Swarm and Evolutionary Computation最新文献

{"title":"An attention-based joint value estimation strategy for multi-agent coordination optimization","authors":"Ze Wang ,&nbsp;Ni Li ,&nbsp;Guanghong Gong ,&nbsp;Haitao Yuan","doi":"10.1016/j.swevo.2025.102132","DOIUrl":"10.1016/j.swevo.2025.102132","url":null,"abstract":"<div><div>Coordination optimization plays a vital role in complex multi-agent systems, and Multi-Agent Reinforcement Learning (MARL) has emerged as a widely adopted solution. However, MARL still faces significant challenges in this domain, including low coordination efficiency and inaccurate value estimation. To address these issues, we propose MVAPO, a novel Multi-Head Joint Value Attention-based Policy Optimization algorithm that improves policy learning through enhanced value approximation and selective attention to agent contributions. The key innovation of MVAPO lies in the introduction of a joint value network augmented with a multi-head attention mechanism. In this mechanism, context-aware team rewards serve as query inputs, directing attention to the most relevant agents in different situations. This allows the model to dynamically focus on the agents that are most critical at any given time, thus improving coordination efficiency and the accuracy of value estimates. Furthermore, MVAPO incorporates feedforward and residual layers, eliminating linear and monotonic constraints, which significantly enhances its representational capacity. Extensive experiments on a multi-UAV benchmark across a variety of scenarios demonstrate that MVAPO consistently outperforms state-of-the-art methods in both reward acquisition and win rates, highlighting its superior performance and robustness.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102132"},"PeriodicalIF":8.5,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145011142","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 self-learning classification-based multi-objective evolutionary algorithm for machine multi-state energy-efficient flexible job shop scheduling under time-of-use pricing","authors":"Da Wang ,&nbsp;Lina Qian ,&nbsp;Kai Zhang ,&nbsp;Dengwang Li ,&nbsp;Shicun Zhao ,&nbsp;Junqing Li","doi":"10.1016/j.swevo.2025.102142","DOIUrl":"10.1016/j.swevo.2025.102142","url":null,"abstract":"<div><div>Driven by the “dual carbon” strategic goals, the coordinated optimization of energy consumption and production efficiency has become a core issue for manufacturing industries. As an important means to promote energy structure transformation, electric substitution has made significant progress in industrial manufacturing, transportation, household electrification, and other fields. Among them, industrial production accounts for over 60% of the total electric energy substitution, becoming the largest electricity consumer. Note that the electricity price is based on time-of-use pricing (TOU), meanwhile, electric consumption is related to the machine multi-state (MM). Regarding these matters, this study focuses on determining sensible machine states and formulating reasonable production scheduling plan, to minimize both production time and power consumption. First, a novel energy-efficient flexible job shop scheduling problem is developed, which considers both the TOU strategy and the MM conditions (EFJSP-MM-TOU). Second, a self-learning classification-based multi-objective evolutionary algorithm (SCMOEA) is proposed to solve the EFJSP-MM-TOU. In specific, the SCMOEA enhances population diversity through a hybrid initialization strategy, adopts a dynamic selection of cross individuals based on the self-learning classification mechanism to improve the search efficiency, and designs four local search operators to increase the potential for approaching better positions. Third, by employing the MK standard dataset in EFJSP-MM-TOU, the proposed SCMOEA is compared with its three variants and five state-of-the-art algorithms to verify its optimization performance. The experimental results suggest that SCMOEA has advantages in terms of Pareto optimal solutions’ diversity and convergence. Finally, by testing in an actual enterprise case, the results further support the effectiveness of the EFJSP-MM-TOU and the significance of SCMOEA.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102142"},"PeriodicalIF":8.5,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145011295","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":"PheroCom: Decentralised and asynchronous robot swarm coordination framework based on virtual pheromone and vibroacoustic communication","authors":"Claudiney R. Tinoco,&nbsp;Luiz Gustavo A. Martins,&nbsp;Gina M.B. Oliveira","doi":"10.1016/j.swevo.2025.102083","DOIUrl":"10.1016/j.swevo.2025.102083","url":null,"abstract":"<div><div>Representing and controlling the dynamics of stigmergic substances used by bio-inspired approaches pose significant challenges when applied to robotics. In order to overcome this challenge, this work proposes a framework based on the virtualisation and control of these substances at a local scope, with the primary goal of coordinating