Search

Evolutionary algorithms built on Pareto-dominance suffer from a loss of selection pressure as the number of objectives increases and the probability of finding non-dominated solutions in the population decreases. Furthermore, Pareto-dominance is computationally expensive because of the pairwise comparisons necessary to rank the individuals of a population. The paper introduces a new genetic algorithm for multi-objective optimization based on Apparent Front Ranking and crowding distance, called Controlled Apparent Front Zones Genetic Algorithm (CAFZGA). To avoid pairwise comparisons, CAFZGA first generates the main Apparent Front Boundary (AFB) with the help of a small set of support vectors and creates secondary AFBs as shifted versions of the main AFB. The space of objectives is divided into zones by the AFBs. The zones play a similar role in CAFZGA as the non-dominated fronts do in the Controlled Non-dominated Sorting Genetic Algorithm CNSGA-II. The zone ranking becomes the main criterion in differentiating individuals, with crowding distance being the tiebreak criterion. The method is shown to be extremely flexible because the AFBs are adjusted for each generation. Computationally, CAFZGA is more efficient than GAs using Pareto-dominance because the set of support vectors for generating the AFBs is significantly smaller than the population. CAFZGA is applied to the problem of optimizing the configuration of a Grid ALU Processor (GAP) and is shown to be competitive and sometimes even better than well-established GAs like CNSGA-II and Fuzzy-Dominance-Driven Genetic Algorithm FDD-GA. [Display omitted] [ABSTRACT FROM AUTHOR]

In recent years cryptographic tokens have gained popularity as they can be used as a form of emerging alternative financing and as a means of building platforms. The token markets innovate quickly through technology and decentralization, and they are constantly changing, and they have a high risk. Negotiation strategies must therefore be suited to these new circumstances. The genetic algorithm offers a very appropriate approach to resolving these complex issues. However, very little is known about genetic algorithm methods in cryptographic tokens. Accordingly, this paper presents a case study of the simulation of Fan Tokens trading by implementing selected best trading rule sets by a genetic algorithm that simulates a negotiation system through the Monte Carlo method. We have applied Adaptive Boosting and Genetic Algorithms, Deep Learning Neural Network-Genetic Algorithms, Adaptive Genetic Algorithms with Fuzzy Logic, and Quantum Genetic Algorithm techniques. The period selected is from December 1, 2021 to August 25, 2022, and we have used data from the Fan Tokens of Paris Saint-Germain, Manchester City, and Barcelona , leaders in the market. Our results conclude that the Hybrid and Quantum Genetic algorithm display a good execution during the training and testing period. Our study has a major impact on the current decentralized markets and future business opportunities. [ABSTRACT FROM AUTHOR]

There is a coupling effect between the liquid and the solid in the hydraulic engine mount (HEM). The accurate estimation and optimization of the dynamic characteristics including dynamic stiffness and lag angle for HEM in the frequency domain are still intractable problems. To this end, a novel model of a deep neural network (DNN)-based on dynamic modeling method is developed by using dataset to estimate the dynamic characteristics, and coupled with a genetic algorithm (GA) for the optimization design. Here, the dataset can be divided into two parts. One part is the input of DNN model and contains feature parameters from simulation; other part is the output of DNN model and composed of dynamic stiffness and lag angle from experiment. They are applied to train, test and validate the DNN model. Besides, the conventional model based on the lumped parameter is also presented to achieve the dynamic characteristics and used for comparison. The performed experiments demonstrate that the estimation accuracy of the DNN model is higher than that of the lumped parameter model. Finally, an optimal design method for the lag angle corresponding to frequency is proposed by combining the DNN model and the GA under the prescribed cost function and constraint conditions. The optimization results are approximately close to the desired values and verify the effectiveness of the proposed method which can improve the isolation performance of HEM. • A novel model of a deep neural network (DNN)-based on dynamic modeling method is developed by using dataset to estimate the dynamic characteristics for hydraulic engine mount (HEM). • A unique data driven-based approach by integrating genetic algorithms into DNN for optimizing HEM is presented. • The integration algorithm allows for fast and effective optimization of dynamic characteristics, providing a practical solution for real-world industrial applications. [ABSTRACT FROM AUTHOR]

Multi-surrogate assisted evolutionary algorithms, utilizing inexpensive global and local surrogate models to assist evolutionary search, have demonstrated a remarkable ability to solve expensive optimization problems (EOPs). The selected training samples of the local surrogate model used in the current algorithms may not be updated continuously, which could bring about a waste of computing resources. This paper proposes a multi-surrogate multi-tasking genetic algorithm with an adaptive training sample selection strategy (MS-MTGAwA), in which the local surrogate model is updated adaptively based on information from established local models to enhance the capability in exploiting the optimal solution. The adaptive training sample selection strategy (ATS) applies optimal points found in historical local models as training samples. In addition, MS-MTGAwA inherits the optimization framework of a multi-tasking genetic algorithm. The radial basis function model is chosen as the modeling basis of the global and local surrogate models. A set of common benchmark functions whose dimensions vary from 10 to 100 and the tension/compression spring design problem are adopted to validate the performance of the proposed algorithm. Based on the average best fitness value over thirty independent runs on the benchmark functions, MS-MTGAwA ranks first in Friedman's test compared to five state-of-the-art algorithms including S-JADE, GORS-SSLPSO, SAHO, SAMSO, and MS-MTO. The proposed ATS can be expected to be applied in any algorithms with local surrogate to make full use of historical information of the local model to improve the prediction accuracy of the next local model. If this paper is accepted, MATLAB codes associated with this paper will be uploaded to https://github.com/Zhongbo-Hu/Prediction-Evolutionary-Algorithm-HOMEPAGE. • Use the predictive best of historical local surrogate models to update the local model. • Design an adaptive training sample selection strategy for the local surrogate model. • Develop a framework that combined multi-surrogate and multi-population-based multitasking. • Propose a multi-surrogate multi-tasking genetic algorithm with an adaptive training sample selection strategy. • MSMTGAwA shows strong competitiveness in comparison with other state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Precision Agriculture technology can use high-tech sensors and analytical tools to help farmers better manage their crop yields, allowing them to demonstrate their techniques on the field. The greater use of remote sensing technologies in farm productivity and nitrogen management tools to increase yields and reduce operational costs contributes to reduced energy consumption. This concept describes the need to fulfill nutrient and crop production objectives for the near and long term. Based on our proposed approach, we concluded that the predictors of individual flow occurrences are variables linked to surface temperatures, moisture, and nitrogen availability. We have developed an automated Convolutional Neural Network creation approach genetic algorithms approach to handling image classification problems more efficiently. The best feature of the method is that it is "automated" and doesn't need domain knowledge about CNNs. When compared to conventional encoding, we compare the performance and adaptability of our processing. Moreover, the comparison of individual Genetic algorithms with CNN schemes employed for training the systems was compared to that of other GENCNN schemes ResNet-34, ResNet-50, AlexNet, DenseNet, VGGNet in terms of convergence speed and overall accuracy, among other things. Our studies data indicate that our Soil Nitrogen Management significantly reduced commercial system costs while providing notable outcomes. The experimental results of soil nutrient deficiency 99%and 97% show that the suggested method may be used to automatically identify the most efficient GENCNN model compared to existing models in a wide range of domains. [ABSTRACT FROM AUTHOR]

This study proposes the combined use of a phenomenological model and neural networks trained by experimental data to predict Carbon dioxide adsorption on Zeolitic imidazolate framework – 8 (ZIF-8) considering different temperature and surface area conditions. The Sips model presented the best fit among the models studied and the perceptron networks presented the best performance to represent the phenomenon. The use of genetic algorithms assisted in the optimization of the neural network architecture, in a shorter computational time. The model indicates that at high pressures the area presents a strong positive correlation for carbon capture, however at lower pressures (of the order of 1 bar) the influence was negative, but with less significance. The study contributes to the development of more efficient carbon capture processes, helping to direct future research in the area of greenhouse gas mitigation, particularly in the field of developing products for carbon capture, among other applications involving carbon dioxide adsorption. [Display omitted] • Combination of phenomenological and neural networks modeling. • A novel approach to model CO 2 adsorption on ZIF-8. • The model includes a wide range of temperature and surface area. • Especially at high pressures, surface area had a positive influence on adsorption. • The developed model reached R2 of 0.9996 and MAE of 0.0303. [ABSTRACT FROM AUTHOR]

