly, a sharing strategy is proposed to mitigate for the capacity loss at unreliable facilities by permitting transportation of products from reliable facilities to unreliable facilities. Additionally, an improved genetic algorithm (IGA) is devised for the MINLP model efficiently. The efficacy of the model and the tailored algorithm is evaluated by means of computational experiments. Finally, a sensitivity analysis is ultimately performed to furnish managers with meaningful insights. For example, the sharing strategies demonstrate a significant decrease of total cost. This paper contributes to enhancing resilience in logistics networks, ensuring their reliable performance even in the face of disruptions. [ABSTRACT FROM AUTHOR]

• An energy model is set up based on energy consumption analyses of machining processes; • NC codes are structurally evaluated and then popularised into the energy model; • A hybrid GA-ACA algorithm is designed to minimise air-cutting toolpaths to optimise the energy model. Machining processes comprise numerous energy consumption activities. Given the significance of the circular economy and manufacturing sustainability to modern societies, it is paramount to design effective methodologies to accomplish energy-efficient machining processes. With this aim, this research presents a new approach of energy consumption optimisation for machining processes based on numerical control (NC) programs. In the approach, the following innovative characteristics are exhibited: (i) An energy model is systematically established based on a detailed analysis of energy consumption activities in machining processes; (ii) NC programs for specific machining processes are assessed in detail and popularised into the energy model for instantiation; (iii) An optimisation algorithm hybridising the genetic algorithm and the ant colony algorithm is designed to minimise air-cutting toolpaths to optimise the energy model. Two case studies were conducted to validate the presented approach. The case studies revealed that the accuracy of the energy model was 95.3% of the actual energy consumption. The studies also showed that, based on the optimised energy model, the total length of air-cutting toolpaths was reduced by 43.8%, and the total machining time was diminished by 25.8%. It can be concluded that the developed approach can achieve substantial energy savings, and therefore it is highly promising to support machining industries to meet their sustainable targets. [ABSTRACT FROM AUTHOR]

Motivated by the service-oriented intelligent maintenance activity under the development of the digital economy, we study the operation and maintenance strategies of the Maintenance Scheduling and Technician Routing Problem (MSTRP) for medical equipment implemented by multi-skilled technicians. We consider the maintenance services provided for both preventive maintenance (PM, e.g., periodic maintenance based on the contract schedule) and corrective maintenance (CM, e.g., repair action in the event of a sudden breakdown). As these activities are performed on equipment with various types of requirements at different geographically distributed systems, technicians with diverse qualifications need to build teams and visit the matching maintenance requests. Meanwhile, we consider the maintenance actions concerning the dynamic change in equipment failure rate and establish a flexible maintenance mechanism by balancing the total cost. This paper develops an efficient hybrid heuristic based on a genetic algorithm, variable neighborhood search, and other optimization strategies to solve this multi-period equipment maintenance scheduling and multi-skilled technician routing and scheduling problem. Based on the settings of the studied MSTRP, this paper designs different sizes of testing instances and multi-period explanatory cases before conducting extensive comparative experiments and sensitivity analysis to demonstrate the effectiveness of our proposed dynamic maintenance strategy and optimization method. [ABSTRACT FROM AUTHOR]

Prefabrication is a sustainable construction method that improves productivity and reduces environmental footprints. Prefabricated panel is one typical type of components widely used in practice. When designing panel stacking plans for transportation, practitioners should consider numerous practical factors, e.g., panel stability and reshufflings. However, in current practice, stacking plans are made on the basis of planners' intuition and experience. A scientific method that can yield satisfactory panel stacking plans is thus required. Motivated by this real need, this paper aims to address a practical problem for practitioners: how to stack prefabricated panels transported by A-frame trailers with high stability and little reshuffling. This paper extends the literature by innovatively formulating a bi-objective mixed-integer programming model that incorporates all the aforementioned factors and objectives. A non-dominated sorting genetic algorithm is designed to approximate the Pareto optimal solutions. We conduct extensive numerical experiments to validate the effectiveness, efficiency, and applicability of our proposed model and algorithm. The developed model and algorithm are capable of automatically performing calculations and effectively generating vertical stacking plans of high quality for planners in real-world settings within seconds. Future research may be devoted to the optimization of stacking plans for other types of prefabricated components or the horizontal stacking plans for components transported by flat-bed trailers. [ABSTRACT FROM AUTHOR]

