Swarm and Evolutionary Computation最新文献

{"title":"A constrained multimodal multi-objective evolutionary algorithm based on adaptive epsilon method and two-level environmental selection","authors":"Fengxia Wang ,&nbsp;Min Huang ,&nbsp;Shengxiang Yang ,&nbsp;Xingwei Wang","doi":"10.1016/j.swevo.2025.101845","DOIUrl":"10.1016/j.swevo.2025.101845","url":null,"abstract":"<div><div>Constrained multimodal multi-objective optimization problems (CMMOPs) commonly arise in practical problems in which multiple Pareto optimal sets (POSs) correspond to one Pareto optimal front (POF). The existence of constraints and multimodal characteristics makes it challenging to design effective algorithms that promote diversity in the decision space and convergence in the objective space. Therefore, this paper proposes a novel constrained multimodal multi-objective evolutionary algorithm, namely CM-MOEA, to address CMMOPs. In CM-MOEA, an adaptive epsilon-constrained method is designed to utilize promising infeasible solutions, promoting exploration in the search space. Then, a diversity-based offspring generation method is performed to select diverse solutions for mutation, searching for more equivalent POSs. Furthermore, the two-level environmental selection strategy that combines local and global environmental selection is developed to guarantee diversity and convergence of solutions. Finally, we design an archive update strategy that stores well-distributed excellent solutions, which more effectively approach the true POF. The proposed CM-MOEA is compared with several state-of-the-art algorithms on 17 test problems. The experimental results demonstrate that the proposed CM-MOEA has significant advantages in solving CMMOPs.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"93 ","pages":"Article 101845"},"PeriodicalIF":8.2,"publicationDate":"2025-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143098151","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A correlation-guided cooperative coevolutionary method for feature selection via interaction learning-based space division","authors":"Yaqing Hou ,&nbsp;Huiyue Sun ,&nbsp;Gonglin Yuan ,&nbsp;Yijing Li ,&nbsp;Zifan Che ,&nbsp;Hongwei Ge","doi":"10.1016/j.swevo.2025.101846","DOIUrl":"10.1016/j.swevo.2025.101846","url":null,"abstract":"<div><div>Feature selection (FS) is an important data preprocessing technique that optimizes the learning process and results by selecting a small subset of features. With the increasing of data dimension, challenges such as high computational complexity and the tendency to fall into local optima also appear. Cooperative Coevolution (CC) shows promising prospects in FS because of its “divide and conquer” approach, which decomposes a complex high-dimensional problem into several lower-dimensional subproblems to solve simultaneously. However, most existing CC-based FS methods suffer from issues such as the coarse space division, inadequate subspace exploration, singular collaboration model, and costly fitness evaluations. To alleviate these limitations, we propose a CC method via interaction learning-based space division for high-dimensional FS problems. First of all, this method proposes an interaction learning-based space division method, which divides the whole feature space into several subspaces of different importance. Then, a correlation-guided search strategy is designed to select relevant features, eliminate redundant features, and escape local optima. Finally, a surrogate-assisted particle recombination strategy also explores the combinatorial performance of non-optimal particles across different subswarms to ensure the comprehensiveness of the exploration and the efficiency of evaluations. The results on 16 typical datasets show that the proposed method evolves the feature subset with the lowest number of features and highest classification accuracy compared to seven state-of-the-art algorithms.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"93 ","pages":"Article 101846"},"PeriodicalIF":8.2,"publicationDate":"2025-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143102712","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-objective optimal power flow problem using constrained dynamic multitasking multi-objective optimization algorithm","authors":"Junhua Zhu ,&nbsp;Xiaobing Yu ,&nbsp;Feng Wang ,&nbsp;Yaqi Mao","doi":"10.1016/j.swevo.2025.101850","DOIUrl":"10.1016/j.swevo.2025.101850","url":null,"abstract":"<div><div>The multi-objective optimal power flow (MOOPF) problem involves conflicting objectives and complex constraints, presenting a significant challenge for existing optimization methods. To address constrained multi-objective optimization problems (CMOPs), a recent evolutionary multi-tasking (EMT) framework has been proposed, involving a primary task and several auxiliary tasks running in parallel. The design of these auxiliary tasks is critical for supporting the solution of the primary task. This paper introduces a novel constrained dynamic multitasking multi-objective optimization algorithm (CDMTMO) to solve CMOPs. The proposed algorithm comprises three populations, each assigned a specific task: the first population focuses on solving the primary CMOP, the second population tackles a constraint-relaxed problem, and the third population gradually transitions from solving an unconstrained problem to the CMOP. To ensure effective collaboration among