This study proposed a satellite constellation method to cluster targets using machine learning and achieve the desired revisit performance for each target in an area of interest. Because a typical satellite constellation, such as the Walker method, does not consider the geometrical dynamics between the satellite and Earth's surface target, the target-oriented revisit performance analysis and the application of optimization method are limited. To overcome these limitations, this study developed a satellite constellation method that is customized according to the target using a repeating ground track orbit and machine learning techniques. The proposed model groups aim to maximize satellite revisit performance via K-means clustering and determine the optimal number of satellites required to achieve the desired revisit performance for each target through a quick and accurate accessibility analysis of various target distributions. This study revealed that three to five target clusters are more effective in improving the satellite revisit performance than a typical single target cluster, while identifying that the density of the target distribution is a significant influencing factor. Finally, the proposed model achieved the desired revisit performance, guaranteeing a timely response for 30 major disaster locations in East Asia and validating the model's optimal performance. [ABSTRACT FROM AUTHOR]

A photovoltaic thermal collector (PV/T) is incorporated into the combined cooling, heating, and power (CCHP) system to decrease primary energy consumption, operation cost and carbon dioxide emission. Six CCHP scenarios based on three strategies are investigated for the energy utilization of the CCHP system. Also, a non-dominated sorting genetic algorithm III using competition crossover and opposition-based learning (NSGA-III-CO) is proposed for the six scenarios. The competition crossover determines the starting point of search in the decision space, and more potential regions can be exploited sufficiently. The opposition-based learning (OBL) carries out random searches and facilitates the convergence of the population. In addition, a constraint handling approach (CHA) is presented to decrease the constraint violations of the infeasible individuals and turn them into feasible individuals, and it guarantees the feasibility of the population. Experimental results suggest that each CCHP scenario with PV/T is more efficient than the one with photovoltaic panel (PV) and the one neglecting PV/T and PV, and it achieves the lowest primary energy consumption, operation cost and carbon dioxide emission. Also, NSGA-III-CO and the other three algorithms are applied to the six CCHP scenarios with PV/T, and it obtains the highest hypervolume and coverage rate for each scenario. [Display omitted] • This paper studies dispatch of combined cooling, heating, and power system (CCHP). • The CCHP involves photovoltaic thermal collector. • An improved NSGA-III is proposed for CCHP. • The improved NSGA-III has exhibited good comprehensive performance for CCHP. • A constraint handling approach is presented to satisfy the constraints of CCHP. [ABSTRACT FROM AUTHOR]

The modeling and simulation of engineering materials using constitutive equations requires a large set of optimized coefficients that characterize the hardening or softening behavior. Optimizing this large set of coefficients with multiple constraints is very challenging using conventional optimization methods. This research proposes a novel model-agnostic framework based on genetic algorithms to identify and optimize the set of coefficients of the constitutive equations of engineering materials. The proposed framework demonstrates solution convergence, scalability to available data, and high explainability over a wide range of engineering materials, including titanium-based, iron-based, and aluminum-based alloys. The experimental test data for necessary validation of the computational framework was generated using the MTS fatigue testing machine equipped with a highly sensitive extensometer. The Chaboche unified visco-plasticity material model has been implemented as an example in this study which can deal with both cyclic effects, such as combined isotropic and kinematic hardening, and rate-dependent effects associated with visco-plasticity. The experimentally obtained cyclic response of three different classes of materials was compared with their optimized simulated constitutive equations, and the results were in good agreement. Furthermore, the proposed multi-objective optimization using genetic algorithm methodology avoids local optimality and converges to the optimal solution much faster than commercially available software. [ABSTRACT FROM AUTHOR]

Home health care (HHC) is a fast-growing area of research that has received increasing attention in recent years due to the rise in life expectancy and the fall in birth rates. It is expected to reduce length of stay in hospital as well as to allow patients to receive care and assistance in their home. Several models and algorithms have been proposed to manage the various constraints and criteria. However, multi-objectives HHC routing and scheduling problems (HHCRSP) with practical and complicated constraints as well as Pareto dominance based approaches are still scarce in the literature. In this paper, we focus on modeling and solving the multi-objectives HHCRSP with lunch breaks, multiple time windows, duty length, skill requirement and a mix of soft and hard services. A mixed integer linear programming (MILP) model with three objectives is proposed, which aims to minimize tardiness of soft services and caregivers' waiting times and also to balance their workload. Two approaches, Pareto and decomposition based, with multi-objective evolutionary algorithms are proposed to solve the multi-objective HHCRSP. We developed a memetic algorithm (NSGA-LS), resulting from the hybridization of non-dominated sorting genetic algorithm II (NSGA-II) with multidirectional local search (MDLS), to trade-off between exploration and exploitation. Computational results and performance measures inferred that NSGA-LS has a better performance compared to other algorithms. According to the hyper-volume indicator, NSGA-LS obtained 93.75% better solutions for the tested instances in a pairwise comparison between algorithms while multiobjective evolutionary algorithm based on decomposition (MOEA/D) has a better complexity. Sensitivity analysis of the percentage of soft/flexible services and the maximum allowed tardiness showed that increasing them significantly improved the quality and trade-off of solutions. [ABSTRACT FROM AUTHOR]

This study introduces a novel framework called Reliable Pollution Scheduling and Routing with Multi-Cross-Docking. The core idea behind this framework is to optimize the processing and transportation of products through each vehicle, ensuring that each vehicle serves its specified destinations only once before returning to the cross-dock. The objective of this research is to develop a multi-objective mixed integer programming model that efficiently manages scheduling and routing complexities in a multi-cross-docking system, taking into account uncertainties in demand. The primary goals of the model are to minimize the overall costs associated with pollution, loading/unloading, and transportation, while simultaneously reducing distribution and shipping durations. Additionally, the model aims to maximize the reliability of the supply chain for perishable products, contributing to sustainable development in the supply chain domain. To address the inherent complexity of the model, especially in dealing with demand uncertainties, we employ a robust optimization method. The multi-objective challenge is tackled using an innovative hybrid approach that combines goal programming and genetic algorithms. The effectiveness of our proposed solution strategies is rigorously assessed using a variety of metrics and subjected to comprehensive statistical testing. Furthermore, we validate the competence of our methodology by conducting a real-world case study, which includes sensitivity analysis of demand parameters and robustness analysis. Our findings confirm that our solution technique produces high-quality solutions. Notably, our approach optimizes route planning for delivery and pick-up phases, resulting in a significant 36.5% reduction in transportation time and a noteworthy 17.27% decrease in the overall system costs compared to existing conditions. Additionally, we observe a substantial 23.15% reduction in vehicle arrival times at cross-docks, contributing significantly to reduced product expenses. [ABSTRACT FROM AUTHOR]

In recent decades, medical tourism especially cosmetic tourism has increased in many countries. In this study, a bi-objective mathematical model is designed for the cosmetic tourism supply chain for the first time. In the first objective function, the profit margin in the supply chain is maximized. In contrast, the distance between the medical and accommodation centers is minimized in the second objective function. The general revised multi-choice goal programming (GRMCGP) method is an exact method for optimizing the problem. To solve the problem, model is run by Lingo software. Then, a new hybrid genetic algorithm named genetic algorithm-general revised multi-choice goal programming (GA-GRMCGP), to solve large-scale instances, is utilized. Moreover, a case study is considered to validate the proposed mathematical model. Furthermore, sensitivity analyses are carried out to evaluate the effectiveness of the research. Finally, the results of the exact and meta-heuristic algorithms are compared and discussed. [ABSTRACT FROM AUTHOR]