Graphical abstract Highlights • The community detection problem is applied in national-scale power grids. • Real power grids (Italy, France, Germany, Iberian peninsula and Texas) are evaluated. • Genetic algorithms are able to obtain any number of communities in a fast way. • Topological information of national-scale power grids is analyzed. • These methods allow to study the vulnerabilities of real power grids. Abstract The large-scale interconnection of electricity networks has been one of the most important investments made by electric companies, and this trend is expected to continue in the future. One of the research topics in this field is the application of graph-based analysis to identify the characteristics of power grids. In particular, the application of community detection techniques allows for the identification of network elements that share valuable properties by partitioning a network into some loosely coupled sub-networks (communities) of similar scale, such that nodes within a community are densely linked, while connections between different communities are sparser. This paper proposes the use of competitive genetic algorithms to rapidly detect any number of community structures in complex grid networks. Results obtained in several national- scale high voltage transmission networks, including Italy, Germany, France, the Iberian peninsula (Spain and Portugal), Texas (US), and the IEEE 118 bus test case that represents a portion of the American Electric Power System (in the Midwestern US), show the good performance of genetic algorithms to detect communities in power grids. In addition to the topological analysis of power grids, the implications of these results from an engineering point of view are discussed, as well as how they could be used to analyze the vulnerability risk of power grids to avoid large-scale cascade failures. [ABSTRACT FROM AUTHOR]

Highlights • Integrates a framework for design exploration using styling strategy and genetic algorithm. • Provides a suitable balance between homogeneity and heterogeneity of design. • Provides basic functions for automatic design alternative generation in the early design stage. Abstract Designers constantly and consistently draft and develop both general concepts and directions to identify the solution that best fits the styling objectives of the lead designer. Designers often confront design fixations that cognitively clash to explore different design combinations. As design teams explore the range of possible design spaces of a certain design strategy, there is an opportunity for computational approaches to improve the styling process. By implementing product appearance similarity and styling strategy in computational design synthesis, it is possible to discover combinations that would otherwise remain unexplored by human designers. Numerous studies on design synthesis have been conducted. However, there has been no focus on the morphological synthesis of designs with strategic styling decisions. Considering this, the proposed study develops a method to synthesize car styling based on product appearance similarity for effective design exploration in the concept generation phase. The similarities of products across different generations, product portfolios, and competitors' products are calculated to evaluate the strategic styling decision. The results of the strategic styling decision are used to formulate a fitness function. Car styling is then synthesized with a genetic algorithm based on this fitness function to generate car styling in accordance with the target strategic styling decision. In this respect, designers can computationally synthesize novel design alternatives that consider both homogeneity (family look in design) and heterogeneity (design trend in the market) by pinpointing the desired design exploration area. Ultimately, the style synthesis methodology proposed in this research can help designers to utilize the gradual visualization of styling strategies for more effective and efficient managerial design decisions. To do this, we conduct five major tasks: first, car design data are collected for design synthesis; second, the product appearance similarity is calculated to measure the strategic styling decision; third, synthesis validation is conducted to test whether the proposed methodology can create outside-the-box designs; fourth, a genetic algorithm is used to synthesize car designs in consideration of the strategic styling decision; finally, a series of in-depth interviews with experts and validation experiments are conducted with in-house automobile designers to examine the impact of the proposed methodology. The results showed that designers can quantitatively measure and compare the styling strategies of each car brand, then implement design upgrades, while still maintaining that specific style. Correspondingly, computationally generated design alternatives improve the satisfaction in ease, time, objective reflection and novelty of design outcomes when formulating design strategies in the concept generation phase. [ABSTRACT FROM AUTHOR]

This research proposes an adaptive clustering-based genetic algorithm (ACGA) to optimize the pick-and-place operation of a dual-gantry component placement machine, which has two independent gantries that alternately place components onto a printed circuit board (PCB). The proposed optimization problem consists of several highly interrelated sub-problems, such as component allocation, nozzle and feeder setups, pick-and-place sequences, etc. In the proposed ACGA, the nozzle and component allocation decisions are made before the evolutionary search of a genetic algorithm to improve the algorithm efficiency. First, the nozzle allocation problem is modeled as a nonlinear integer programming problem and solved by a search-based heuristic that minimizes the total number of the dual-gantry cycles. Then, an adaptive clustering approach is developed to allocate components to each gantry cycle by evaluating the gantry traveling distances over the PCB and the component feeders. Numerical experiments compare the proposed ACGA to another clustering-based genetic algorithm LCO and a heuristic algorithm mPhase in the literature using 30 industrial PCB samples. The experiment results show that the proposed ACGA algorithm reduces the total gantry moving distance by 5.71% and 4.07% on average compared to the LCO and mPhase algorithms, respectively. [ABSTRACT FROM AUTHOR]

