{"title":"A knowledge-guided evolutionary algorithm incorporating reinforcement learning for energy efficient dynamic flexible job shop scheduling problem with machine breakdowns","authors":"Zhixiao Li , Guohui Zhang , Nana Yu , Shenghui Guo , Wenqiang Zhang","doi":"10.1016/j.swevo.2025.102050","DOIUrl":"10.1016/j.swevo.2025.102050","url":null,"abstract":"<div><div>The flexible job shop scheduling problem is gradually developing towards greening and intelligence. However, in the real production, there are often various dynamic disturbances that result in lower executability of scheduling solutions. Therefore, this paper first investigates the energy efficient dynamic flexible job shop scheduling problem with machine breakdowns. To solve this problem, a knowledge-guided evolutionary algorithm incorporating reinforcement learning (KEARL) is established to minimize maximum completion time, total energy consumption, and workload of critical machines, which is a mixed-integer linear programming model with transportation time of jobs and setup time of machines included. In KEARL, a new rescheduling strategy is designed to reduce the possibility of the machine's second breakdown. In addition, four knowledge-guided initialization methods are also designed and a reinforcement learning-based parameter adaptive strategy is used to optimize the crossover probability and mutation probability, while a knowledge-guided variable neighborhood search strategy enhances the search capability of KEARL. More importantly, three energy efficient methods are implemented to reduce the energy consumption of the production process. Finally, through extensive experiments, the KEARL is compared with several well-known algorithms. The experimental results indicate that KEARL outperforms the other algorithms.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"97 ","pages":"Article 102050"},"PeriodicalIF":8.2,"publicationDate":"2025-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144489587","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 augmented variable neighborhood search for mixed-model two-sided assembly line balancing considering PM scenarios","authors":"Lianpeng Zhao , Qiuhua Tang","doi":"10.1016/j.swevo.2025.102043","DOIUrl":"10.1016/j.swevo.2025.102043","url":null,"abstract":"<div><div>In a real mixed-model two-sided assembly line, preventive maintenance (PM) activities cause capacity waste at available stations and production halts. To mitigate these issues, multiple task assignment schemes with high interchangeability are required, each tailored to one specific scenario. However, the resulting mixed-model two-sided assembly line balancing problem considering PM scenarios (MTALBP-PM) has not been studied. Therefore, a mixed-integer linear programming model is formulated to minimize total cycle time and task adjustment simultaneously. Meanwhile, driven by knowledge and learning, an augmented variable neighborhood search (AVNS) is designed. Concretely, with the guidance of problem-specific knowledge, a decoding mechanism and three objective-oriented neighborhood structures are designed to achieve solutions with better objectives. Using unsupervised learning, an initialization heuristic is mined from tacit information to obtain high-quality initial solutions. With historical search information, a self-adaptive strategy based on Q-learning is proposed to recommend the best-fit neighborhood structure for higher efficiency. Besides, an auto-tuning restart operator based on multi-domain knowledge is employed to escape local optima. Experimental results show that the espoused policy is effective, and AVNS outperforms eight other state-of-the-art meta-heuristics in deriving well-converged and -distributed Pareto fronts. In a statistical sense, the average <em>GD, IGD</em>, and <em>HVR</em> of AVNS reach the best values among all tested meta-heuristics based on 40 benchmark cases, which are 0.4599, 0.8021, and 0.8943, respectively.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"97 ","pages":"Article 102043"},"PeriodicalIF":8.2,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144489590","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}
Yulan Lu , Xueyi Guo , Jinggai Geng , Shuai Wang , Zhenkun Wang , Aimin Zhou , Jianyong Sun , Xinhui Si , Xin Sun , Hu Zhang