The problem of flexible job shop batch scheduling represents an extension of the flexible manufacturing problem. Different batching strategy significantly affects the effect of subsequent scheduling optimization. Therefore, collaborative optimization of the batching strategy and subsequent scheduling is crucial for improving production efficiency. In this paper, a flexible job shop batch production model considering the set-up time is established by analyzing the coupling relationship between batching strategies and scheduling optimization. Based on the proposed model, an improved genetic algorithm with a three-layer coding mechanism is developed to solve the problem of the solution space expansion caused by batch collaboration. The three-layer coding mechanism includes the batch strategy, operation selection, and machine selection. In addition, an operation overlapping strategy is introduced to improve production efficiency and make production more resilient. The proposed algorithm is tested by experiments on scheduling problems of different sizes. The experimental result shows that the proposed method can significantly improve and efficiently optimize flexible job shop scheduling in batch production. [ABSTRACT FROM AUTHOR]

Optimal aiming strategies are essential for efficient solar power tower technology operation. However, the high calculation complexity makes it difficult for existing optimization methods to solve the optimization problem in real-time directly. This work proposes a real-time optimal heliostat aiming strategy generation method via deep learning. First, a two-stage learning scheme where the neural network models are trained by genetic algorithm (GA) benchmark solutions to produce an optimal aiming strategy is presented. Then, an end-to-end model without needing GA solutions for training is developed and discussed. Furthermore, a robust end-to-end training method using randomly sampled flux maps is also proposed. The proposed models demonstrated comparable performance as GA with two orders of magnitude less computation time through case studies. Among the proposed models, the end-to-end model shows significantly better generalization ability than the pure data-driven two-stage model on the test set. A robust end-to-end model with data enhancement has better robustness on unseen flux maps. [ABSTRACT FROM AUTHOR]

In recent decades, research on supply chain management (SCM) has enabled companies to improve their environmental, social, and economic performance. This paper presents an industrial application of logistics that can be classified as an inventory-route problem. The problem consists of assigning orders to the available warehouses. The orders are composed of items that must be loaded within a week. The warehouses provide an inventory of the number of items available for each day of the week, so the objective is to minimize the total transportation costs and the costs of producing extra stock to satisfy the weekly demand. To solve this problem a formal mathematical model is proposed. Then a hybrid approach that involves two metaheuristics: a greedy randomized adaptive search procedure (GRASP) and a genetic algorithm (GA) is proposed. Additionally, a meta-learning tuning method is incorporated into our hybridized approach, which yields better results but with a longer computation time. Thus, the trade-off of using it is analyzed. An extensive evaluation was carried out over realistic instances provided by an industrial partner. The proposed technique was evaluated and compared with several complete and incomplete solvers from the state of the art (CP Optimizer, Yuck, OR-Tools, etc.). The results showed that our hybrid metaheuristic outperformed the behavior of these well-known solvers, mainly in large-scale instances (2000 orders per week). This hybrid algorithm provides the company with a powerful tool to solve its supply chain management problem, delivering significant economic benefits every week. [ABSTRACT FROM AUTHOR]

This paper presents a new initialization method that combines adaptive visibility graphs and the A* algorithm to improve the exploration, accuracy, and computing efficiency of hybrid path planning for mobile robots. First, segments/links in the full visibility graphs are removed randomly in an iterative and adaptive manner, yielding adaptive visibility graphs. Then the A* algorithm is applied to find the shortest paths in these adaptive visibility graphs. Next, high-quality paths featuring low fitness values are chosen to initialize the subsequent heuristic optimization in hybrid path planning. Specifically, in the present study, the genetic algorithm (GA) is implemented on a CPU/GPU edge computing device (Jetson AGX Xavier) to exploit its massively parallel processing threads, and the strategy for judicious CPU/GPU resource utilization is also developed. Numerical experiments are conducted to determine proper hyperparameters and configure GA with balanced performance. Various optimal paths with differential consideration of practical factors for robot path planning are obtained by the proposed method. Compared to the other benchmark methods, ours significantly improves the diversity of initial path and exploration, optimization accuracy, and computing speed (within 5 s with most less than 2 s). Furthermore, real-time experiments are carried out to demonstrate the effectiveness and application of the proposed algorithm on mobile robots. • The adaptive visibility graph eliminates the issue of biased path search. • A new approach generates multiple initial paths with enriched diversity. • Judicious utilization of CPU/GPU resources improves computing efficiency. • Implementation on light-weight edge device is well-suited for robot applications. [ABSTRACT FROM AUTHOR]

In the era of internet and big data, recommender systems are necessary to filter out useless information. Collaborative filtering (CF) is one of the most successful technique used in recommender systems, this technique suffers from a large number of users and items, or the scalability issue. In this paper, a new hybrid method based on K-means clustering (KM) and Singular Value Decomposition (SVD) which uses evolutionary algorithms is proposed to deal with scalability issue. On the one hand, KM optimized by Particle Swarm Optimization (PSO) and denoising by Density-Based Spatial Clustering of Applications with Noise (DBSCAN) is used to cluster users and reduce the number of comparisons between the target user and other users. On the other hand, SVD optimized by Genetic Algorithm (GA) is used to reduce the number of items. The proposed method is assessed on three standard datasets and the results are compared with basic CF and other extended versions that use clustering algorithms, evolutionary algorithms and dimensionality reduction techniques. The results show that our proposed method performed better than other methods in terms of precision , recall and MAE , and the scalability problem was improved by reducing the time complexity. Also, the combined clustering method was optimized in terms of Davies-Bouldin and the Dunn's index compared to the basic clustering methods. • The quality of the recommender system was improved by combining model-based and memory-based collaborative filtering methods. • The problem of scalability has been solved by using different clustering methods and matrix dimensionality reduction method. • Combination of K-means clustering and DBSCAN was performed to clean the clusters from noise. • The quality of clustering and matrix dimensionality reduction has been improved by using evolutionary algorithms. [ABSTRACT FROM AUTHOR]

The design of the suspension model of road cars is extremely important for both driver comfort and safety. Road traffic injuries are currently the eighth leading cause of death worldwide and are predicted to become the seventh leading cause of death by 2030, according to the World Health Organization (WHO). Among other factors, road infrastructure is one of the most important factors to consider when designing the suspension model, especially as autonomous driving becomes more popular. In addition, there is a growing need to test and study the impact of road roughness and pavement type on autonomous vehicles (AVs). This work provides the following contributions: a platform for suspension system optimization–flexible, highly parameterizable and rich in applications. Our solution implements, tests and optimises the 2, 3 and 4 degrees of freedom (DOF) quarter-car models (QCM) with linear seat suspension and the 5 DOF half-car model (HCM), respectively, based on random road profiles generated according to ISO 8608 standards. In addition, many multi-objective optimization (MOO) algorithms have been implemented and applied in isolation to see their comparative performance (in terms of Pareto fronts, hypervolume, coverage, spread, and epsilon), but most importantly we have also introduced a meta-optimization technique by super-positioning Non-dominated Sorting Genetic Algorithm II and Strength Pareto Evolutionary Algorithm 2, where each of the algorithms, running concurrently, produces a percentage of the total number of individuals in a generation, percentage that is calculated according to the solution quality of the individuals in the previous generation. The experimental results were obtained by simulating the optimal configurations generated by the MOO heuristic algorithms on different categories of cars (small/medium). Finally, the Taguchi method and regression analysis were used to evaluate the effects of specific parameters on system responses such as sprung mass acceleration and displacement. The Taguchi simulation was performed to experimentally validate the results obtained by metaheuristic optimization methods. [ABSTRACT FROM AUTHOR]

Upper limb amputation is a widespread problem worldwide, leading to massive loss of functionality for the victims. While a few solutions exist, these are often very expensive and involve expensive and dangerous surgical procedures. In this paper, we propose a low-cost, highly functional upper limb prosthesis controlled via Electroencephalogram (EEG) signals captured in real-time, classified into upper limb motion intention using a novel Genetic Algorithm (GA) optimized Long Short-Term Memory (LSTM) deep learning model to rehabilitate amputees. The proposed 3D-printed prosthetic arm has 3 Degrees-of-Freedom (DOF), which allows it to perform complex movements that can accurately emulate an actual human arm. The results discussed later in this paper conclude that our approach, while being low-cost, is highly accurate and functional. [ABSTRACT FROM AUTHOR]