auxiliary tasks and the main task, CDMTMO incorporates an improved ε-constrained method and an enhanced dual ranking method. Furthermore, a pre-selection strategy for solution sets is integrated to discern promising individuals and facilitate knowledge transfer. CDMTMO has been evaluated using two IEEE standard systems, to demonstrate its capability and suitability in efficiently tackling the MOOPF problem. A thorough analysis of CDMTMO's results was performed, comparing it with seven state-of-the-art algorithms: AGEMOEA, CCMO, DSPCMDE, CMOEMT, ToP, EMCMO and DEST. After evaluating across eight test cases, CDMTMO achieved the best inverted generational distance plus (IGD+) and hypervolume (HV) values in seven cases. Furthermore, CDMTMO achieves a feasible rate (FR) of 1 in all cases, demonstrating its consistent ability to find feasible solutions.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"93 ","pages":"Article 101850"},"PeriodicalIF":8.2,"publicationDate":"2025-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143102711","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An enhanced decomposition-based multi-objective evolutionary algorithm with neighborhood search for multi-resource constrained job shop scheduling problem","authors":"Bohan Zhang,&nbsp;Ada Che","doi":"10.1016/j.swevo.2024.101834","DOIUrl":"10.1016/j.swevo.2024.101834","url":null,"abstract":"<div><div>Today’s industry 5.0 emphasizes the synergy between humans and equipment to raise productivity. This paper investigates a multi-resource constrained job shop scheduling problem, aiming to minimize both the makespan, the total energy consumption of automated guided vehicles (AGVs), and the total workload of workers. To address the problem, we apply an extended disjunctive graph and establish a multi-objective mixed integer linear programming model based on it. Afterward, we develop an enhanced decomposition-based multi-objective evolutionary algorithm with neighborhood search (EMOEA/D-NS) to efficiently solve this problem. In this algorithm, we design a three-layer solution representation and propose a hybrid heuristic based on priority weights to yield high-quality individuals, which comprises a congestion alleviation assignment rule for AGVs and a shortest-earliest rule for allocating workers. Furthermore, three lemmas for determining non-critical tasks are given and six neighborhood search approaches are designed to improve the quality of solutions. To enhance the exploration and exploitation capabilities of the EMOEA/D-NS, we propose a multi-rank individual-driven evolutionary mechanism that classifies the individuals into guiding, working, and following groups. For the individuals within the guiding group, we propose a self-evolution strategy that allows themselves to evolve in the way of utilizing their own experiences. For the individuals of the working group, we design a collaborative evolutionary strategy, consisting of a priority weights-based crossover and mutation operators, to evolve them with other individuals to explore promising space and exploit known space. The individuals of the following group are evolved toward the direction of those within the guiding group by an oriented evolutionary strategy, which aims to improve the quality of population and accelerate the convergence of the algorithm. Numerical experiments are carried out on 40 modified benchmarks to highlight the efficiency of the EMOEA/D-NS. Lastly, we conclude our work and outline further research directions.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"93 ","pages":"Article 101834"},"PeriodicalIF":8.2,"publicationDate":"2025-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143102708","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An improved shuffled frog leaping algorithm based on support vector machine for hybrid flow shop rescheduling with disturbance prediction","authors":"Wei Zhang ,&nbsp;Shaofeng Yan ,&nbsp;Hongtao Tang ,&nbsp;Xing Li ,&nbsp;Deming Lei","doi":"10.1016/j.swevo.2025.101843","DOIUrl":"10.1016/j.swevo.2025.101843","url":null,"abstract":"<div><div>The current research on casting scheduling focuses on multi-resource constraints and the disturbances on machines, transport, and other resources. The uncertain disturbance have the potential to significantly disrupt the casting process, leading instability and delay in the whole manufacturing system. The evaluation and prediction for disturbance is a key challenge in solving the casting scheduling problem. This paper presents a hybrid flow shop rescheduling model for the casting process, which considers the maximum completion time and delay as the optimization targets under a disturbing environment. A disturbance degree evaluation system comprising five indicators in casting was established, and a classification prediction model based on a support vector machine (SVM) is developed. An improved shuffled frog leaping algorithm (ISFLA) is proposed. First, an coding and an improved NEH initialization method are designed for the single-batch coupling and resource-constrained problems. Secondly, an improved module group search strategy based on a multi-classified SVM prediction model to establish the relationship between the severity of disturbance events and the algorithmic strategy. Finally, the efficacy of the proposed casting rescheduling model under disturbance environment and the ISFLA are validated through the simulation experiments and case study.