Understanding the affective needs of customers is crucial to the success of product design. Hybrid Kansei engineering system (HKES) is an expert system capable of generating products in accordance with the affective responses. HKES consists of two subsystems: forward Kansei engineering system (FKES) and backward Kansei engineering system (BKES). In previous studies, HKES was based primarily on single-objective optimization, such that only one optimal design was obtained in a given simulation run. The use of multi-objective evolutionary algorithm (MOEA) in HKES was only attempted using the non-dominated sorting genetic algorithm-II (NSGA-II), such that very little work has been conducted to compare different MOEAs. In this paper, we propose an approach to HKES combining the methodologies of support vector regression (SVR) and MOEAs. In BKES, we constructed predictive models using SVR. In FKES, optimal design alternatives were generated using MOEAs. Representative designs were obtained using fuzzy c-means algorithm for clustering the Pareto front into groups. To enable comparison, we employed three typical MOEAs: NSGA-II, the Pareto envelope-based selection algorithm-II (PESA-II), and the strength Pareto evolutionary algorithm-2 (SPEA2). A case study of vase form design was provided to demonstrate the proposed approach. Our results suggest that NSGA-II has good convergence performance and hybrid performance; in contrast, SPEA2 provides the strong diversity required by designers. The proposed HKES is applicable to a wide variety of product design problems, while providing creative design ideas through the exploration of numerous Pareto optimal solutions. [ABSTRACT FROM AUTHOR]

Soft sensors have been widely used in industrial processes over the past two decades because they use easy-to-measure process variables to predict difficult-to-measure ones. Some success has been achieved by the dominant traditional methods of modeling soft sensors based on statistics, such as principal components analysis (PCA) and partial least square (PLS), but such sensors usually become inaccurate and inefficient when processing strong nonlinear data. In this paper, a new soft sensor modeling approach is proposed based on a deep learning network. First, stacked auto-encoders (SAEs) are employed to extract high-level feature representations of the input data. In the process of training each layer of a SAE, the Limited-memory Broyden-Fletcher-Goldfarb-Shanno algorithm (L-BFGS) is adopted to optimize the weights parameters. Then, a support vector regression (SVR) is added to predict the target value on the basis of the features obtained from the SAE. To improve the model performance, Genetic Algorithm (GA) is used to obtain the optimal parameters of the SVR. To evaluate the proposed method, a soft sensor model for estimating the rotor deformation of air preheaters in a thermal power plant boiler is studied. The experimental results demonstrate that the soft sensor based on the SAE-SVR algorithm is more effective than the existing methods are. [ABSTRACT FROM AUTHOR]

Evacuation modeling offers challenging research topics to solve problems related to the development of emergency planning strategies. In this paper, we built an agent-based evacuation simulation model to study the pedestrian dynamics and learning process by applying the NeuroEvolution of Augmenting Topologies (NEAT) which is a powerful method to evolve artificial neural networks (ANNs) through genetic algorithms (GAs). The NEAT method strengthens the analogy between GAs and biological evolution by both optimizing and complexifying the solutions simultaneously. We set our main goal to develop a model by identifying the most appropriate fitness function for the agents that can learn how to change and improve their behaviors in a simulation environment such as moving towards the visible targets, producing efficient locomotion, communicating with each other, and avoiding obstacles while reaching targets. The fitness function we chose captured the learning process effectively and our NEAT-based implementation evolved suitable structures for the ANNs autonomously. According to our experiments and observations in the simulated environment, the agents accomplished their tasks successfully and found their ways to the exits. [ABSTRACT FROM AUTHOR]