robot swarms. This framework introduces a novel pheromone representation that enables decentralisation and decision asynchronicity, while its lightweight design ensures accessibility to resource-constrained platforms. Each robot maintains an independent virtual pheromone map in its memory, which is continuously updated through its own pheromone deposits and evaporation. Additionally, each robot’s pheromone map is also updated by aggregating information from other robots that are exploring nearby areas. Consequently, individual and independent maps eliminate the need for a centralised agent to manage and distribute pheromone information. This propagation mechanism is inspired by ants’ vibroacoustic communication, which is characterised as a form of indirect communication. The framework was evaluated using an agent-based mass simulation tool and a real-world simulation platform. Experiments were conducted to validate the framework in diverse environments, with variations in shapes, sizes, and the number of robots. Results demonstrated that this proposal can effectively perform the coordination of robot swarms, and the robots have exhibited satisfactory performance while executing the surveillance task.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102083"},"PeriodicalIF":8.5,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145011297","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":"Exploratory landscape analysis on black-box optimization problems via Graph Neural Network","authors":"Xu Yang ,&nbsp;Rui Wang ,&nbsp;Kaiwen Li ,&nbsp;Wenhua Li ,&nbsp;Tao Zhang","doi":"10.1016/j.swevo.2025.102136","DOIUrl":"10.1016/j.swevo.2025.102136","url":null,"abstract":"<div><div>Most real-world optimization problems are poorly understood, some of which are black-box optimization problems (BBOPs). Exploratory landscape analysis (ELA) paves the way for algorithm design to deal with BBOPs. Existing ELA methods have limitations on unseen problems and lack analysis on the problem itself. To this end, this study introduces a novel ELA framework leveraging Graph Neural Network (GNN) upon BBOP’s surrogate model. Specifically, a neural network surrogate model is constructed whose architecture is utilized to represent BBOP in the form of graph. Then, GNN is responsible for capturing the relationships between the graph-represented BBOP and high-level features. As one of the most notable features in optimization, multimodality of multi-objective problems is to be identified for illustration. More than 99% accuracy on independent test set demonstrates the effectiveness of the proposed framework with simultaneously avoiding the effect of problem dimensions.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102136"},"PeriodicalIF":8.5,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145020703","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":"Adaptive landscape-aware repelling restart covariance matrix adaptation-evolution strategy for multimodal and global optimization","authors":"Xikang Wang,&nbsp;Tongxi Wang,&nbsp;Hua Xiang","doi":"10.1016/j.swevo.2025.102143","DOIUrl":"10.1016/j.swevo.2025.102143","url":null,"abstract":"<div><div>In multimodal optimization using Covariance Matrix Adaptation-Evolution Strategy (CMA-ES), redundant restarts are caused by repeated convergence to previously explored local basins, which leads to significant computational resource waste. To address this problem, previous research proposed the concept of Repelling Restart and developed RR-CMA-ES, but issues remain regarding rigid repulsion and gradient information of local basin structures. Building on this foundation, we propose an Adaptive Landscape-aware Repelling Restart CMA-ES (ALR-CMA-ES) that enhances the original RR-CMA-ES through three key improvements: 1) A fitness sensitive dynamic exclusion mechanism that adaptively adjusts tabu region radius based on local optimality and convergence frequency, prioritizing avoidance of high-quality basins; 2) A covariance matrix mechanism preserving convergence history to geometrically align hyper-ellipsoidal exclusion regions with explored local basin landscapes; 3) A Boltzmann-like probabilistic acceptance scheme incorporating exclusion regions, permit- ting controlled exploration near tabu boundaries. Experiments on the BBOB benchmark demonstrate that ALR-CMA-ES outperforms RR-CMA-ES in 90% of tested problems spanning 2D to 50D. This method provides a practical solution for expensive black-box optimization by systematically integrating landscape topology awareness into tabu mechanisms, while proposing a new solution for multimodal optimization problems.