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"93 ","pages":"Article 101843"},"PeriodicalIF":8.2,"publicationDate":"2025-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143102709","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-objective orbital maneuver optimization of multi-satellite using an adaptive feedback learning NSGA-II","authors":"Qian Yin ,&nbsp;Guohua Wu ,&nbsp;Guang Sun ,&nbsp;Yi Gu","doi":"10.1016/j.swevo.2024.101835","DOIUrl":"10.1016/j.swevo.2024.101835","url":null,"abstract":"<div><div>Earth observation satellite (EOS) systems play a crucial role in performing emergency monitoring tasks such as natural disasters. In terms of urgent observation tasks within a limited period, manipulating the orbit of EOSs to meet emergency requirements is an efficient scheme. The traditional multiple satellite orbit maneuver optimization problem (MSOMOP) almost considers single objective optimization, neglecting the optimization of conflicting objectives in practical applications. This paper is devoted to conducting multi-objective optimization research for the MSOMOP. First, a multi-objective mathematical model is established, where the response time, imaging resolution, and fuel cost are considered as optimization objectives. Subsequently, an adaptive feedback learning of non-dominated sorting genetic algorithm-II (AFL-NSGA-II) is proposed, which introduces the idea of adaptive strategy and a feedback learning mechanism into the traditional NSGA-II. The AFL-NSGA-II incorporates an increased learning mechanism and adaptive strategies, which facilitates efficient solution search and reduces the risk of converging to a local optimum. Moreover, several problem-specific designed operators are incorporated into the algorithm to enhance the search capability. Finally, we conduct extensive experimental studies to verify the efficiency of the proposed algorithm. Experiment results demonstrate that the proposed AFL-NSGA-II outperforms three existing algorithms and exhibits superior performance in typical scheduling scenarios.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"93 ","pages":"Article 101835"},"PeriodicalIF":8.2,"publicationDate":"2025-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143164101","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A reinforcement learning-based ranking teaching-learning-based optimization algorithm for parameters estimation of photovoltaic models","authors":"Haoyu Wang ,&nbsp;Xiaobing Yu ,&nbsp;Yangchen Lu","doi":"10.1016/j.swevo.2025.101844","DOIUrl":"10.1016/j.swevo.2025.101844","url":null,"abstract":"<div><div>Accurate parameter control and optimization are vital issues in the process of utilizing solar energy through photovoltaic systems, which poses significant challenges due to the inherent complexity of photovoltaic systems. The paper proposes a new algorithm, reinforcement learning-based ranking teaching-learning-based optimization, for identifying photovoltaic model parameters. Parameters play a major role in the performance of many optimization algorithms, and the same parameter is not appropriate for all problems. Reinforcement learning adjusts parameters by accumulating returns to meet the requirements of the environmental model. The proposed algorithm divides learners into superior and inferior groups based on fitness rankings and uses a reinforcement learning approach to dynamically adjust learner classification divisions to ensure adaptability in different optimization scenarios. In the teacher phase, superior learners emulate top performers, while inferior learners engage in guided mutual learning to enhance global search capabilities. In the learner phase, superior learners communicate with better peers, while inferior learners seek a wider range of information sources, balancing exploration and exploitation. In the experimental evaluation of five different photovoltaic models, the comparative analysis of eleven established algorithms verified its superior performance in accuracy, convergence speed, and complexity.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"93 ","pages":"Article 101844"},"PeriodicalIF":8.2,"publicationDate":"2025-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143102710","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-modal multi-objective wolf pack algorithm with circumferential scouting and intra-niche interactions","authors":"Jia Zhao ,&nbsp;Fujun Chen ,&nbsp;Renbin Xiao ,&nbsp;Runxiu Wu ,&nbsp;Jeng-Shyang Pan ,&nbsp;Hui Wang ,&nbsp;Ivan Lee","doi":"10.1016/j.swevo.2024.101842","DOIUrl":"10.1016/j.swevo.2024.101842","url":null,"abstract":"<div><div>The multi-objective wolf pack algorithm faces issues in solving multi-modal multi-objective optimization problems, such as the dominance of high-performing individuals, excessive reliance on the lead wolf, and ineffective learning among certain individuals, which result in poor diversity and convergence. Therefore, this paper presents a multi-modal multi-objective wolf pack algorithm with circumferential scouting and intra-niche interactions (MMOWPA-CN). To enhance algorithm's local search capability, a circumferential scouting technique is proposed. It divides the subpopulation using two different non-dominated levels of individuals, allowing individuals within the subpopulation to search in all directions, thereby improving the