This paper focuses on the container loading and unloading problem with dynamic ship arrival times. Using a determined berth plan, in combination with the reality of a container terminal production scheduling environment, this paper proposes a scheduling method for quay cranes that can be used for multiple vessels in a container terminal, based on a dynamic rolling-horizon strategy. The goal of this method is to minimize the operation time of all ships at port and obtain operation equilibrium of quay cranes by establishing a mathematical model and using a genetic algorithm to solve the model. Numerical simulations are applied to calculate the optimal loading and unloading order and the completion time of container tasks on a ship. By comparing this result with the traditional method of quay crane loading and unloading, the paper verifies that the quay crane scheduling method for multiple vessels based on a dynamic rolling-horizon strategy can provide a positive contribution to improve the efficiency of container terminal quay crane loading and unloading and reduce resource wastage. [ABSTRACT FROM AUTHOR]

Developing optimal operation policy for single or multi-purposes dams and reservoirs is a complex engineering application. The main reasons for such complexity are the stochastic nature of the system input and slow convergence of the optimization method. Furthermore, searching optimal operation for multi-purposes or chain reservoir systems, becomes even more complex because of interfering operations between successive dams. In this study, a new hybrid algorithm has been introduced by merging the genetic algorithm (GA) with the krill algorithm. In fact, the proposed hybrid algorithm amalgamates the advantages of both algorithms, first, the ability to converge fast for global optimum and, second, considering the effect of stochastic nature of the system. Three benchmark functions were used to evaluate the performance of this proposed optimization model. In addition, the proposed hybrid algorithm was examined for Karun-4 reservoir in Iran as an example for a hydro-power generation dam. Two benchmark problems of hydropower operations for multi-purposes reservoir systems, namely four-reservoir and ten-reservoir systems were considered in the study. Results showed that the proposed hybrid algorithm outperformed the well-developed traditional nonlinear programming solvers, such as Lingo 8 software. [ABSTRACT FROM AUTHOR]

Normal operation of a Mechanical, Electrical, and Plumbing (MEP) system under random and intentional attacks is important to a building. A systematic research framework is proposed to analyze the resilience of an MEP system and optimize its design. The resilience magnitude in this research measures the ability of the MEP system to keep standard operation when component failures appear. First, the MEP model in Building Information Modelling (BIM) environment is extracted to a graph database by using Revit API, which represents the complex network of an MEP system. Second, the importance of the components and the resilience of the MEP system is analyzed based on the network theory and topological metrics. Third, the failure simulation is carried out by attacking the node of the system randomly and intentionally. Finally, the genetic algorithm is used to optimize the design of the MEP system by adding new edges. The results show that: (i) the graph database is a good representation of the MEP system, and it can convert the 3D model to a format that can be analyzed by data analysis measures, (ii) the same component in the MEP system could have different importance from different perspectives, (iii) the proposed network is more resilient with bridge ratio index and average path length improved by 6.16% and 40.58%, respectively, and (iv) the proposed intentional attack strategy is more conform to reality, and it can cause more severe results to a system. The research can contribute to the implementation of the resilience design theory in the MEP discipline, and create a bond between the 3D model and data analysis. [ABSTRACT FROM AUTHOR]

The design of complex systems has been the forte of and therefore limited to engineers while algorithms have only aided in the representation and optimization of the design options. This human-centric approach has several limitations that leads to sub-optimal design results. In this work, a novel algorithmic design framework ALGINEER is proposed with a concept-proof implementation that overcomes the limitations. ALGINEER formalizes the design processes to isolate the design process from the design problem. It implements genetic algorithms and machine learning to explore the complete solution space, achieve trade-off among many design objectives, and demonstrate design behavior and learning akin to engineers. It is modular, scalable and empowers engineers to concentrate more on problem formulation. The work also suggests future research possibilities towards extending ALGINEER's abilities. [ABSTRACT FROM AUTHOR]

In product engineering, design changes are unavoidable because of the evolution of customer requirements and key technologies. Modularization has been extensively applied as an effective strategy for the development of complex products. In a modular product, although modules are supposed to be functionally independent, the design change in a module can still propagate to other modules because of the interfaces between modules. This propagation leads to numerous potential propagation paths and uncontrollable change in costs. Thus, it is valuable to consider the impact of the modular structure during the optimization of design change propagation. This research offers a glimpse into design change propagation routing in modular products to accelerate design change propagation within the modules. First, the relationships between components were determined to construct the product network model. Subsequently, an optimization model of the design change propagation path was established to minimize the change propagation intensity, which includes the importance of the components, change in cost, and path length. To decrease the number of modules that the design change propagates into, the correlation between the modules was explored as penalties, considering the dependencies of a module on the interface(s) that are defined from aspects of the topological characteristics, change in the propagation characteristics, and cost of design changes. A multiple-population genetic algorithm was developed to solve this model. Finally, a case study of the design change propagation routing of a potato lifter was implemented to expound on the utility and effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