This paper presents an aircraft conceptual design technique with more than three objective functions, called many-objective optimisation. The selection of aircraft configuration is usually achieved using a system engineering approach. This selection approach has the design variables assigned to remove the configuration decision-making process. The design problem is demonstrated for the conceptual design of a fixed-wing unmanned aerial vehicle. Eight objective functions, including power required, take-off weight, take-off distance, landing distance, endurance, range, lift coefficient at cruise and drag coefficient at cruise, are posed, while the constraints are aircraft stability, performance required and take-off distance. Design variables simultaneously determine an aircraft configuration, shape and sizing parameters. Hence a new, many-objective metaheuristic is developed to increase the design performance. A grid-based many-objective metaheuristic with iterative parameter distribution estimation (MM-IPDE-Gr) is developed. It is an enhanced variant of the MM-IPDE with improved reproduction schemes, adaptive parameters and a grid-based clustering technique. Several additional reproduction schemes in mutation and crossover processes with two additional adaptive parameters are integrated to increase population diversity and improve the exploration ability of the algorithm. In addition, the grid-based method is integrated as a clustering technique to improve the Pareto clustering process in many-objective optimisation. The proposed method, with established newly invented metaheuristics, is used to solve the new design problem and its performance compared with existing design methods. It is shown that the proposed many-objective metaheuristic gives the best results. [ABSTRACT FROM AUTHOR]

The three-fluid heat exchanger is a multi-stream heat exchanger that provides three streams at different temperatures for certain usage objectives. Recently, optimizing heat exchangers generally and three-fluid heat exchangers especially have received great attention in the energy field. The current study objective is to get the optimal mass flow rate of the three streams to minimize the entropy generation number and maximize effectiveness. Optimization is performed on the inserted coiled tube type for 24 cases with different usage objectives and different mass flow rate constraints. The optimization is done using an analytical approach for the temperature profiles and a regressed model for the overall heat transfer coefficients. Single-objective and multi-objective genetic algorithms are the optimization algorithms. The result shows that cooling the hot fluid is the most efficient usage objective; however, heating the intermediate-temperature fluid is the less efficient usage objective. Also, the study shows that the entropy generation number is dependent on the mass flow rate constraint. The maximum effectiveness of 0.693 occurs when the usage objective is cooling the hot fluid. In this case, the mass flow rate of the cold and the intermediate-temperature fluids are maxima while the mass flow rate of the hot fluid is minimum. The minimum entropy generation number 0.0486 occurs when the mass flow rate of the cold fluid is maximum while the mass flow rate of the hot fluid and intermediate-temperature fluid are minimal. [Display omitted] • Heat transfer of a three-fluid heat exchanger is investigated. • A model for optimizing the inserted coiled tube three-fluid heat exchanger is developed. • The maximum errors for the regression models of heat transfer coefficients are 1.72% and 2.81%. • The maximum effectiveness reached about 0.69323. [ABSTRACT FROM AUTHOR]

Assessing project portfolio risk (PPR) is essential for organizations to grasp the overall risk levels of project portfolios (PPs) and realize PPR mitigation. However, current research is inadequate to effectively assess PPR, which brings challenges to managing PPR. In this context, the purpose of this study is to develop a PPR assessment model via an enhanced backpropagation neural network (BPNN). First, PPR assessment criteria considering project interdependencies are determined. Second, fuzzy logic is used to obtain original data for assessment criteria. Principal component analysis (PCA) is then employed to reduce the dimensionality of assessment criteria and derive the input and output of BPNN. Third, an improved genetic algorithm (IGA) is designed to optimize the initial weights and thresholds of BPNN. On this basis, the PCA-IGA-BPNN assessment model is constructed, followed by training and testing, possessing a test accuracy of 98.6%. Finally, comparison experiments are conducted from both internal and external perspectives. For internal comparison, the proposed model yields less mean absolute percentage error (MAPE), mean square error (MSE), and root mean square error (RMSE) than PCA-GA-BPNN, IGA-BPNN, PCA-BPNN and BPNN and offers the largest convergence speed (γ). As for external comparison, the presented model produces lower MAPE, MSE, and RMSE than Decision Tree (DT), Support Vector Machine (SVM), Random Forest (RF), and K-Nearest Neighbor (KNN) and has the largest coefficient of determination (R 2). Results indicate that the established model performs more satisfactorily in assessing PPR. This research enriches PPR assessment methods and provides managers with a useful tool to evaluate PPR. • A hybrid project portfolio risk assessment model based on an enhanced BPNN is established. • GA is improved by adaptively adjusting the probability of crossover and mutation operations. • PCA and improved GA are integrated with BPNN to enhance its evaluation performance. • The effectiveness and superiority of the proposed model are validated through experiments. [ABSTRACT FROM AUTHOR]

Solar PV generation is renewable energy source that in the last years has contributed to reduce the use of fossil fuels. Controlling the efficiency of photovoltaic (PV) installations is essential in order for its use to spread. Deep learning (DL) models have demonstrated their efficiency in this context. The purpose of this work is to show a methodology to optimised deep learning models and show a specific application of the optimised model to characterise PV installations. The estimations yielded by this model are obtained without the need for real PV data to the training process; synthetic data are used. The built multivariate neural network is optimised through the use of a multiobjective genetic algorithm and the use of a feature engineering tool based on functional data analysis (FDA) clustering. This method was applied on synthetic data and on a PV installation located in north-western Spain, where the number of parallel modules, the azimuth and the slope in the installation are estimated. The results show that the model optimised using the non-dominant sorting genetic algorithm (NSGA-II) and the customised dataset with FDA, achieve lower errors or in the same range by reducing the complexity of the model or the complexity of the dataset used. Specifically, the bests models generates estimations with an average error between 6% and 18% on synthetic data and between 6% and 12% on real data. • An optimisation methodology is proposed to select the best LSTM architecture. • The optimisation is based on the multiobjective genetic algorithm NSGA-II. • A feature engineering tool based on FDA clustering is also proposed. • The models are evaluated based on the PV installations characterisation. • The results show the improvements provided by the optimisations presented. [ABSTRACT FROM AUTHOR]

The proportional–integral regulator is one of the most famous and widely used linear controllers in the industrial field because of its easy adjustment of the results, simplicity, and inexpensive. The negative of this controller is that it is affected by the change of system parameters, which causes several problems, including the high value of ripples and low quality of the power/current. To reduce this defect, a new proportional–integral regulator is suggested based on the use of the terminal sliding surface technique and genetic algorithm. This suggested nonlinear controller is discussed for the first time in this work to use it to ameliorate the quality of energy output from a dual-rotor wind power system using an induction generator. The use of a new controller to regulate the generator energy will inevitably lead to an increase in power errors, power ripples, and harmonic distortion of current especially if the system parameters are changed. Also, the traditional direct power control based on the proposed nonlinear controllers of the induction generator is a robust control, does not need the mathematical form of the generator, is simple, and does not need a specialist. Moreover, the suggested method is applied to the rotor inverter of the 1.5 MW generator using Matlab software. The suggested technique takes full advantage and the power regulation objective of the suggested power system is confirmed by the numerical results compared to the conventional technique and other published controls. [ABSTRACT FROM AUTHOR]

The efficient coordination of directional overcurrent relays is a challenging problem. Recent papers in the literature have focused on developing efficient algorithms to solve this issue. However, there is a lack of reporting practical procedures that affects the mathematical formulation to ensure accurate outputs by any optimization algorithm. In this paper, we show a comprehensive procedure that considers mutually coupled circuits via Graph Theory, the requirements of short-circuit current calculation, contingencies to be considered as well as the setting of Instantaneous Overcurrent (IOC) and Time Overcurrent (TOC) units, both for phase and ground protection. The simulation has been carried out by using a real electric network and C++ language programming. Computational simulation has shown the performance of the method, which provides high-quality solutions, with adequate coordination of the selected protective devices that increase the efficiency in setting this type of protection with safety. Evolutionary Particle Swarm Optimization (EPSO) outperformed conventional Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) as the best metaheuristic for optimally minimizing the proposed problem, achieving faster convergence times for the phase overcurrent protection scenario (16.36 s vs. 22.27 s and 17.03 s, respectively), and also for the ground overcurrent protection scenario (19.87 s vs. 20.44 s and 20.11 s, respectively). [Display omitted] • Relay coordination considering devices from different manufactures. • Contingencies in transmission lines formed by several mutually coupled circuits. • Inclusion of mutual impedances in the relay coordination and they are processed using Graph Theory. • An optimization model that combines three performance indices using metaheuristics. • A new strategy to determine pick-up current values for phase IOC unit. [ABSTRACT FROM AUTHOR]

