Swarm and Evolutionary Computation最新文献

{"title":"Combined model assisted evolutionary algorithm with complementary fill sampling criterion for expensive multi/many-objective optimization","authors":"Qiuzhen Wang ,&nbsp;Feng Xie ,&nbsp;Yuan Liu ,&nbsp;Juan Zou ,&nbsp;Jinhua Zheng","doi":"10.1016/j.swevo.2025.101980","DOIUrl":"10.1016/j.swevo.2025.101980","url":null,"abstract":"<div><div>In this paper, we design an algorithm to address the challenges of expensive multi-objective optimization problems by improving the surrogate model and sampling criterion. Firstly, we introduce a combined model which aims to enhance the impact of points that do not play a negative role, thus improving prediction accuracy. Subsequently, we develop two complementary indicators to accommodate various shapes of Pareto frontiers to better balance convergence and diversity in the sampling criterion. Experimental results on several benchmarks show that our proposed method is highly competitive in solving expensive multi-objective optimization problems compared to other state-of-the-art algorithms.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"96 ","pages":"Article 101980"},"PeriodicalIF":8.2,"publicationDate":"2025-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144139624","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 Q-learning-assisted memetic algorithm for joint vehicle scheduling problem for harvesting and transportation in smart agriculture","authors":"Xiang Guo ,&nbsp;Quan-Ke Pan ,&nbsp;Wei Zhang ,&nbsp;Zhong-Hua Miao ,&nbsp;Xue-Lei Jing ,&nbsp;Hong-Yan Sang","doi":"10.1016/j.swevo.2025.102007","DOIUrl":"10.1016/j.swevo.2025.102007","url":null,"abstract":"<div><div>As smart agriculture continues to advance, the integration of agricultural activities with intelligent vehicle technologies is offering significant opportunities while posing new challenges. This paper focuses on the harvesting and transportation joint vehicle scheduling problem (HTJVSP) in smart agriculture, aiming to minimize the maximum completion time. The study proposes a joint vehicle scheduling model and introduces a novel solution approach based on a Q-learning-assisted memetic algorithm (Q-MA). The Q-MA algorithm features a hybrid initialization strategy that generates a diverse and high-quality initial population. During the evolutionary phase, three tailored crossover strategies are proposed, specifically designed to align with the unique characteristics of HTJVSP. These strategies enhance the exploration of the search space and promote faster convergence. In the local search phase, Q-learning acts as an adaptive decision- making agent, dynamically selecting the most effective operator from four specialized local search methods, thereby improving solution refinement and accelerating convergence. Finally, the experimental results and ANOVA analysis confirm that the Q-MA outperforms state-of-the-art competitors from the benchmark set, demonstrating the effectiveness of the proposed algorithmic components and its superior performance in solving the HTJVSP.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"96 ","pages":"Article 102007"},"PeriodicalIF":8.2,"publicationDate":"2025-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144146780","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 dynamic multi-objective evolutionary algorithm based on geometric prediction and vector–scalar transformation strategy","authors":"Yan Zhao,&nbsp;Yongjie Ma,&nbsp;Yue Yang","doi":"10.1016/j.swevo.2025.101987","DOIUrl":"10.1016/j.swevo.2025.101987","url":null,"abstract":"<div><div>Dynamic multi-objective evolutionary algorithms (DMOEAs) have attracted significant attention from scholars due to their strong robustness and wide range of applications across various fields. A current research focus is on how to quickly track the changing Pareto Set (PS) and Pareto Front (PF); however, the distribution of optimal individuals on the PF is often overlooked. To address this issue, we propose a dynamic multi-objective evolutionary algorithm based on geometric prediction and vector–scalar transformation strategy (GPVS). By combining memory and diversity strategies, we propose a zoom-in and zoom-out prediction strategy for population range estimation based on a geometric center point. The mirror adjustment strategy is introduced as a prediction adjustment mechanism to accelerate the algorithm’s convergence. The vector–scalar transformation strategy optimizes the distribution of the evolved population following geometric prediction, ensuring that individuals carry the maximum possible evolutionary information. This strategy provides valuable population information for the next evolution. We evaluated the performance of the proposed algorithm through experimental comparisons with classical algorithms on 22 test functions, demonstrating its effectiveness and robustness in solving dynamic multi-objective optimization problems (DMOPs).