Drones are a type of unmanned aerial vehicle. They use several rotors to control their flight motion and stabilize their attitude. This study aims to determine the optimal proportional-integral–differential (PID) gain values that can stabilize a quadcopter with four rotors quickly when its attitude is disturbed. Generally, expert knowledge and a great deal of time and money are required to obtain the PID gain values to stabilize the attitude of a drone. However, in this study, long short-term memory (LSTM), which is a type of neural network algorithm, was used to evaluate the flight motions based on PID gain values without expert knowledge of quadcopters (quadcopter flight motion, PID control algorithms, and quadcopter expert experience). To obtain the optimal values of the PID gain for stabilizing the attitude of a drone, a PID simulator algorithm was developed in this study. The developed algorithm used dynamic equations of motion for the drones. Simulations were used to acquire the drone stabilization data, and a back-propagation neural network was applied to establish an approximate model. Subsequently, the non-dominated sorting genetic algorithm-II was used to obtain the optimal PID gain values that could restore the attitude of the drone quickly. The drone motion data obtained using the simulations were used as the LSTM training data, and the optimal PID gain values obtained using the genetic algorithm were used by the LSTM to predict the motion. To verify the results, a drone was constructed, and the LSTM and dynamic simulation values were compared with the drone experimental values using an experimental device that allowed the motion of the drone to be examined. The results indicated that the motions resulting from the optimal values and the experimental results were in agreement. [ABSTRACT FROM AUTHOR]

In recent years, the “power of the crowd” has been repeatedly demonstrated and various Internet platforms have been used to support applications of collaborative intelligence in tasks ranging from open innovation to image analysis. However, crowdsourcing applications in the fields of design research and creative innovation have been much slower to emerge. So, although there have been reports of systems and researchers using Internet crowdsourcing to carry out generative design, there are still many gaps in knowledge about the capability and limitations of the technology. Indeed the process models developed to support traditional commercial design (e.g. Pugh’s Total Design, Agile, Double-Diamond etc.) have yet to be established for Crowdsourced Design (cDesign). As a contribution to the development of such a general model this paper proposes a cDesign framework to support the creation of crowdsourced design activities. Within the cDesign framework the effective evaluation of design quality is identified as a key component that not only enables the leveraging of a large, virtual workforce’s creative activities but is also fundamental to almost all iterative optimisation processes. This paper reports an experimental investigation into two different Crowdsourced design evaluation approaches; free evaluation and ‘Crowdsourced Design Evaluation Criteria’ (cDEC). The results are benchmarked against a ‘manual’ evaluation carried out by a panel of experienced designers. The results suggest that the cDEC approach produces design rankings that correlate strongly with the judgements of an “expert panel”. The paper concludes that cDEC assessment methodology demonstrates how Crowdsourcing can be effectively used to evaluate, as well as generate, new design solutions. [ABSTRACT FROM AUTHOR]

In the past, designers developed new products by virtue of their own accumulation of aesthetic perception and experience. Because the information that designers could master was limited, it was difficult to develop quickly the capability of satisfying consumer-oriented markets. This limitation may consequently cause the enterprises unnecessary risks. Based on this, a set of aesthetic evaluations and an optimization system for form aesthetics are proposed in this study. Calculations of explicit equations were performed so as to assisting in measuring the aesthetic characteristics; next, a fuzzy judgment was invoked to calculate the perceptual aesthetic measures of a product style so as to establish the overall aesthetic standard for the product. Aesthetic measurement principles were combined with the genetic algorithm (GA) and applied to the optimization of the product’s shape. All-in-one stereos were chosen to serve as the target products as a case study. Further form optimization was conducted on two of the stereos and the questionnaire survey indicated that their aesthetic measures increased after the optimization. The errors that resulted from the equations for aesthetic measurements and judgments were also reduced accordingly. The precision and feasibility of the aesthetic theory that was constructed in this study were assessed, tested, and verified. [ABSTRACT FROM AUTHOR]