exploration capability. To prevent population from generating aggregation phenomenon, the neighbourhood elite bootstrapping strategy is introduced. It utilizes the neighbourhood lead wolf and the historical optimal solutions stored in archive to constrain individual movements, guiding them toward sparser areas and ensuring decision space's diversity. Additionally, in the decision space, a mechanism of intra-niche interactions is designed to avoid ineffective learning among individuals on different fronts. It allows individuals to interact with each other with information in a restricted area, ensuring the convergence of the algorithm. Comparative experiments on 17 test functions and one practical application have shown that MMOWPA-CN outperforms seven state-of-the-art algorithms, demonstrating its stronger optimization capabilities.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"93 ","pages":"Article 101842"},"PeriodicalIF":8.2,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143102750","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Meta-Black-Box optimization for evolutionary algorithms: Review and perspective","authors":"Xu Yang ,&nbsp;Rui Wang ,&nbsp;Kaiwen Li ,&nbsp;Hisao Ishibuchi","doi":"10.1016/j.swevo.2024.101838","DOIUrl":"10.1016/j.swevo.2024.101838","url":null,"abstract":"<div><div>Black-Box Optimization (BBO) is increasingly vital for addressing complex real-world optimization challenges, where traditional methods fall short due to their reliance on expert knowledge and time-consuming processes. Meta-Black-Box Optimization (MetaBBO) emerges as a pivotal solution, leveraging meta-learning to enhance or discover optimization algorithms automatically. Originating from Automatic Algorithm Design (AAD), MetaBBO has branched into areas such as Learn to Optimize (L2O), Automated Design of Meta-heuristic Algorithm (ADMA), and Automatic Evolutionary Computation (AEC), each contributing to the advancement of the field. This comprehensive survey integrates and synthesizes the extant research within MetaBBO for Evolutionary Algorithms (EAs) to develop a consistent community of this research topic. Specifically, a mathematical model for MetaBBO is established, and its boundaries and scope are clarified. The potential optimization objects in MetaBBO for EAs is explored, providing insights into design space. A taxonomy of MetaBBO methodologies is introduced, reflecting the state-of-the-art from a meta-level perspective. Additionally, a comprehensive overview of benchmarks, evaluation metrics, and platforms is presented, streamlining the research process for those engaged in learning and experimentation in MetaBBO for EA. The survey concludes with an outlook on research, underscoring future directions and the pivotal role of MetaBBO in automatic algorithm design and optimization problem-solving.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"93 ","pages":"Article 101838"},"PeriodicalIF":8.2,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143102753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A consensus optimization mechanism with Q-learning-based distributed PSO for large-scale group decision-making","authors":"Qingyang Jia ,&nbsp;Kewei Yang ,&nbsp;Yajie Dou ,&nbsp;Ziyi Chen ,&nbsp;Nan Xiang ,&nbsp;Lining Xing","doi":"10.1016/j.swevo.2024.101841","DOIUrl":"10.1016/j.swevo.2024.101841","url":null,"abstract":"<div><div>Current industrial product design evaluation faces multiple challenges including shortened research and development (R&amp;D) cycles, increased technical complexity and expanding expert teams, which exacerbate problems of incomplete information, uncertainty, and expert preference conflicts. Existing evaluation methods are difficult to capture linguistic ambiguity effectively and exhibit low consensus efficiency and optimization performance in large-scale group decision making (LSGDM). To address these challenges, this paper proposes a Q-learning-based distributed particle swarm optimization (QLDPSO) consensus mechanism for industrial product design evaluation. The proposed approach utilizes probabilistic linguistic term sets (PLTSs) to express expert preferences and capture evaluation uncertainties. An automated consensus optimization model is developed to eliminate preference conflicts, improve consensus efficiency and minimize time and effort spent on repeated negotiations by identifying optimization objectives and adjustment ranges. To overcome slow convergence and local optima issues in high-dimensional optimization, the method integrates Q-learning with distributed PSO, dividing the population into collaboratively evolving subpopulations and dynamically adjusting subpopulation sizes through reinforcement learning to balance exploration and exploitation. Finally, the proposed algorithm was validated through an aeroengine design case study and compared with existing algorithms. The experimental results demonstrate that the QLDPSO consensus optimization mechanism significantly improves both the consensus optimization efficiency and evaluation accuracy in LSGDM scenarios, offering an innovative and practical solution for design alternative selection of complex industrial products.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"93 ","pages":"Article 101841"},"PeriodicalIF":8.2,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143102752","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}