</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"96 ","pages":"Article 101987"},"PeriodicalIF":8.2,"publicationDate":"2025-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144130919","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":"Solving machine overload for re-scheduling of dynamic flexible job shop by adaptive tripartite game theory-based genetic algorithm","authors":"Zeyu Feng,&nbsp;Zhiyuan Zou,&nbsp;Xu Liang","doi":"10.1016/j.swevo.2025.101938","DOIUrl":"10.1016/j.swevo.2025.101938","url":null,"abstract":"<div><div>In the production process of a flexible job shop, the dynamic events could disrupt the original production scheduling plan. The existing methods typically use rescheduling, but they only ensure the resumption of normal production without considering the machine load, which would lead to a machine overload vicious cycle. This paper studies the dynamic flexible job shop scheduling problem (DFJSP) considering machine load under the constraint of machine breakdown as a dynamic event, and proposes an adaptive tripartite game theory-based genetic algorithm (ATGA). Firstly, a population initialization strategy based on a pre-scheduling scheme is designed to obtain a better initial population. Then, in order to better balance multiple objectives, a machine selection strategy based on tripartite game is designed. Finally, for improving the search ability and convergence performance of the algorithm, the adaptive probability selection strategy of binary tournament is designed. The experimental results show that the algorithm surpasses other advanced algorithms in scheduling effectiveness.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"96 ","pages":"Article 101938"},"PeriodicalIF":8.2,"publicationDate":"2025-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144131020","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 systematic review of metaheuristics-based and machine learning-driven intrusion detection systems in IoT","authors":"Mohammad Shamim Ahsan ,&nbsp;Salekul Islam ,&nbsp;Swakkhar Shatabda","doi":"10.1016/j.swevo.2025.101984","DOIUrl":"10.1016/j.swevo.2025.101984","url":null,"abstract":"<div><div>The widespread adoption of the Internet of Things (IoT) has raised a new challenge for developers since it is prone to known and unknown cyberattacks due to its heterogeneity, flexibility, and close connectivity. To defend against such security breaches, researchers have focused on building sophisticated intrusion detection systems (IDSs) using machine learning (ML) techniques. Although these algorithms notably improve detection performance, they require excessive computing power and resources, which are crucial issues in IoT networks considering the recent trends of decentralized data processing and computing systems. Consequently, many optimization techniques have been incorporated with these ML models. Specifically, a special category of optimizer adopted from the behavior of living creatures and different aspects of natural phenomena, known as metaheuristic algorithms, has been a central focus in recent years and brought about remarkable results. Considering this vital significance, we present a comprehensive and systematic review of various applications of metaheuristics algorithms in developing a machine learning-based IDS, especially for IoT. A significant contribution of this study is the discovery of hidden correlations between these optimization techniques and machine learning models integrated with state-of-the-art IoT-IDSs. In addition, the effectiveness of these metaheuristic algorithms in different applications, such as feature selection, parameter or hyperparameter tuning, and hybrid usages are separately analyzed. Moreover, a taxonomy of existing IoT-IDSs is proposed. Furthermore, we investigate several critical issues related to such integration. Our extensive exploration ends with a discussion of promising optimization algorithms and technologies that can enhance the efficiency of IoT-IDSs.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"96 ","pages":"Article 101984"},"PeriodicalIF":8.2,"publicationDate":"2025-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144130918","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":"Remanufacturing scheduling under uncertainty considering remanufacturability assessment with adaptive hybrid optimization algorithm","authors":"Qinyu Jin ,&nbsp;Jifeng Xu ,&nbsp;Xiaoling Huang ,&nbsp;Xiangqi Liu ,&nbsp;Liang Huang ,&nbsp;Kang Jiang","doi":"10.1016/j.swevo.2025.101990","DOIUrl":"10.1016/j.swevo.2025.101990","url":null,"abstract":"<div><div>Remanufacturing enables the values contained in end-of-life products to be developed and utilized to the maximum extent, which is greatly significant to economic and social development. The remanufacturing process is characterized by uncertainties such as the quality of end-of-life products and the required remanufacturing time. Some studies have focused on remanufacturing scheduling under uncertainty. However, these studies ignored the direct effects of uncertainties on the assessment of remanufacturability and the selection of remanufacturing lines. Therefore, this study proposed a new decision tree-based remanufacturing scheduling model under uncertainty considering remanufacturability assessment, which constructs decision trees and combines fuzzy numbers to assess remanufacturability and select appropriate remanufacturing lines. Experiments have shown that the proposed model increases the total profits by approximately 2.8 %. To solve this model effectively, an adaptive hybrid optimization algorithm is proposed, with