Product design is a multidisciplinary activity that requires the integration of concurrent engineering approaches into a design process that secures competitive advantages in product quality. In concurrent engineering, the Taguchi method has demonstrated an efficient design approach for product quality improvement. However, the Taguchi method intuitively uses parameters and levels in measuring the optimum combination of design parameter values, which might not guarantee that the final solution is the most optimal. This work proposes an integrated procedure that involves neural network training and genetic algorithm simulation within the Taguchi quality design process to aid in searching for the optimum solution with more precise design parameter values for improving the product development. The concept of fractals in computer graphics is also considered in the generation of product form alternatives to demonstrate its application in product design. The stages in the general approach of the proposed procedures include: (1) use of the Taguchi experimental design procedure, (2) analysis of the neural network and genetic algorithm process, and (3) generation of design alternatives. An electric fan design is used as an example to describe the development and explore the applicability of the proposed procedures. The results indicate that the proposed procedures could enhance the efficiency of product design efforts by approximately 7.8%. It is also expected that the proposed design procedure will provide designers with a more effective approach to product development. [ABSTRACT FROM AUTHOR]

[Display omitted] • Long-short term memory to capture interrelationship between energy data and time. • GA to select LSTM architecture to improve forecasting accuracy and robustness. • Effects of historical weather and time horizon of past information is analyzed. • Real-life testing of the proposed model is conducted on two educational buildings. • A powerful tool to assist net-zero building operation and mitigate climate change. The real-world building can be regarded as a comprehensive energy engineering system; its actual energy consumption depends on complex affecting factors, including various weather data and time signature. Accurate energy consumption forecasting and effective energy system management play an essential part in improving building energy efficiency. The multi-source weather profile and energy consumption data could enable integrating data-driven models and evolutionary algorithms to achieve higher forecasting accuracy and robustness. The proposed building energy consumption forecasting system consists of three layers: data acquisition and storage layer, data pre-processing layer and data analytics layer. The core part of the data analytics layer is a hybrid genetic algorithm (GA) and long-short term memory (LSTM) neural network model for accurate and robust energy prediction. LSTM neural network is adopted to capture the interrelationship between energy consumption data and time. GA is adopted to select the optimal architecture for LSTM neural networks to improve its forecasting accuracy and robustness. The hyper-parameters for determining LSTM architecture include the number of LSTM layers, number of neurons in each LSTM layer, dropping rate of each LSTM layer and network learning rate. Meanwhile, the effects of historical weather profile and time horizon of past information are also investigated. Two real-life educational buildings are adopted to test the performance of the proposed building energy consumption forecasting system. Experiments reveal that the proposed adaptive LSTM neural network performs better than the existing feedforward neural network and LSTM-based prediction models in accuracy and robustness. It also outperforms those LSTM networks whose hyper-parameters are determined by grid search, Bayesian optimisation and PSO. Such accurate energy consumption prediction can play an essential role in various areas, including daily building energy management, decision making of facility managers, building information model designs, net-zero energy operation, climate change mitigation and circular economy. [ABSTRACT FROM AUTHOR]

Optimizing collaborative operations for yard cranes (YCs) and yard trucks (YTs) is vital to the overall performance of a container terminal. This research investigates four different hybrid approaches developed for dealing with yard crane scheduling problem (YCSP) and yard truck scheduling problem (YTSP) simultaneously for export containers in the yard side area of a container terminal. First, these approaches use a load-balancing heuristic to assign containers to YCs evenly. Following this, each of them employs a specific heuristic/metaheuristic, such as genetic algorithm (GA), particle swarm optimization (PSO) or subgroups PSO (SGPSO), to generate alternative container loading sequences for each YC. Finally, a simulation model is used to simulate loading and transporting of these export containers, evaluate alternative planning results, and finally output the best planning result. Experiments have been conducted to compare these hybrid approaches. The results show Hybrid4 (SGPSO) outperforms Hybrid1 (Sort-by-bay), Hybrid2 (GA), and Hybrid3 (PSO) in terms of makespan. [ABSTRACT FROM AUTHOR]