Several optimization algorithms have been developed in the past for application in a variety of complex real world problems. Herein, a new hybrid variant of Genetic Algorithm (GA), utilizing the sample space reduction technique of Cohort Intelligence (CI) Algorithm, referred to as Adaptive Range Genetic Algorithm (ARGA) is presented and discussed. The correctness of the proposed algorithm is established by solving 50 standard test functions. Also, the practical applicability of the proposed algorithm is demonstrated by applying it to the design and economic optimization problem of a Shell and Tube Heat Exchanger. Heat exchangers are critical industrial components used in a wide range of industrial applications. Owing to their high manufacturing costs and complicated design criteria, their economic efficiency of operation depends heavily on optimization of their design. A relative assessment, comparing the performance of the proposed method of solution to other existing algorithmic methods, is conducted. The quality and robustness of the proposed solution method highlights its adeptness in solving real world computational problems in core mechanical engineering applications. [ABSTRACT FROM AUTHOR]

The segmentation of multivariate temporal series has been studied in a wide range of applications. This study investigates a challenging segmentation problem on traffic engineering, namely, identification of time-of-day breakpoints for pre-fixed traffic signal timing plans. A large number of urban centres have traffic control strategies based on time-of-day intervals. We propose a bilevel optimization model to address simultaneously the segmentation problems and the traffic control problems over these time intervals. Efficient memetic algorithms have been developed for the bilevel model based on the hybridization of the particle swarm optimization, genetic algorithms or simulated annealing with the Nelder–Mead method. Numerically the effectiveness of the algorithms using real and synthetic data sets is demonstrated. We address the problem of automatically estimating the number of time-of-day segments that can be reliably discovered. We adapt the Bayesian Information Criterion, the PETE algorithm and a novel oriented-problem approach. The experiments show that this last method gives interpretable results about the number of reliably necessary segments from the traffic-engineering perspective. The experimental results show that the proposed methodology provides an automatic method to determine the time-of-day segments and timing plans simultaneously. • A bilevel model for simultaneously determining TODs and traffic signal optimization. • Memetic algorithms based on PSO, SA, GA and Nelder Mead are proposed for bilevel model. • A set of computational tests on synthetic and real data sets. • A novel oriented-problem approach to determine optimal number of TODs. [ABSTRACT FROM AUTHOR]

Nowadays, the medical tourism industry encourages many healthcare practitioners to provide high-quality and low-cost medical services for patients worldwide. The development of operations research models and algorithms is one important instrument for improving the medical tourism industry based on the economic, political, social, and cultural aspects. In this paper, a medical tourism trip design problem is developed where patients travel from their city of residence to a destination city that may be in another country to receive high-quality and low-cost medical care. The most important part of this problem is to visit a number of tourist cities for each patient individually in the destination. In addition to the total cost, the patients prefer to increase the attractiveness of trips by referring to the quality of medical services and the attractiveness of visiting tourist cities. As far as we know in the area of medical tourist studies, no study has considered the minimization of total cost and maximization of the attractiveness of trips, simultaneously using utility function. The proposed multi-objective optimization model assigns the patients from the origin country to the hospitals in the destination country while making their routing and scheduling decisions to visit the tourist cities. The proposed model is limited by patients' interests and time restrictions while allocating patients to the hospital and orienteering the patients toward visiting tourist attractions. As a complex optimization problem, another significant novelty of this paper is the proposal of a local search-based non-dominated sorting genetic algorithm (LSNSGA-II) for solving the proposed multi-objective optimization model. The proposed algorithm is compared with the original non-dominated sorting genetic algorithm (NSGA-II) and epsilon constraint (EC) method based on different multi-objective criteria. Finally, one main finding from our analyses is finding a trade-off between the total cost and attractiveness of trips as a challenging decision while proposing high-quality solutions in a reasonable time (i.e., less than one hour). • A new multi-objective medical tourism trip design problem is developed. • Proposed model minimizes the total cost while maximizing the attractivity of trips. • A utility function is used to compute variable staying time for visiting tourist places. • An efficient local search-based non-dominated sorting genetic algorithm is introduced. [ABSTRACT FROM AUTHOR]

This paper presents an Optimal Robust Model Reference Adaptive Proportional–Integral controller for grid-tied power systems, which is automatically parametrised using a Genetic Algorithm. The task of choosing the adaptation rate and the controller's initial gains set is now performed in a computational and automatic way, saving design time and ensuring satisfactory transient regimes.The use of a Genetic Algorithm for complete parametrisation of robust adaptive controllers applied to renewable energy power generation systems with LCL filter is first here explored. Five cost functions are considered in the optimisation process. A comparison of simulation results applying the optimised controller in a grid-tied Voltage Source Inverter with LCL filter points out the benefits and drawbacks of using each cost function. A framework for performance analysis and an in-depth discussion are provided. The function determined as the best one, which is integral of absolute error, is used in the controller implemented experimentally, ensuring suitable tracking of grid current references and rejection of grid voltage harmonics, even under unbalanced grid voltage conditions. An experimental comparison in a three-phase 6.75 kW power converter with the non-optimised controller is also provided, where the optimised controller obtains a reduction in the overshoot in all transient regimes and smaller tracking error, resulting in a reduction in the mean error of 14.31% and 60.30% in α and β coordinates, respectively. In addition, the optimised controller reduces total harmonic content, improving energy quality. This parametrisation procedure can be immediately extended to the control of any other power converter systems. [ABSTRACT FROM AUTHOR]

Teaching Learning Algorithm (TLA) is a recently developed nature-inspired optimization technique applicable to complex optimization problems. This paper proposes an improved TLA version using the Laplacian operator of the Genetic Algorithm (GA), named LX-TLA. The proposed algorithm is tested on benchmark optimization problems, including unimodal and multimodal problems. The numerical results are obtained in the form of objective function values, and a t-test is applied to compare the performance of LX-TLA and basic TLA. Convergence plots are given to provide insight into the convergence behavior of LX-TLA. The results reveal that proposed algorithm provides effective and efficient performance in solving benchmark test functions. The proposed algorithm is also applied to engineering design problems, such as Tuned Mass Damper (TMD), truss structure, welded beam, tension string, and pressure vessel. The results obtained using LX-TLA are compared with other nature-inspired optimization algorithms. The results demonstrate that the proposed algorithm is a robust and effective tool for solving complex optimization problems. • Proposed an improved Teaching Learning Algorithm using Laplacian Operator of Genetic algorithm. • Tested on benchmarks functions of varying complexity. • Applied to real life problems in civil and mechanical engineering problems. [ABSTRACT FROM AUTHOR]

Dropout is an effective method of mitigating over-fitting while training deep neural networks (DNNs). This method consists of switching off (dropping) some of the neurons of the DNN and training it by keeping the remaining neurons active. This approach makes the DNN general and resilient to changes in its inputs. However, the probability of a neuron belonging to a layer to be dropped, the 'dropout rate', is a hard-to-tune parameter that affects the performance of the trained model. Moreover, there is no reason, besides being more practical during parameter tuning, why the dropout rate should be the same for all neurons across a layer. This paper proposes a novel method to guide the dropout rate based on an evolutionary algorithm. In contrast to previous studies, we associate a dropout with each individual neuron of the network, thus allowing more flexibility in the training phase. The vector encoding the dropouts for the entire network is interpreted as the candidate solution of a bi-objective optimisation problem, where the first objective is the error reduction due to a set of dropout rates for a given data batch, while the second objective is the distance of the used dropout rates from a pre-arranged constant. The second objective is used to control the dropout rates and prevent them from becoming too small, hence ineffective; or too large, thereby dropping a too-large portion of the network. Experimental results show that the proposed method, namely GADropout, produces DNNs that consistently outperform DNNs designed by other dropout methods, some of them being modern advanced dropout methods representing the state-of-the-art. GADroput has been tested on multiple datasets and network architectures. [ABSTRACT FROM AUTHOR]