a new solution representation scheme, an adaptive adjustment function and a new population updating strategy. Simulated comparison experiments with other baseline algorithms and a real case study demonstrate that, the proposed algorithm has better performance in solution exploration and has superior stability in solving the remanufacturing scheduling model proposed in this study. Specifically, for improving the efficiency of remanufacturing, the proposed algorithm performs 0.5 % better than the differential evolutionary algorithm, 3.3 % better than the teaching-learning-based optimization algorithm, 0.2 % better than the extended particle swarm optimization algorithm, 1.7 % better than the improved ant colony optimization algorithm, and 2.7 % better than the simulated annealing algorithm, approximately. Finally, a real case study demonstrates the superior performance of the proposed model and algorithm in real industrial applications.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"96 ","pages":"Article 101990"},"PeriodicalIF":8.2,"publicationDate":"2025-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144124513","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":"Solving spherical multi-aircraft path planning problem using self-adaptive update strategy differential evolution algorithm","authors":"Lun Zhu ,&nbsp;Yongquan Zhou ,&nbsp;Guo Zhou ,&nbsp;Qifang Luo ,&nbsp;Yuanfei Wei","doi":"10.1016/j.swevo.2025.102004","DOIUrl":"10.1016/j.swevo.2025.102004","url":null,"abstract":"<div><div>The path planning problem is a class of NP-hard problems, and in recent years, researchers have endeavored to address path planning issues in various environments. This paper addresses, for the first time, a spherical multi-aircraft path planning problem (SMAPPP), which is proposed to simulate future research on path planning for spacecraft exploring extraterrestrial planets or missile trajectories. The differential evolution (DE) algorithm, a classic meta-heuristic algorithm, does not yield satisfactory results when applied to the SMAPPP. Constrained by the imbalance between the global and local search capabilities of DE, this paper proposes a self-adaptive update strategy differential evolution algorithm (SaUSDE) to enhance its performance in solving global optimization problems. The SaUSDE algorithm incorporates four distinct DE variant strategies. Among these strategies, the algorithm adaptively and frequently selects the more suitable strategy for different problems, thereby effectively balancing global and local search capabilities. To validate the algorithm's efficacy, tests were conducted on the CEC2017 benchmark functions at 10D, 30D, 50D, and 100D, as well as 48 real-world constrained optimization problems from CEC2020. Based on these problems, our algorithm was compared with several advanced DE variants and a few algorithms that won the CEC2020 competition. Compared with these advanced algorithms, the proposed algorithm generally achieved superior results. Finally, SaUSDE was applied to a simulated spherical multi-aircraft path planning problem. In scenarios involving 3 to 12 aircraft, the SaUSDE algorithm effectively avoided collisions between each aircraft and certain no-fly zones (static obstacles) as well as unknown flying objects (dynamic obstacles). The source code for the algorithm is available at: <span><span>https://ww2.mathworks.cn/matlabcentral/fileexchange/173495-sausde</span><svg><path></path></svg></span></div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"96 ","pages":"Article 102004"},"PeriodicalIF":8.2,"publicationDate":"2025-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144124514","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":"KDDCP: Kernel density driven cumulative probability-cooperated framework for constrained muti-objective optimization","authors":"Shijie Zhao ,&nbsp;Xin Yu ,&nbsp;Lei Zhang ,&nbsp;Jinling Song","doi":"10.1016/j.swevo.2025.101989","DOIUrl":"10.1016/j.swevo.2025.101989","url":null,"abstract":"<div><div>In constrained multi-objective problems (CMOPs), striking a balance between population diversity, convergence, and feasibility requirements is a critical challenge for constrained multi-objective evolutionary algorithms (CMOEAs). Overemphasizing feasibility in traditional methods may lead to the search being trapped in local optima, whereas ignoring constraints may result in resource waste. To address these challenges, in this paper, a constrained multi-objective evolutionary algorithm (KDDCP) based on kernel density-driven cooperative cumulative probability is proposed, which can reasonably utilize feasible solutions as well as valuable infeasible solutions in order to approach the real Pareto frontier effectively. Specifically, KDDCP achieves multi-objective search cooperation by constructing the main population, the global auxiliary population and the local auxiliary population: the main population focuses on finding the feasible Pareto front (CPF). Meanwhile, the global auxiliary population approaches the unconstrained Pareto frontier (UPF) under the condition of ignoring constraints, and only runs in the first stage of the algorithm, and stops evolving after convergence, thus avoiding the computational overhead in the subsequent stage. For the local auxiliary population, KDD-RDS and CP-RDB strategies are combined to facilitate detailed local exploration, enhancing both the convergence and diversity of the algorithm. The experimental results show that KDDCP significantly outperforms 13 comparison algorithms on 33 test problems such as MW, DTLZ and DASCMOP as well as on realistic constrained multi-objective optimization problems commonly used in six engineering fields. The results confirm that KDDCP achieves an optimal balance among feasibility, diversity, and convergence under complex constraints.