• A yard sharing strategy is to balance the shortage and surplus of two yards. • A multi-objective model was developed about container storage space assignment. • Travel distance, imbalance of two yards and shared storage space are considered. • Fuzzy theory and trigonometric weight were introduced. • The problem is analyzed and solved by a non-dominated sorting genetic algorithm II. Facing the shortage of storage space of container terminal yard, a yard sharing strategy that uses dry port's surplus storage space to ease container congestion is proposed. This novel strategy can address the container storage space assignment problem for inbound containers. The problem is studied based on the storage yard of the combined container terminal and dry port. First, a multiple-objective mixed integer programming model that considers yard sharing strategy with the objectives of minimizing total travel distance, minimizing imbalance in number of containers, maximizing shared storage space of the dry port is formulated to obtain optimal solutions. Second, a non-dominated sorting genetic algorithm II (NSGA-II) is proposed. Next, the performance of the algorithm is verified by a set of instances. Numerical experiments are conducted to elucidate the problem with yard sharing strategy intuitively. Furthermore, the performance of the model in four aspects proclaims the advantages of yard sharing strategy and certifies the comprehensiveness. Finally, sensitivity analysis is conducted by two aspects which are weight coefficient and feasible distance to verify the efficiency of the proposed method. [ABSTRACT FROM AUTHOR]

This paper studies an intelligent maritime search and rescue (SAR) system problem. According to historical accidents and available SAR equipment information, a bi-level mixed-integer programming (MIP) model is proposed to determine the type and number of SAR equipment allocated to activated stations. Particle swarm optimization (PSO) algorithm and genetic algorithm (GA) algorithm are applied to solve the proposed mathematical model. Computational experiments based on real instances in the East Sea China not only validate the effectiveness of the bi-level MIP model in balancing two objectives during decision process, but also indicate that PSO algorithm is better than GA algorithm to solve the proposed model and generate reasonable equipment allocation plans. Some managerial implications are also outlined on the basis of the numerical experiments. [ABSTRACT FROM AUTHOR]

Surrogate models have been widely used in engineering design because of their capability to approximate computationally complex engineering systems. In practice, the choice of surrogate models is extremely important since there are many types of surrogate models, and they also have different hyper-parameters. Traditional manual selection approaches are very time-consuming and cannot be generalized. To address these challenges, an evolutionary algorithm (EA)-based approaches are proposed and studied. However, they lack interpretability and are computationally expensive. To address these gaps, we create a rule-based method for an automatic surrogate model selection called AutoSM. The drastic increase in the selection pace by pre-screening of surrogate model types based on selection rule extraction is the scientific contribution of our proposed method. First, an interpretable decision tree is built to map four critical features, including problem scale, noise, size of sample and nonlinearity, to the types of surrogate model and select the promising surrogate model; then, a genetic algorithm (GA) is used to find the appropriate hyper-parameters for each selected surrogate model. The AutoSM is tested with three theoretical problems and two engineering problems, including a hot rod rolling and a blowpipe design problem. According to the empirical results, using the proposed AutoSM, we can find the promising surrogate model and associated hyper-parameter in 9 times less than other automatic selection approaches such as concurrent surrogate model selection (COSMOS) while maintaining the same accuracy and robustness in surrogate model selection. Besides, the proposed AutoSM, unlike previous EA-based automatic surrogate model selection methods, is not a black box and is interpretable. [ABSTRACT FROM AUTHOR]

Configuration has been recognized as an effective methodology to provide high product variety that caters to individual customer's needs in the manufacturing industry. For healthcare service configuration, this decision-making process can be taken as service package creation and treated analogously as a project, which is defined as a collection of tasks. This research develops a decision support model to integrate individual patients into the healthcare service configuration process. The healthcare service configuration is formulated as a Resource Constrained Project Scheduling Problem (RCPSP), and a bi-level optimization algorithm based on Genetic Algorithm (GA) is developed for problem solving. The methodology and algorithm are implemented with a case study based on the data obtained from a general hospital in Singapore, which has demonstrated the applicability of healthcare service configuration. [ABSTRACT FROM AUTHOR]

Yard template is a space assignment at the tactical level, which is kept unchanged within a long period of time and significantly impacts the handling efficiency of a container terminal. This paper addresses a yard template planning problem considering uncertainty and traffic congestion. A two-stage stochastic programming model is formulated for minimizing the risk of containers with no available slots in the designated yard area and minimizing total transportation distances. The first-stage model is formulated for assigning vessels in each block without considering the physical location properties of blocks, and the second-stage model is formulated for designating physical locations to all blocks. Subsequently, a solving framework based on genetic algorithm is proposed for solving the first-stage model, and the CPLEX (a commercial solver) is used for solving the second-stage model. Finally, numerical experiments and scenario analysis are conducted to validate the effectiveness of the proposed model and the efficiency of the proposed solution approach. [ABSTRACT FROM AUTHOR]