Swarm and Evolutionary Computation最新文献

{"title":"K-DPSA: A hierarchical surrogate-assisted evolutionary algorithm for expensive constrained multi-objective optimization with application to UAV path planning","authors":"Ai-Qing Tian,&nbsp;Fei-Fei Liu,&nbsp;Shao-Quan Ni,&nbsp;Jeng-Shyang Pan","doi":"10.1016/j.swevo.2026.102490","DOIUrl":"10.1016/j.swevo.2026.102490","url":null,"abstract":"<div><div>Expensive Constrained Multi-Objective Optimization Problems (ECMOPs) arise in engineering domains where each fitness evaluation requires costly simulation and hard constraints fragment the feasible region. This paper proposes K-DPSA, a hierarchical surrogate-assisted evolutionary algorithm with a three-stage architecture for ECMOPs. An Ensemble Classifier layer pre-filters infeasible candidates before expensive evaluation, concentrating the budget within feasible corridors. A Two-Indicator Surrogate maps the objective space onto convergence and diversity indicators, halving the surrogate model count while preserving the Pareto trade-off structure. An uncertainty-aware adaptive infill criterion balances exploitation and exploration through a fixed split without additional hyper-parameters. Validated on a UAV post-disaster path planning benchmark with four objectives and six constraints, K-DPSA was compared against six baselines spanning classical MOEAs, Kriging-based SAEAs, scalarization-based SAEAs, and hierarchical methods. Across three scenarios of increasing complexity under a 300-evaluation budget, K-DPSA achieved Inverted Generational Distance values of 0.045, 0.052, and 0.068, Hypervolume values of 0.782, 0.756, and 0.712, and success rates of 100%, 95%, and 90%. All pairwise comparisons were statistically significant (<span><math><mrow><mi>p</mi><mo>&lt;</mo><mn>0.05</mn></mrow></math></span>, Holm–Bonferroni). Removing the EC-layer caused the success rate to drop from 90% to 32%, and replacing the TIS-layer increased IGD by 29%–40%, confirming each component’s contribution.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"107 ","pages":"Article 102490"},"PeriodicalIF":9.6,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148730831","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 population partition and prediction strategy-based evolutionary algorithm for dynamic multi-objective optimization","authors":"Chunliang Zhao,&nbsp;Zhihao Xiao,&nbsp;Jing Sun,&nbsp;Yi Xiang,&nbsp;Dunwei Gong","doi":"10.1016/j.swevo.2026.102476","DOIUrl":"10.1016/j.swevo.2026.102476","url":null,"abstract":"<div><div>Dynamic multi-objective optimization problems (DMOPs) are difficult because the Pareto set (PS) and Pareto front (PF) change over time, requiring algorithms to respond quickly while maintaining convergence and diversity. Existing prediction-based dynamic multi-objective evolutionary algorithms (DMOEAs) often rely on linear assumptions or individual-level modeling, which may limit their ability to respond to non-affine or structurally changing PS trajectories. To address this issue, this paper proposes a dynamic multi-objective evolutionary algorithm based on population partition and prediction strategy, termed PPDMOEA. It employs a cooperative response mechanism in which elite solutions mainly support convergence recovery and regular solutions preserve diversity. First, a two-dimensional evaluation strategy is used to select representative elite solutions for prediction by considering both convergence and distribution. Then, an order-insensitive set-encoded LSTM model is used to describe the temporal evolution of elite solution sets without relying on fixed element positions. Finally, a dynamic solution redistribution strategy is introduced to improve diversity and adaptability in new environments. Experiments on the CEC2018 dynamic multi-objective benchmark against seven representative DMOEAs show that the proposed algorithm achieves competitive MIGD and MHV performance on this benchmark, while its advantages are not uniform across all problem instances.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"107 ","pages":"Article 102476"},"PeriodicalIF":9.6,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148642254","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 hybridized Thompson Sampling hyperheuristic for the truck scheduling problem at a cross-docking facility","authors":"Bokang Li,&nbsp;Marta Borowska-Stefańska,&nbsp;Szymon Wiśniewski,&nbsp;Qianwen Guo,&nbsp;Yui-yip Lau,&nbsp;Maxim A. Dulebenets","doi":"10.1016/j.swevo.2026.102502","DOIUrl":"10.1016/j.swevo.2026.102502","url":null,"abstract":"<div><div>Cross-docking has become a widely adopted logistics strategy for improving distribution efficiency by reducing inventory requirements and accelerating freight transshipment across supply chains. However, efficient terminal operations require reliable coordination of inbound and outbound truck activities under tightly synchronized and time-sensitive operating conditions, making the development of efficient service schedules highly challenging in practice. This study investigates the truck scheduling problem at a mixed-mode cross-docking facility. A mixed-integer linear programming model aimed at reducing total operational costs is adopted. To effectively solve this computationally complex problem, a Hybridized Thompson Sampling Hyperheuristic (HTSH) algorithm is developed, which integrates a population-based evolutionary framework with a Thompson