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102143"},"PeriodicalIF":8.5,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145011296","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 intelligent automated guided vehicle scheduling framework for manufacturing: Balancing energy, efficiency, and task completion","authors":"Xiang Huo,&nbsp;Lei Nie","doi":"10.1016/j.swevo.2025.102127","DOIUrl":"10.1016/j.swevo.2025.102127","url":null,"abstract":"<div><div>In recent years, the widespread usage of Automated Guided Vehicles (AGVs) has become prevalent in material transportation systems of industries. The AGVs are known for their operational flexibility and efficiency, but efficient scheduling remains a crucial issue due to the conflicting factors, including deviation penalties for task execution times, power consumption, overall task completion time, collision risk, and utilization efficiency. To address this, this research employs a multi-objective mixed-integer programming model (MO-MIP) to formulate the scheduling problem of AGVs. The optimization algorithms, such as Non-dominated Sorting Genetic Algorithm II (NSGA-II) and the Reference Point-based Non-dominated Sorting Genetic Algorithm (NSGA-III) are utilized to obtain the Pareto optimal solutions in solving the scheduling problem. The simulation experiment on three distinct manufacturing workshop scenarios was conducted to examine the effectiveness of the model. The outcomes illustrated that the NSGA-II and NSGA-III exhibit reduced penalty cost, power consumption, collision risk, task completion time, and higher utilization efficiency. These algorithms also showed better computational efficiency and outperformed baseline algorithms under three manufacturing scenarios. These outcomes indicate that the proposed method is a promising solution for the industrial sector to perform material transportation in an efficient manner.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"98 ","pages":"Article 102127"},"PeriodicalIF":8.5,"publicationDate":"2025-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145004301","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 bi-objective routing problem for cooperated trucks and drones in river water quality monitoring","authors":"Shuo Dang ,&nbsp;Zhihao Luo ,&nbsp;Zhong Liu ,&nbsp;Yuzhen Zhou ,&nbsp;Jianmai Shi","doi":"10.1016/j.swevo.2025.102148","DOIUrl":"10.1016/j.swevo.2025.102148","url":null,"abstract":"<div><div>The preservation of urban river ecosystems constitutes a fundamental component in sustainable water resource management. Effective water quality evaluation relies on systematic sampling approaches that encompass diverse aquatic environments. Traditional vessel-based sampling methods are often time-consuming, labor-intensive, and pose a risk of contamination to the aquatic ecosystem. In contrast, the cooperation between ground vehicles (referred to as trucks) and drones offers an efficient and environmentally friendly sampling method. This paper presents a novel approach to water sampling in which drones serve as the primary tools for collecting water samples while trucks extend the flight range of drones by acting as mobile depots. A Mixed Integer Quadratic Programming (MIQP) model is formulated which considers the influence of payload variations of drones on energy consumption during the sampling process. To synchronize the routing of drones and trucks, an improved Non-dominated Sorting Genetic Algorithm-II (NSGA-II) is developed, incorporating an iterative front optimization strategy to enhance the solution diversity. Furthermore, three specialized genetic operators tailored to the specific problem scenario are designed to improve the quality of the population. The practicality and efficiency of the proposed algorithm are validated using real-world case studies, highlighting the transformative potential of truck-drone cooperation in urban river water quality monitoring.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"98 ","pages":"Article 102148"},"PeriodicalIF":8.5,"publicationDate":"2025-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144996492","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":"Adaptive information fusion–driven evolutionary algorithm via balancing the information from unconstrained and constrained pareto fronts","authors":"Xujie Tan ,&nbsp;Yalin Wang ,&nbsp;Chenliang Liu ,&nbsp;Jing Liao ,&nbsp;Yong Wang ,&nbsp;Weihua Gui","doi":"10.1016/j.swevo.2025.102150","DOIUrl":"10.1016/j.swevo.2025.102150","url":null,"abstract":"<div><div>Obtaining well–converged and well–distributed constrained Pareto fronts (CPFs) is the ultimate goal of solving constrained multi–objective optimization problems (CMOPs). In recent years, leveraging information from the unconstrained Pareto front (UPF) has become a prevalent method for CMOPs. However, the equilibrium and representation of information from CPF and UPF are crucial to the performance of evolutionary algorithms. To balance the information from CPF and UPF adaptively, this paper proposes an adaptive information fusion–driven evolutionary algorithm, referred to as AIFDEA. Specifically, the evolutionary process of AIFDEA is divided into infeasible and feasible stages. During the infeasible stage, a clustering–based individual selection strategy is proposed to balance diversity and feasibility. Furthermore, a novel fitness function that integrates UPF, CPF, and diversity information adaptively is designed to balance convergence, feasibility, and diversity in the feasible stage. The superiority of the proposed method is substantiated throught extensive comparison experiments across 34 benchmark functions and parameter analysis experiments. Additionally, application experiments on 16 real–world benchmark CMOPs and an energy consumption optimization problem in copper electrolysis process are conducted, to validate the practical applicability of AIFDEA in diverse real–world and complex industrial environments. Moreover, this paper demonstrates that fusing UPF, CPF, and diversity information adaptively is promising for CMOPs.