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"96 ","pages":"Article 101989"},"PeriodicalIF":8.2,"publicationDate":"2025-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144115057","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":"Knowledge-driven inverse diffusion prediction algorithm for flexible job shop scheduling problem considering transportation resources and multiple breakdowns","authors":"Cong Wang,&nbsp;Lixin Wei,&nbsp;Hao Sun,&nbsp;Ziyu Hu","doi":"10.1016/j.swevo.2025.101979","DOIUrl":"10.1016/j.swevo.2025.101979","url":null,"abstract":"<div><div>The complex coupling between transportation and production poses significant challenges for scheduling algorithms, particularly in dynamic environments where contingencies are frequent. The traditional Flexible Job Shop Scheduling Problem (FJSP) does not fully account for these factors. Therefore, a two-stage knowledge-driven inverse diffusion prediction algorithm (TKIDP) is proposed for FJSP with transportation resources and multiple breakdowns (MBFJSPT). In the first phase, TKIDP operates in a static environment and aims to minimize the makespan and load imbalance, thereby reducing the risk of machine breakdowns. When a machine breakdown occurs, TKIDP enters a rescheduling phase and minimizes total energy consumption as an additional objective. An inverse diffusion prediction strategy is proposed to address the occurrence of multiple machine breakdowns. By learning from optimization processes of historical breakdowns, the proposed strategy identifies potential evolutionary patterns and predicts the evolutionary trends of future breakdowns. A similarity measure based on mutual entropy is introduced to select the most relevant historical breakdowns for learning. This method enhances prediction accuracy. To further improve exploration and exploitation capabilities, an adaptive competitive reconfiguration strategy is proposed. This strategy incorporates multiple deletion and reconstruction operators to guide the algorithm toward the optimal solution. To verify the effectiveness of MBFJSPT, a dedicated MB test set is constructed. On the MB1-18 test, TKIDP is compared with five other algorithms. Based on the Inverted generational distance (IGD), hypervolume (HV), and relative deviation (RD) metrics, TKIDP achieves the best results on 15, 18, and 18 out of the 18 test instances, respectively.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"96 ","pages":"Article 101979"},"PeriodicalIF":8.2,"publicationDate":"2025-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144107669","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":"Potential-driven multi-learning particle swarm optimisation","authors":"Weitao Zhao ,&nbsp;Zati Hakim Azizul ,&nbsp;Chaw Seng Woo ,&nbsp;Weijie Kuang ,&nbsp;Yafeng Li","doi":"10.1016/j.swevo.2025.101993","DOIUrl":"10.1016/j.swevo.2025.101993","url":null,"abstract":"<div><div>The particle swarm optimisation (PSO) algorithm is a simple and effective metaheuristic algorithm. However, its search strategy may lead to issues such as getting trapped in local optima and premature convergence when solving complex multimodal problems. This paper proposes a Potential-Driven Multi-Learning PSO (PDML-PSO) to enhance the global search capability of the PSO algorithm. In this algorithm, particles are classified into three levels based on their performance: elite particles, potential particles, and regular particles. A multi-learning approach is employed to assign different search priorities to each level. Specifically, a non-traditional selection criterion based on the number of consecutive gradient steps related to fitness is used to classify potential particles. This method retains the advantage of gradient descent in accelerating the search while mitigating its drawback of quickly getting trapped in local optima. To validate the performance of PDML-PSO, tests were conducted on 30-dimensional, 50-dimensional, and 100-dimensional problems from the CEC2017 benchmark suite and on 10-dimensional and 20-dimensional problems from the CEC2022 benchmark suite. The results were compared with those of nine other PSO algorithms. The experimental results demonstrate that PDML-PSO exhibits superior global search capability compared to the nine algorithms. Furthermore, ablation experiments confirmed the effectiveness of the improvements made in PDML-PSO. All experimental results highlight PDML-PSO’s performance advantage in solving complex multimodal problems.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"96 ","pages":"Article 101993"},"PeriodicalIF":8.2,"publicationDate":"2025-05-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144107591","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}