Sampling-based adaptive control mechanism that dynamically learns and selects search operators according to observed search performance. A mixed repair-penalization infeasibility handling strategy is incorporated to efficiently manage inbound-outbound coordination constraints. Moreover, a hybrid fitness evaluation mechanism combining fast approximate assessment with exact optimization evaluation is applied for efficient exploration together with accurate solution refinement. A set of extensive computational experiments demonstrates that HTSH consistently produces high-quality scheduling solutions comparable with exact optimization on small instances and substantially outperforms a set of classical and more recent metaheuristics across a wide range of problem instances. The HTSH algorithm is able to achieve improvement margins of up to approximately 24% while maintaining an average runtime of 186.71 s. The experimental results demonstrate that the proposed HTSH framework provides an effective and scalable solution approach for cross-docking truck scheduling, contributing to improved operational efficiency and service reliability in distribution terminal operations.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"107 ","pages":"Article 102502"},"PeriodicalIF":9.6,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148730832","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 novel particle swarm optimizer with failure-aware searching framework for neural mass model inversion","authors":"Han Li,&nbsp;Yang Zou,&nbsp;Jiyu Tan,&nbsp;Zihao Jia,&nbsp;Yurong Li,&nbsp;Nianyin Zeng","doi":"10.1016/j.swevo.2026.102504","DOIUrl":"10.1016/j.swevo.2026.102504","url":null,"abstract":"<div><div>In this paper, a novel particle swarm optimizer (PSO) is proposed based on a designed failure-aware searching (FAS) framework. Named FAS-PSO, the proposed algorithm seeks to learn potential valuable information from those non-improved evolutionary experiences rather than directly discarding them. Considering the topological status of population, each particle is equipped with a structural probe to assess local regions, which integrates information from different individuals and enhances the communication among population. Next, in case of a probing location with better fitness, the swarm will be updated accordingly. Otherwise, by measuring the deterioration extent of the probe as compared to the initial particle, an information reinvestigation mechanism is adopted to generate a positional modification for swarm update, including continuous forward search and escape in the opposite direction. Extensive experiments demonstrate the performance of FAS-PSO is superior to that of other state-of-the-art swarm intelligence-based algorithms. Moreover, the proposed FAS-PSO is further extended to a multi-objective version and is successfully applied to accomplish the reliable and robust parameter estimation of a Wendling neural mass model, showing considerable engineering practicality in solving the complicated biomedical model inversion task.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"107 ","pages":"Article 102504"},"PeriodicalIF":9.6,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148730824","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":"Bi-objective optimization of 3D track layout for railway humps: rolling efficiency and construction cost","authors":"Wei Li,&nbsp;Yang Ran,&nbsp;Hao Pu,&nbsp;Xinjie Wan,&nbsp;Paul Schonfeld,&nbsp;Shaoxiang Huang,&nbsp;Shicheng Zhang","doi":"10.1016/j.swevo.2026.102503","DOIUrl":"10.1016/j.swevo.2026.102503","url":null,"abstract":"<div><div>The railway hump is a critical facility in train classification, and its geometric design directly affects operational efficiency and investment. Existing studies mainly focus on vertical profile optimization while keeping the horizontal alignment fixed, which restricts the search space and may overlook superior design alternatives. To resolve this issue, a 3D track layout optimization model for railway humps is developed, jointly optimizing horizontal and vertical hump alignments under their coupling constraints. Rolling efficiency and construction cost are formulated as bi-objective functions. To efficiently solve the constrained optimization problem, a particle swarm optimization (PSO) algorithm with a feasibility-bound preprocessing strategy is proposed to avoid infeasible search regions and enhance the search efficiency. In a real-world case study, the recommended solution reduces rolling time and intervals between successive railcars by 15.21% and construction cost by 4.02% compared with the best manual design. Comparative experiments with genetic algorithm (GA), original PSO, and a customized PSO variant demonstrate that the proposed method achieves improved solution quality and computational efficiency.