{"title":"Regularity-based evolutionary multi-objective optimization review","authors":"Yulan Lu , Xueyi Guo , Jinggai Geng , Shuai Wang , Zhenkun Wang , Aimin Zhou , Jianyong Sun , Xinhui Si , Xin Sun , Hu Zhang","doi":"10.1016/j.swevo.2025.101999","DOIUrl":"10.1016/j.swevo.2025.101999","url":null,"abstract":"<div><div>Under mild conditions, the Pareto optimal solutions of a continuous <span><math><mi>m</mi></math></span>-dimensional multi-objective optimization problem (MOP) have been proved to form a piecewise (<span><math><mi>m</mi></math></span>-1)-dimensional manifold structure in the search space, a characteristic known as the regularity property. As a domain knowledge of MOP, since the first proposal in 2008, this regularity property has demonstrated significant potential for enhancing the performance of multiobjective evolutionary algorithms (MOEAs). However, there has yet to be a systematic survey of the regularity property within the design of MOEAs. This article aims to address this gap by providing a comprehensive review of regularity-based evolutionary multi-objective optimization (REMO) approaches. We hope that this survey will help EMO researchers to have a comprehensive understanding of REMO.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"97 ","pages":"Article 101999"},"PeriodicalIF":8.2,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144471282","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}
Wei Liang , Zeqiang Zhang , Yanqing Zeng , Dan Ji , Yu Zhang , Haiye Chen , Yan Li , Lixia Zhu
{"title":"Modelling and optimization of mixed-parallel straight and two-sided disassembly line balancing problem","authors":"Wei Liang , Zeqiang Zhang , Yanqing Zeng , Dan Ji , Yu Zhang , Haiye Chen , Yan Li , Lixia Zhu","doi":"10.1016/j.swevo.2025.102045","DOIUrl":"10.1016/j.swevo.2025.102045","url":null,"abstract":"<div><div>The recycling of end-of-life (EoL) products is an urgent challenge at present. Disassembly line layout plays a crucial role among the factors that affect recycling efficiency. Subsequently, this study proposed a mixed-parallel straight and two-sided disassembly line layout, combining the advantages of the three layouts to enhance the efficiency of recycling EoL products. Additionally, a mixed-integer non-linear programming (MINLP) model was developed to minimize the number of workstations, idle time balancing, demand, and hazard indices. To solve the mixed-parallel straight and two-sided disassembly line balancing problem (k-MPSTDLBP), this study designed a two-layer non-dominated sorting genetic algorithm-II (NSGA-Ⅱ) with universal encoding and decoding mechanisms. The algorithm’s effectiveness was validated by solving two hybrid cases using both the MINLP model and the two-layer NSGA-II. Moreover, comparative analyses with the non-dominated sorting genetic algorithm-III, the improved artificial fish swarm algorithm, and the improved firefly algorithm demonstrated the superiority of the two-layer NSGA-II. Finally, the two-layer NSGA-II was applied to a hybrid case study involving four types of EoL products under both the mixed-parallel straight and two-sided disassembly line and straight disassembly line layouts, confirming the higher recycling efficiency of the k-MPSTDLBP. Meanwhile, the sensitivity of the two-layer NSGA-II on the k-MPSTDLBP was analyzed using orthogonal experiments.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"97 ","pages":"Article 102045"},"PeriodicalIF":8.2,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144471283","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":"Teaching–learning based optimization of a multi-server machining model with emergency vacation and retention of reneged machines","authors":"C.K. Anjali, Sreekanth Kolledath","doi":"10.1016/j.swevo.2025.102036","DOIUrl":"10.1016/j.swevo.2025.102036","url":null,"abstract":"<div><div>Emergency vacations refers to the immediate and unplanned leave of absence that repairmen take in response to unforeseen and urgent crises, such as natural calamities, health emergencies, or significant disruptions like the COVID-19 pandemic. During these vacations, the repair facility halts the service of failed units waiting before completion. This study focuses on developing a multi-server machining model with emergency vacation. The system comprises of K operating machines, S standbys and R repairmen. It also resolves the challenge of failed units reneging while waiting for service by implementing retention strategies. The steady state evaluation of the model is conducted utilizing the matrix analytic method, and various performance metrics are obtained. A graphical analysis of the obtained metrics is conducted to identify factors that can optimize these measures. The cost optimization of the model is implemented through TLBO and the results obtained through it is compared using PSO and GA techniques.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"97 ","pages":"Article 102036"},"PeriodicalIF":8.2,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144471284","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}
Wei Zhang, Jianchang Liu, Jiaxin Tian, Yuanchao Liu, Honghai Wang