This paper presents a nondominated sorting genetic algorithm III with three crossover strategies (NSGA-III-TCS) for the combined heat and power dynamic economic emission dispatch (CHPDEED) problems with or without prohibited operating zones. The first strategy updates the individuals of the first nondomination rank by a modified improvisation of the harmony search algorithm, which accelerates the information exchanges of the elite individuals. The second strategy regulates the individuals of the last nondomination rank according to a modified single-point crossover of the genetic algorithm, which strengthens the convergence of the population. The third strategy adjusts the remaining individuals in terms of a modified DE/rand-to-best/1 mutation of the differential evolution algorithm, and it guides these individuals to move towards the ones which have at least one minimum objective function value, which is beneficial for expanding the distribution range of the population. NSGA-III-TCS obtains the largest average hypervolumes and shows a desirable convergence for six CHPDEED problems. NSGA-III-TCS achieves relatively high coverage rates in most cases, indicating the dominance of the Pareto sets of NSGA-III-TCS over those of the other algorithms. Also, it has relatively low average spacings for all problems, suggesting the good evenness of the Pareto sets. • Combined heat and power dynamic economic emission dispatch (CHPDEED) is studied. • A NSGA-III with three crossover strategies (NSGA-III-TCS) is presented for CHPDEED. • The Pareto sets of the NSGA-III-TCS have high hypervolumes and coverage rates. • NSGA-III-TCS obtains good solutions with low generation cost and pollutant emission. [ABSTRACT FROM AUTHOR]

Petroleum production forecasting is the process of predicting fluid production from the wells using historical data. In contrast to the traditional methods of analysing surface and subsurface parameters governing fluid production, machine learning (ML) techniques are being applied to forecast the production. The major drawback of traditional and conventional ML techniques is that they are time-consuming and often lack good forecasting power. In this work, time-series forecast models based on powerful and efficient ML techniques are developed to forecast production with historical data. We have fused the attention mechanism into the long short-term memory network, which is referred as the attention-based long short-term memory (A-LSTM) network. The A-LSTM network is fast and accurate, thus solving the low forecasting power problem. To ensure no data leakage occurs during training, and to build a reliable data-driven forecasting approach, we construct the dynamic floating window with varying window sizes over the entire production data. The dynamic floating window slides one-step forward after every prediction and continues till the last production window enabling the model to fit the new data automatically. We have tested and validated the proposed forecasting models with the ML algorithm using actual production data for three wells from entirely different geographies. We then compared them with statistical, deep learning, hybrid, and ML approaches. The genetic algorithm (GA) is applied to optimize the hyper-parameters of the A-LSTM. The results of a comparative analysis show that the A-LSTM network statistically and computationally outperforms the other models for forecasting petroleum production. [Display omitted] • Attention-based Long Short-Term Memory (A-LSTM) was used for forecasting the production. • A genetic algorithm (GA) was used to optimize the hyper-parameters of A-LSTM. • Seven forecasting models, consisting of several inputs and output were developed and tested. • A well-trained A-LSTM algorithm was observed superior, easy, and computationally efficient. [ABSTRACT FROM AUTHOR]

Data science techniques (DST) are the most popular approaches for estimating the discharge coefficient (C d) of labyrinth weirs (LW). This study proposes a hydraulic-informed multi-layer perceptron (HI-MLP) by incorporating the effects of hydraulic phenomena on C d , which were previously neglected due to their immeasurability. The HI-MLP is not a black box and selects its internal parameters considering the nape behavior attributes. HI-MLP is trained using Levenberg–Marquardt and Genetic algorithms, resulting in HI-MLP-LM and HI-MLP-GA, which are compared against adaptive neuro-fuzzy inference systems (ANFIS), support vector regression (SVR), and standard MLP. Evaluating the models through randomly selected testing data shows that HI-MLP-GA is the most accurate, with an MAEP of 0.732%, while all techniques also had acceptable performances. Conversely, when estimating C d for LWs with excluded intermediate and extrapolated geometries, only HI-MLP-LM and HI-MLP-GA could predict accurate values with average MAEP of 1.94% and 0.62%, respectively. This value in ANFIS and SVR was equivalent to 27.01% and 6.47%, respectively. Furthermore, hydraulic-informed techniques could be trained with at least a 25% smaller dataset, while the robustness analysis indicated that they are less prone to overfitting. HI-MLP-GA has the highest computational effort and is almost five times slower than HI-MLP-LM. Nevertheless, due to the disappearance of the vanishing gradient and considering the higher generalizability of HI-MLP-GA, GA is still a more desirable learning algorithm than LM. [ABSTRACT FROM AUTHOR]

In actual air combat, there is an inevitable risk that an unmanned aerial vehicle (UAV) will be destroyed. However, this risk is rarely considered in the mission planning phase. In this paper, we focus on cooperative mission assignment for heterogeneous UAVs. We develop a multi-objective optimization model to find a balance between mission gains and UAV losses. The objective function is expressed using conditional probability theory by introducing the probabilities of mission success and UAV loss. Munitions loading capacity, time constraints, and priority constraints are modeled as constraints. To solve this combinatorial problem, an improved multi-objective genetic algorithm, which incorporates a natural chromosome encoding format and specially designed genetic operators, is developed. An efficient unlocking method is constructed to address the unavoidable dead-lock phenomenon meanwhile maintaining the population randomness. Numerical simulations for different problem sizes and ammunition stocks are performed, and the proposed algorithm is compared with the Multi-objective Particle Swarm Optimization and the Multi-objective Grey Wolf Optimization, respectively, using different unlocking approaches. The simulation and comparison results demonstrate the practical value and effectiveness of the developed model and the proposed algorithm. • A risk-based multi-objective model for cooperative task assignment is established. • An improved genetic algorithm is proposed with all ammunition situations considered. • A novel fast positioning unlocking method is given to improve unlocking efficiency. [ABSTRACT FROM AUTHOR]

One of the major economic activities of a business concern is transportation of its products without any hassle. Therefore, the research on a transportation problem is very important now-a-days. In a classical transportation problem, the amount incurred for shipping the goods from the source to a destination is determined on the basis of unit transportation cost. But now, in many practical cases, it is seen that the unit transportation cost is settled on the basis of transported amount. In such a direction, this article develops a transportation problem to transport the produced items from a production house to some retailers through some distributors. The system has two cost parameters, one is actual unit transportation cost and other is demanded unit transportation cost. The actual unit transportation cost is the cost incurred during the transportation while the demanded unit transportation cost is the cost demanded by the distributors. Here, the demanded unit transportation cost has been decided depending on the amount of transported item. Since in the system, there are more than one distributor, distributors offer a credit period policy to attract the retailers. On the basis of the deterioration of the loaded amount, the transported items are classified into two categories. For the higher deteriorated items, the transportation cost is relatively larger and for the lower deteriorated items, the transportation cost is relatively lower with the increase in the loading amount. Considering the above two categories, two crisp models are developed. Due to the uncertain information of the actual unit transportation cost, the models are illustrated considering the actual unit transportation cost as a triangular intuitionistic fuzzy number. The main objective of this research is to minimize the total cost of the retailers and to maximize the total profit of the distributors simultaneously under the above circumstances. For crispification, two different methods have been used here — one is an existing method namely ordered weighted average operator and other is a developed average value approach. Also, the elitist non-dominated sorting genetic algorithm (NSGA-II) and the fuzzy programming technique have been used to get the optimal solution of this problem. To justify the impact and significance of the proposed models on this decision making system, some numerical studies have been done and it is observed that distributors and retailers both benefit more under the credit period policy than the policy without credit period. This study discusses the situation under which the uncertain parameters are to be expressed in terms of triangular intuitionistic fuzzy numbers. Also, it is identified that more accurate crispification has been done by the average value approach than the existing approach. [ABSTRACT FROM AUTHOR]