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"107 ","pages":"Article 102503"},"PeriodicalIF":9.6,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148730828","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":"Feasibility-guaranteed multi-objective evolutionary path planning with knowledge-based local guidance","authors":"Tonghao Wang,&nbsp;Xiaodong Yang,&nbsp;Handing Wang,&nbsp;Chaoyue Zhao","doi":"10.1016/j.swevo.2026.102481","DOIUrl":"10.1016/j.swevo.2026.102481","url":null,"abstract":"<div><div>Multi-objective path planning is an important yet challenging problem for mobile agents working in complex environments with obstacles and risks. While evolutionary algorithms show significant promise, they often struggle to guarantee collision-free paths. To address this, we propose a novel multi-objective evolutionary path planning framework that ensures path feasibility and employs two local guidance operators to accelerate convergence. First, a two-stage constraint-handling mechanism is designed to ensure path feasibility. It resolves point violations in which waypoints lie within obstacles via a remapping strategy, and corrects segments that cross obstacles via a deterministic correction strategy. This ensures all individuals in the population represent feasible, i.e., obstacle-free, paths, enabling accurate fitness evaluation without penalty functions. Second, we design two local guidance operators that leverage problem-specific knowledge to guide elite solutions toward desirable regions in the objective space, thereby accelerating convergence to the extremities of the Pareto front. These operators guide the elite solutions, stored in two separate archives for path length and risk respectively, toward shorter or safer regions before reintegrating them into the population with an adapted replacement strategy. Extensive experiments on 30 randomly generated problems demonstrate the superiority of the proposed algorithm. Furthermore, ablation studies confirm that both the constraint-handling mechanism and the local guidance operators perform as expected.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"107 ","pages":"Article 102481"},"PeriodicalIF":9.6,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148642256","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":"Energy-aware hybrid flow-shop scheduling problem considering ageing workforce: A Multi-Tasking Metaheuristic","authors":"Jiamei Cao,&nbsp;Zhongshi Shao,&nbsp;Weishi Shao","doi":"10.1016/j.swevo.2026.102480","DOIUrl":"10.1016/j.swevo.2026.102480","url":null,"abstract":"<div><div>In modern manufacturing system, despite the advancement of automation, human workers still remain an indispensable resource. However, the impacts of the ageing workforce are frequently overlooked in task assignment and resource allocation, leading to suboptimal productivity. Meanwhile, modern manufacturing systems are facing increasing pressure to reduce energy consumption. The workforce characteristics directly influence machine utilization and processing efficiency, thereby affecting overall energy consumption. Therefore, this paper investigates the energy-aware hybrid flow-shop scheduling problem considering ageing workforce (EAHFSP-AW), aiming to minimize makespan and total energy consumption simultaneously. First, a mixed-integer linear programming (MILP) model is formulated to characterize the EAHFSP-AW. To solve the problem efficiently, a Multi-Tasking Metaheuristic (MTM) is proposed. A specialized decoding heuristic based on two dispatching rules is developed. Unlike traditional multi-objective metaheuristics that rely on serial or synchronous structure, the MTM adopts a parallel and asynchronous search framework composed of three search tasks with differentiated optimization responsibilities. Two convergence-oriented auxiliary tasks intensify the search for boundary solutions corresponding to makespan and energy consumption, respectively, while a diversity-oriented main task focuses on exploring well-distributed non-dominated solutions. These tasks interact asynchronously through a shared external archive, where the convergence-oriented tasks continuously contribute promising boundary solutions and the diversity-oriented task exploits such information to search a well-converged and well-distributed non-dominated solution set. Finally, comprehensive computational experiments are conducted. Experimental results validate the effectiveness of the MTM’s architectural design and its superior performance in solving the EAHFSP-AW compared to state-of-the-art algorithms.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"107 ","pages":"Article 102480"},"PeriodicalIF":9.6,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148642690","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 novel information entropy-based decomposition-integration multi-objective evolutionary algorithm for balancing exploration and exploitation","authors":"Deling Yuan,&nbsp;Zhen Peng,&nbsp;Zitao Hong,&nbsp;Xueping Liu","doi":"10.1016/j.swevo.2026.102489","DOIUrl":"10.1016/j.swevo.2026.102489","url":null,"abstract":"<div><div>Decomposition-based multi-objective evolutionary algorithms (MOEAs) are widely adopted paradigms for solving multi-objective optimization problems (MOPs). While their efficacy is well-established, these algorithms face three core unresolved challenges: inherent difficulty in balancing exploration and exploitation, susceptibility to local optima during neighborhood updates, and Pareto front degradation. Most existing works prioritize objective space optimization, neglecting the non-negligible impact of decision space dispersion on algorithms’ global exploration capability. To address these gaps, this paper proposes an information entropy-based decomposition-integration MOEA (IE-MOEA/DI) to advance the performance of decomposition-based MOEAs. By replacing conventional random initialization with a <em>Riesz s-Energy</em> based uniform population seeding strategy, the algorithm delivers