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"98 ","pages":"Article 102150"},"PeriodicalIF":8.5,"publicationDate":"2025-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144988484","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":"Asynchronous storming-norming ant colony system for project scheduling under scenario perturbation of activity attributes","authors":"Wen Shi ,&nbsp;Feng-Feng Wei ,&nbsp;Xiaolin Bo ,&nbsp;Zhisheng Bi ,&nbsp;Jianing Xi ,&nbsp;Wei-Neng Chen","doi":"10.1016/j.swevo.2025.102141","DOIUrl":"10.1016/j.swevo.2025.102141","url":null,"abstract":"<div><div>Project scheduling is essential in complex fields like infrastructure and healthcare, but current methods often fail to capture the subtle interplay between different activity attributes, such as cost and duration. These factors, fraught with uncertainty, can lead to unpredictable changes and unreliable assessments. To alleviate these issues, we introduce a novel approach called Asynchronous Storming-Norming Ant Colony System based on Scenario Unification simulation(ASN-ACS-SU). Firstly, we introduce a scenario unification framework, ensuring simulations are consistent across scenarios, thus greatly enhancing stability. Next, we integrate a storming-norming strategy, adapting different tactics based on the evolution stage, effectively accelerating convergence. Lastly, an asynchronous scheme, tailored for handling various activity attributes asynchronously, is incorporated to improve the effectiveness of the solution. These components are skillfully integrated into the Ant Colony System framework, ensuring a harmonious combination of their individual strengths. Comprehensive tests using simulation datasets and demonstrate that ASN-ACS-SU significantly outperforms existing algorithms, including CH-GA, LRBH, Hybrid DE and Two-stage GA, which are state-of-the-art algorithms for multi-mode resource-constrained project scheduling problems. The proposed method demonstrates non-inferiority to CH-GA and Hybrid DE in at least 90% of scenarios, to Two-stage GA in at least 80% of scenarios and to LRBH in at least 70% of scenarios in all datasets. Thus, the validity and reliability of the ASN-ACS-SU can be demonstrated.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"98 ","pages":"Article 102141"},"PeriodicalIF":8.5,"publicationDate":"2025-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144932118","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":"Non-probabilistic set-based selection strategy for multi-objective optimization with interval uncertainties","authors":"Qianqian Yu ,&nbsp;Chen Yang ,&nbsp;Guangming Dai ,&nbsp;Lei Peng","doi":"10.1016/j.swevo.2025.102153","DOIUrl":"10.1016/j.swevo.2025.102153","url":null,"abstract":"<div><div>Uncertainty introduces great challenges for multi-objective optimization, as the solutions are no longer deterministic values, complicating both the accurate evaluation of solution quality and the selection of elite solutions. Misjudging the dominance relationship among uncertain solutions may result in the loss of superior solutions, and imprecise crowding distance quantification may fail to maintain population diversity. Therefore, a novel non-probabilistic set-based selection strategy (NPS) is developed to balance the convergence and diversity of uncertain populations. It employs a novel two-dimensional interval dominance relationship and an interval crowding distance model to determine the new parent population. Additionally, a dimension-wise approach (DWA), a non-intrusive uncertainty analysis model, is used to quantify the bounds of uncertain optimization objectives. Furthermore, a novel interval crowding distance-based sample standard deviation metric is proposed to enhance the accuracy of diversity evaluation for uncertain populations. The proposed NPS is integrated into two classical multi-objective optimization frameworks and is compared with other selection strategies across multiple groups of benchmarks. The results indicate that algorithms incorporating NPS and DWA can not only effectively explore the Pareto Front under uncertainties but also directly evaluate uncertain solutions with limited samples. Compared with other selection strategies, NPS can explore an optimal solution set with superior convergence, higher diversity, and lower uncertainty.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"98 ","pages":"Article 102153"},"PeriodicalIF":8.5,"publicationDate":"2025-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144925809","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}