In this paper, we propose a multi-vehicle clustered traveling purchaser problem (MVCluTPP). Here, two types of procurement planning are proposed. In the first setup, the purchaser visits the markets clusterwise and, after satisfying the demand, returns to the depot with the purchased products, which are carried on the same path using a different vehicle. The other set up, in which products are purchased clusterwise and transported directly from the center of the cluster to the depot, but with the mandate that the purchaser visits the markets clusterwise. One of the multi-pronged aims of the model is to select the clusters and identify which markets to visit, determine the amount of procurement available in each market in a cluster, and develop an optimal routing plan in such a way that the overall system cost is minimized. The clusters are generated using k-means algorithm, and a variable-length chromosome genetic algorithm (VLC-GA) is proposed to optimize the cluster paths and to use a local heuristic to link the clusters to minimize the system cost. Furthermore, the superiority of the VLC-GA has been established through standard TPP and TSP instances, compared with exact methods, and some statistical tests are presented. • The mathematical development of the clustered traveling purchaser problem. • Cluster-wise procurement planning in different scenarios. • A variable-length genetic algorithm is developed. • The developed algorithms outperform classical algorithms. [ABSTRACT FROM AUTHOR]

The hub location is a major problem in the systems with high dependence on the transportation of goods, information, and passengers between the constituent parts. The present study is aimed to introduce a multi-commodity multi-model hierarchical hub location problem with uncertain demand. Thus, it is estimated by intuitionistic fuzzy variables. The objective function in the present problem is to minimize the total transportation costs in the network to determine the optimal hub location, allocate non-hub nodes to the hubs and the ground hubs to the air hubs, and identify the type of vehicles required in each route. The model has a hierarchical structure consisting of a three-echelon network (i.e., central hubs, non-central hubs, and demand nodes) connected in the star-tour-star form. A four-index mathematical model was presented and solved by GAMS optimization software. The results indicated that the objective function (cost) has a descending trend with increasing the number of hubs which can be assigned to the decline in the transportation costs. The costs increased by a decrease in the maximum delivery time which could be due to the selection method of the transportation mode. In the multi-product case, the costs declined by increasing the number of hubs as these hubs decremented the network complexity. Regarding the NP-Hard nature of the hub problems, a genetic meta-heuristic algorithm was proposed in the MATLAB software to solve the large-scale problem. According to the results, this algorithm offered close-optimal solutions. Noteworthy, this study employed known AP datasets. [ABSTRACT FROM AUTHOR]

Task-oriented design of modular robot topology is a crucial issue to be solved in modular robot research. In this paper, a Genetic algorithm-based optimal design strategy for the modular robot topology is proposed. Four tuples are used to represent the modular robot topology. By establishing a database of motion transmission forms for a modular unit, a distributed parallel kinematic modeling and analysis method for modular robots is introduced. Taking the four-tuple representation of the modular robot topology as decision variables, an optimization model of the topology design for modular robots is established. To solve the model, under the framework of the Genetic algorithm, the genotype and genetic operations are designed to improve the effectiveness of individual generating in the optimization process. Finally, a modular robot is taken as an example to carry out simulations. It indicates that the modular robot under the designed topology can successfully execute tasks with as few modular units as possible, and the computing time for the kinematics modeling and analysis of modular robots can be reduced by at least an order of magnitude. • Topology is a type of information about the connection state of a modular robot. • Topology design can make modular robots more adaptive to tasks. • The application of optimization can improve the efficiency of topology design. • Robot kinematics with modular modeling and parallel computing has high efficiency. [ABSTRACT FROM AUTHOR]

This paper addresses the rescue unit allocation and scheduling problem (RUASP) with fuzzy processing times through an evolutionary approach. The goal of RUASP is to efficiently assign and schedule the rescue units to process the incidents in the event of a natural disaster. This problem can be considered as a variant of the unrelated parallel-machine scheduling problem with sequence and machine-dependent setup times. A steady-state grouping genetic algorithm (SSGGA) approach is presented to minimize the total weighted completion time of the incidents, where the weights correspond to the severity levels of the incidents. The crossover and mutation operators used in this approach are designed as per the characteristics of RUASP and its objective. The proposed approach uses a combination of greedy and random heuristics while generating the initial solutions, thereby yielding superior quality diverse initial solutions. The performance of the proposed approach is compared with the state-of-the-art approach available in the literature. The proposed approach always yields better results in much shorter execution times in comparison to the state-of-the-art approach. Moreover, the robustness analysis test demonstrates the proposed approach to be more robust in comparison to the state-of-the-art approach. [ABSTRACT FROM AUTHOR]

Variable Selection in large datasets is a commonly procedure in multivariate calibration, which is a field of study from chemometrics. Selecting the most informative variables becomes an important step to build mathematical models through statistical techniques in order to predict some property of interest from an analyzed sample. Recombination-based search methods such as Genetic Algorithms (GAs) have been widely used as variable selection techniques to solve several optimization problems. However, previous works from literature have emphasized the schemata disruption problem caused by genetic operators. Therefore, this paper proposes two versions of an epistasis-based implementation (EbFSA) as a novel approach for variable selection in multivariate calibration problems, where each version is deterministic and performs a different strategy. The use of epistasis concepts becomes important to assess the genes (variables) interdependence. Based on our experimental results, we are able to claim EbFSA can select the most informative variables and overcome some state-of-the-art algorithms. • A novel approach for variable selection in multivariate calibration problems is presented. • We present two versions of an Epistasis-based implementation (EbFSA). • A comparison with traditional algorithms is performed. • We obtained advantages by reducing the number of variable as well as the prediction error. • We experimentally demonstrated that EbFSA can be a good choice for the variable selection problem. [ABSTRACT FROM AUTHOR]

Abstract In this study, we extend the traveling repairman problem with profits (TRPP) by considering the use of multiple vehicles, which occurs commonly in practice. We call the resulting problem the multiple traveling repairman problem with profits (MTRPP). In MTRPP, a fleet of identical vehicles is dispatched to visit a subset of nodes in order to collect time-dependent profits. To solve MTRPP, we present a highly effective memetic algorithm, which combines (i) a randomized greedy construction method for initial solution generation, (ii) a variable neighborhood search for local refinement, and (iii) a dedicated route-based crossover operator for solution recombination. Computational experiments performed on a large set of instances reveal the efficiency of the algorithm. Additionally, we demonstrate the relevance of our algorithm for the classic TRPP. The reported computational results serve as a benchmark for future researchers who wish to investigate the problem. Highlights • The multiple traveling repairman problem with profits (MTRPP) is introduced. • A mathematical model of MTRPP is formulated. • A memetic algorithm (MA) is proposed for solving MTRPP. • Computational study shows that the proposed MA effectively solves MTRPP. [ABSTRACT FROM AUTHOR]

Abstract This paper presents a bi-objective hierarchical hub location problem with hub facilities as servicing centers. The objectives are to minimize the total cost (i.e., fixed cost of establishing hub facilities and transportation cost) and the maximum route length, simultaneously. Hub facilities are categorized as central and local ones. The queuing frameworks for these types of facilities are considered as M ∕ M ∕ c and M ∕ M ∕ 1 , respectively. Moreover, density functions for the traveling time and number of entities are assumed to be Exponential and Poisson functions. The presented mathematical model is solved by a new game theory variable neighborhood fuzzy invasive weed optimization (GVIWO) as introduced in this paper. To evaluate the efficiency of this proposed algorithm, some experiments are conducted and the related results are compared with the non-dominated sorting genetic algorithm (NSGA-II) and hybrid simulated annealing (HSA) algorithm with respect to some comparison metrics. The results show that the proposed GVIWO algorithm outperforms the NSGA-II and HSA. Finally, the conclusion is provided. Graphical abstract Highlights • Presenting a new bi-objective hierarchical hub location problem that minimizes the total cost and the maximum route length. • Considering the M ∕ M ∕ c and M ∕ M ∕ 1 queuing frameworks for central and local hub facilities, respectively. • Applying an efficient method based on the ME measure to cope with the uncertainty. • Proposing a new Game theory Variable neighborhood fuzzy Invasive Weed Optimization (GVIWO) method. • Comparing the GVIWO results with two well-known meta-heuristic algorithms by some comparison metrics. [ABSTRACT FROM AUTHOR]