balanced starting points for all subproblems to boost global exploration potential. To address the oversight of decision-space diversity, a new information entropy metric is introduced to quantify population dispersion in the decision space, complementing conventional objective-space-only exploration evaluation. An integrated framework fusing three mainstream scalarization functions is further developed to balance exploration and exploitation, mitigating premature local convergence during rapid neighborhood updates. Complementing these designs, a search angle-guided elite crossover strategy is incorporated to enhance high-quality gene propagation and convergence speed. Extensive experiments on standard benchmark suites verify that IE-MOEA/DI achieves superior overall performance over state-of-the-art decomposition-based MOEAs on IGD⁺ and HV metrics, effectively alleviating Pareto front degradation and realizing a well-balanced exploration-exploitation trade-off.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"107 ","pages":"Article 102489"},"PeriodicalIF":9.6,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148642691","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":"SNN-RET: A spiking neural network with reference-guided environmental selection and turbulence for multi- and many-objective optimization","authors":"Issam Zidi,&nbsp;Salim El Khediri","doi":"10.1016/j.swevo.2026.102499","DOIUrl":"10.1016/j.swevo.2026.102499","url":null,"abstract":"<div><div>Many-objective optimization is hard because improving several competing objectives at once slows convergence and erodes the spread of the non-dominated set. This work introduces SNN-RET, a spiking-neural framework that couples membrane-state neural adaptation, reference-guided environmental selection, and a turbulence-based recovery mechanism. Unlike methods that adapt only at the operator level, its spiking layer directly drives candidate generation and is reinforced by Pareto survival and diversity contribution, steering the search toward promising regions while preserving a well-distributed Pareto front. SNN-RET is compared with six baselines, including the reference-vector method RVEA, over 31 runs on DTLZ, WFG, ZCAT, and the constrained MW suite, with all indicators computed on an equal-size output set. It gives the strongest overall profile at two and three objectives, leading on hypervolume and non-dominated count; on many-objective problems it is mid-pack at 5000 evaluations but reaches the best median hypervolume on several instances as the budget grows. Overall, SNN-RET raises the average hypervolume by 17.65% and the number of non-dominated solutions by 33.48%, while reducing GD, IGD, spacing, and additive epsilon, and a non-parametric analysis ranks it best on five of six indicators. These results show SNN-RET is a robust framework for multi- and many-objective optimization, combining spiking-neural adaptation, reference-based selection, and diversity recovery in one unified search process.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"107 ","pages":"Article 102499"},"PeriodicalIF":9.6,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148730226","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 global evolutionary algorithm using filled functions and estimation of distribution for robust multiobjective optimization","authors":"Yanhui Tang,&nbsp;Hui Li,&nbsp;Yuxiang Shui,&nbsp;Jianyong Sun,&nbsp;Qingfu Zhang","doi":"10.1016/j.swevo.2026.102496","DOIUrl":"10.1016/j.swevo.2026.102496","url":null,"abstract":"<div><div>In robust multiobjective optimization, the mean effective objective function (MEOF) is commonly used to evaluate the robustness of solutions. However, when the original objective functions are multimodal, the global Pareto front of the MEOF-based problem may correspond to a local Pareto front of the original problem, which poses a significant challenge for the search of robust solutions. Consequently, traditional multiobjective evolutionary algorithms (MOEAs) face challenges in identifying the global robust MEOF-based PFs, as their primary goal is to locate the global PFs of the original MOPs. In this paper, we introduce a novel global multiobjective evolutionary algorithm that incorporates a filled function and estimation of distribution algorithm (EDA), denoted by MO-EDA/FF, to approximate the global robust PFs of MEOF-based MOPs under the framework of MOEA/D. In our proposed algorithm, the filled functions facilitate the search to escape from local optima by detecting robustness-related multimodal decision variables and reinitializing the evolutionary population within MOEA/D. Moreover, the EDA based on multivariate Gaussian distribution is used to optimize a family of filled weighted sum subproblems, enabling the sampling of numerous offspring solutions within promising search regions. We conduct some experiments to compare the performance of our proposed algorithm with several other state-of-the art MOEAs based on MEOF evaluations on a benchmark set of multiobjective test instances with robustness challenges. Our experimental results demonstrate the competitiveness and advantages of our proposed algorithm in solving these test instances.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"107 ","pages":"Article 102496"},"PeriodicalIF":9.6,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148642251","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}