{"title":"A dual ensemble-surrogate assisted evolutionary algorithm with state information driven infill criterion for expensive many-objective optimization","authors":"Wei Zhang, Jianchang Liu, Jiaxin Tian, Yuanchao Liu, Honghai Wang","doi":"10.1016/j.swevo.2025.102019","DOIUrl":"10.1016/j.swevo.2025.102019","url":null,"abstract":"<div><div>Surrogate-assisted evolutionary algorithms (SAEAs) for expensive many-objective optimization have received increasing attention in recent years. However, most of them do not fully exploit the useful information from the real objective evaluations to guide the whole search process of SAEAs, while ignoring the importance of the population state to select individuals for updating the model. These make most of them have the low search efficiency and behave unsatisfactorily. For this purpose, this work develops a dual ensemble-surrogate assisted evolutionary algorithm with state information driven infill criterion (SIDSAEA) for expensive many-objective optimization. In SIDSAEA, two ensemble-surrogate models are built based on the same really evaluated training set for fully exploiting the information in the real objective evaluations, where one is to approximate the objective functions for optimizing and the other is to approximate the proposed comprehensive performance indicator for infilling. Based on the model for objective functions, an inner–outer region based dominance relation is designed to select a promising population, and further improve the algorithm search efficiency. In addition, a state information driven infill criterion is proposed to select the optimal individuals for updating the model. According to the captured state, this criterion adaptively switches between the expected improvement based on the above performance indicator and the uncertainty of approximating objectives, and thus better balances convergence and diversity of individuals and improves the algorithm exploration. Extensive experimental results on three benchmark test suites and two real-world applications demonstrate that SIDSAEA has higher competitiveness in comparison with six state-of-the-art SAEAs on 65 out of 89 test instances.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"97 ","pages":"Article 102019"},"PeriodicalIF":8.2,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144365634","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}
Peng Wang , Yue Chen , Changsheng Zhang , Xiangrong Tong , Yingjie Wang
{"title":"A constrained multi-objective evolutionary algorithm via separate exploration and united exploitation strategy","authors":"Peng Wang , Yue Chen , Changsheng Zhang , Xiangrong Tong , Yingjie Wang","doi":"10.1016/j.swevo.2025.102044","DOIUrl":"10.1016/j.swevo.2025.102044","url":null,"abstract":"<div><div>Efficient allocation of search resources is paramount in solving constrained multi-objective optimization problems (CMOPs). This task becomes particularly challenging when striving to strike a balance among diversity, convergence, and feasibility, especially in CMOPs with intricate infeasible regions. To address this issue, we present an algorithm with two complementary search stages for efficient dynamic resource allocation on diversity, convergence, and feasibility. Firstly, the separate exploration stage independently explores the unconstrained and constrained Pareto fronts, efficiently traversing complex infeasible zones. Subsequently, in the united exploitation stage, the searches collaboratively exploit the constrained Pareto front. Furthermore, a <span><math><mi>θ</mi></math></span>-constraint dominance principle-based environmental selection is incorporated to achieve a balance between constraint convergence and diversity. Comprehensive tests on 47 problems across four benchmark suites and 6 real-world CMOPs reveal that the proposed algorithm outperforms six state-of-the-art algorithms, demonstrating its superior efficacy.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"97 ","pages":"Article 102044"},"PeriodicalIF":8.2,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144365633","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}
Wenji Li , Yifeng Qiu , Zhaojun Wang , Biao Xu , Zhifeng Hao , Qingfu Zhang , Yun Li , Zhun Fan
{"title":"Surrogate-assisted neural learning and evolutionary optimization for expensive constrained multi-objective problems","authors":"Wenji Li , Yifeng Qiu , Zhaojun Wang , Biao Xu , Zhifeng Hao , Qingfu Zhang , Yun Li , Zhun Fan","doi":"10.1016/j.swevo.2025.102020","DOIUrl":"10.1016/j.swevo.2025.102020","url":null,"abstract":"<div><div>Expensive constrained multi-objective optimization problems (ECMOPs) present significant challenges due to the high computational cost of evaluating objective and constraint functions, which severely limits the number of feasible function evaluations. To address this issue, we propose an efficient surrogate-assisted constrained multi-objective evolutionary algorithm, named LEMO. LEMO integrates neural learning with a novel constraint screening strategy to dynamically construct surrogate models for the most relevant constraints. During the optimization process, a neural network is designed to learn the mapping between arbitrary weight vectors and their corresponding constrained Pareto optimal solutions. This enables the generation of high-quality solutions while requiring fewer expensive function evaluations. Additionally, a constraint screening mechanism is introduced to dynamically exclude constraints that are irrelevant to the current search phase, thus simplifying the surrogate models and improving the efficiency of the constrained search process. To evaluate the effectiveness of LEMO, we compare its performance against seven state-of-the-art algorithms on three benchmark suites, LIRCMOP, DASCMOP, and MW, as well as a real-world optimization problem. The experimental results demonstrate that LEMO consistently outperforms these algorithms in both computational efficiency and solution quality.