Abstract The search for interpretable reinforcement learning policies is of high academic and industrial interest. Especially for industrial systems, domain experts are more likely to deploy autonomously learned controllers if they are understandable and convenient to evaluate. Basic algebraic equations are supposed to meet these requirements, as long as they are restricted to an adequate complexity. Here we introduce the genetic programming for reinforcement learning (GPRL) approach based on model-based batch reinforcement learning and genetic programming, which autonomously learns policy equations from pre-existing default state–action trajectory samples. GPRL is compared to a straightforward method which utilizes genetic programming for symbolic regression, yielding policies imitating an existing well-performing, but non-interpretable policy. Experiments on three reinforcement learning benchmarks, i.e., mountain car, cart–pole balancing, and industrial benchmark, demonstrate the superiority of our GPRL approach compared to the symbolic regression method. GPRL is capable of producing well-performing interpretable reinforcement learning policies from pre-existing default trajectory data. [ABSTRACT FROM AUTHOR]

Abstract Many decisions in operations management (OM) belong to the class of Non-deterministic Polynomial hard problems and thus heuristic search methods have been applied to improve OM decisions. While genetic algorithms (GAs) are promising tools for searching fast and good solutions in diverse OM areas, future research will benefit from a review of the OM problems solved by GAs. The purpose of this paper is to review the literature on OM with GA-based solutions and to suggest possible gaps from the point of view of researchers and practitioners. A total of 119 peer reviewed journal papers published from 2007 to 2017 are reviewed and analysed methodologically. The applications of GAs in OM are categorized into process and product design, operations planning and control, and operations improvement. Observations from the existing literature are presented and future research directions are suggested. Although GAs have been one of the most popular heuristic approaches for optimization, there are OM problems that are yet to be investigated. The findings of this review pave the path for future research to apply GAs to solve OM problems. Highlights • Review literature on applications of genetic algorithms in operations management. • Categorize applications in operations management into 3 themes and 9 decision areas. • Provide research and application agenda for further work. [ABSTRACT FROM AUTHOR]

Abstract Source term estimation (STE) of atmospheric dispersion plays an important role in public safety, environmental protection and many other application fields. In this paper, several new composite cost functions for STE using hybrid genetic algorithm are proposed and compared using Nemenyi test based on 68 STE tasks from Prairie Grass field experiment. Results show one of the new composite cost functions, named as WSD, has outstanding performance in estimating both source location and emission rate. Then the patterns in STE results using different cost functions are analyzed based on the 68 tasks mentioned above, which provides further insights into what to expect from STE. At last, the relationship between composite cost functions and multi-objective optimization is analyzed to facilitate the understanding of composite cost functions. To summarize, composite cost functions such as WSD has the potential to achieve a better balance between sensitivity and robustness of cost functions applied in STE, providing the most accurate estimates. Statistical algorithm comparison techniques like Nemenyi test can help us better understand the characteristics and performance of specific settings in STE methods. Graphical abstract Highlights • New composite cost functions are introduced to source term estimation. • Nemenyi test is used to compare different cost functions using multiple data sets. • One of the new composite cost function WSD has outstanding performance. • Bias patterns of the results using different cost functions are analyzed. • The relationship with multi-objective optimization is demonstrated. [ABSTRACT FROM AUTHOR]

Skin cancer is defined as the rapid growth of skin cells due to DNA damage that cannot be repaired. Melanoma is one of the deadliest types of skin cancer, which originates from melanocytes. While other skin cancer types have limited spreading capabilities, the danger of melanoma comes from its ability to spread (metastasize) rapidly. Fortunately, it can be detected by visual inspection of the skin surface, and it is 100% curable if identified in the early stages. However, detection by “subjective visual inspection” creates an important problem, due to investigators’ different levels of experiences and education. Dermoscopy (dermatoscopy) has significantly increased the diagnostic accuracy of melanoma since late 90’s. In addition, several systems have been proposed in order to assist investigators or to perform an automatic melanoma detection. This survey focuses on the algorithms for automated melanoma detection in dermoscopic images through an extensive analysis of the stages in methodologies proposed in the literature, and by examining related concepts and describing possible future directions through open problems in this domain of research. [ABSTRACT FROM AUTHOR]

Floods are natural phenomena that invade different parts of the globe annually, leaving severe adverse effects on the natural landscape, particularly in areas where humans practice large-scale urban development activities. Inadequate corporate strategic stormwater management or maintenance plans make the issue even worse. Therefore, efforts have been consolidated to confront flood hazards by incorporating precautious and proactive strategies toward better forecasting and preventing flood disasters or even mitigating the damage if it occurs. Further, there is inadequacy in incorporating computational intelligence techniques to detect flood susceptibility in road networks. This paper proposes a data-driven flood risk areas detection model incorporating various ensemble and generic machine learning techniques as well as autoML tools. A systematic and iterative approach is followed to examine different solutions to tackle the data balancing problem, including oversampling and undersampling techniques. Also, hyperparameter optimisation is undertaken using both grid search and genetic algorithms, and the models' performance using each designated strategy is reported and assessed. The empirical results from the experiments indicate that the optimised ensembled Extremely Randomized Trees classifier proves an ability to perform well on the given dataset with an averaged ROC AUC value of 90% and outperforms several state-of-the-art generic, ensembled, and autoML classifiers in all evaluation measures. The proposed framework could be possibly generalised to other road networks in similar topographical and geological regional areas. [ABSTRACT FROM AUTHOR]

The accurate and reasonable prediction of natural gas consumption is important for enhancing national energy security and alleviating the environmental problems caused by energy consumption. Starting with the traditional nonlinear grey Bernoulli model, in this paper, we expand the development coefficient a and grey action quantity b, propose a new self-adaptive time-varying grey Bernoulli prediction model, deduce the time response formula of the model, and explore the relationship between model parameters and model accuracy. Then, the nonlinear parameter and power parameter coefficient of the new model are optimized by a genetic algorithm, and the effectiveness of the model is analyzed by the natural gas consumption of China, the United States, and Russia. The results show that, compared with other traditional grey prediction models, the new model has a better simulation effect, higher prediction accuracy, and stronger applicability. Meanwhile, the model was used to forecast the natural gas consumption of China, the United States, and Russia in the period from 2022 to 2026. The prediction results show that China's natural gas consumption will continue to steadily increase in the future, which is consistent with the global goal of adhering to the orderly development of natural gas. The prediction results can provide a theoretical basis for relevant departments to formulate relevant natural gas policies. [ABSTRACT FROM AUTHOR]

The smart city is an idea of upcoming revolution in cities, which includes smart home, smart mobility, smart living, and smart environment that can be operated using a smartphone connected to the internet. By adequately scheduling appliances, residential consumers can reduce their electricity expenses while increasing their comfort. Electricity cost and user comfort being conflicting in nature, can be formulated as a dynamic multi-objective optimization problem with varying user priority to use different home appliances at different times. Further to solve this problem, a Fuzzy Adaptive Dynamic SPEA2 with Borda Ranking Method (FDSPEA2-BR) is proposed based on the dynamic modification of popular SPEA2 algorithm. The algorithm includes an improved Borda count method along with Mamdani fuzzy rules to select the best solutions. The changes in crossover and mutation rate in the SPEA2 algorithm is now being controlled with fuzzy rules. Proposed FDSPEA2-BR is validated on fourteen benchmark dynamic multi-objective functions taken from the FDA, JY, dMOP, and DF test suites, results have validated the optimization model's efficacy. Simulation study is performed on the scheduling of 11 smart home appliances varying over 10 time slots. The results in the form of time varying Pareto Fronts is demonstrated and the schedules corresponding to 5 diversified points on each Pareto Front are reported. The end user can make use of the optimized schedules that were thus obtained to modify his or her patterns of demand and energy use. [ABSTRACT FROM AUTHOR]