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"97 ","pages":"Article 102020"},"PeriodicalIF":8.2,"publicationDate":"2025-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144365083","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":"Constrained multi-objective optimization assisted by competitive mechanism based reward auxiliary optimization problems","authors":"Haoming Zhang , Qianlong Dang , Xinyu Feng , Xiaochuan Gao","doi":"10.1016/j.swevo.2025.102021","DOIUrl":"10.1016/j.swevo.2025.102021","url":null,"abstract":"<div><div>Solving constrained multi-objective optimization problems requires simultaneously satisfying multiple objectives and constraints, presenting a significant challenge for solving tasks. Constructing auxiliary optimization problems to assist the main optimization problem in accelerating convergence is a common approach for constrained multi-objective evolutionary algorithms (CMOEAs). However, this approach may waste computational resources on auxiliary optimization problems that provide little benefit to the main optimization problem at certain stages of the evolution process. Based on the above issue, this paper proposes a constrained multi-objective optimization algorithm to address this issue via competitive mechanism based reward auxiliary optimization problems (RACMO). Specifically, an unconstrained auxiliary optimization problem and a dynamic constrained auxiliary optimization problem are constructed. They are rewarded by the number of solutions provided to the main optimization problem, and the cumulative reward is mapped to the probability to adaptively selecting more valuable auxiliary optimization problems. Moreover, an adaptive stop-update strategy is designed. By controlling the competition between two auxiliary populations and adaptive stop-updating, excellent convergence is guaranteed while significantly saving computational resources. Experimental results demonstrate the competitiveness of RACMO compared to ten advanced CMOEAs on three test suites and eight practical application problems.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"97 ","pages":"Article 102021"},"PeriodicalIF":8.2,"publicationDate":"2025-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144365084","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}
Huixiang Zhen , Bing Xue , Wenyin Gong , Mengjie Zhang , Ling Wang
{"title":"Offline evolutionary optimization with problem-driven model pool design and weighted model selection indicator","authors":"Huixiang Zhen , Bing Xue , Wenyin Gong , Mengjie Zhang , Ling Wang","doi":"10.1016/j.swevo.2025.102034","DOIUrl":"10.1016/j.swevo.2025.102034","url":null,"abstract":"<div><div>Offline data-driven evolutionary algorithms aim to provide a promising solution based on the collected historical data, without online real fitness evaluations. However, the suitability of surrogate models varies significantly across different problem types, and current research often overlooks the relationship between problem characteristics and model performance. To address this gap, we propose a novel offline data-driven evolutionary algorithm, termed MSEA, which integrates a problem-driven model pool design and a weighted indicator-based model selection mechanism. The model pool is carefully designed, incorporating four distinct surrogate models tailored for various optimization landscapes to align with diverse problem characteristics. A weighted selection indicator, derived from both model evaluation and solution quality assessment, is employed to dynamically select the most suitable model for the optimization problem. Extensive experimental results demonstrate that MSEA effectively identifies and utilizes the optimal model from the pool for specific offline optimization tasks. Compared to five state-of-the-art offline data-driven methods, MSEA achieved optimal results for 26 out of 32 functions across dimensions ranging from 10 to 100 and also exhibited faster running times. Furthermore, in high-dimensional spaces, MSEA achieved the best optimization results in dimensions ranging from 200 to 500. Our code is available at <span><span>https://github.com/zhenhuixiang/MSEA</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"97 ","pages":"Article 102034"},"PeriodicalIF":8.2,"publicationDate":"2025-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144364917","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}