Swarm and Evolutionary Computation最新文献_第10页

A consensus-based EDA with multi-scale neighborhood search for vehicle routing problem with pickup and delivery 基于共识的多尺度邻域搜索的车辆取货路径问题EDA

IF 8.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Swarm and Evolutionary Computation

Pub Date : 2025-11-10 DOI: 10.1016/j.swevo.2025.102216

Wei Wang , Yindong Shen

The vehicle routing problem with pickup and delivery (VRPPD) is concerned with planning optimal routes for a fleet of vehicles to meet the diverse demands of customers. In this problem, scenarios involving either simultaneous pickup and delivery (SPD) or mixed pickup and delivery (MPD) cause fluctuations in vehicle loads. Our theoretical analyses reveal that SPD tightens vehicle capacity constraints, reducing the number of feasible solutions, while MPD expands the feasible region, thus increasing the number of local optima. This necessitates a generic algorithm to balance diversification and intensification and to promote persistent exploration. Therefore, this paper considers multiple neighborhood scales and proposes a consensus-based estimation of distribution algorithm (EDA) incorporating the scalable large neighborhood search (SLNS) and the tour fragment recombination (TFR), abbreviated as CEDA-ST. In the CEDA-ST, population consensus is leveraged to estimate the distribution of optimal routes, generating a consensus matrix for individual construction and neighborhood searches. The SLNS operator conducts destroy-and-repair moves in large neighborhoods to promote diversification. Meanwhile, the TFR operator facilitates local improvements in small neighborhoods to enhance intensification. Furthermore, a stagnation-triggered diversity management (STDM) strategy is developed to eliminate redundant individuals, encouraging persistent exploration. Comparative experiments demonstrate its superiority. An effectiveness analysis and two ablation experiments highlight the contributions of consensus information and multi-scale neighborhood search, respectively. Additionally, a real-world case study on JD Logistics further validates the applicability of CEDA-ST in practical scenarios.

取货车辆路线问题（VRPPD）关注的是为车队规划最优路线，以满足客户的不同需求。在这个问题中，涉及同时拾取和交付（SPD）或混合拾取和交付（MPD）的场景会导致车辆负载的波动。理论分析表明，SPD收紧了车辆容量约束，减少了可行解的数量；MPD扩大了可行区域，增加了局部最优解的数量。这就需要一种通用算法来平衡多样化和集约化，并促进持续的探索。因此，本文考虑多个邻域尺度，提出了一种基于共识的分布估计算法（EDA），该算法结合了可扩展大邻域搜索（SLNS）和游片段重组（TFR），简称CEDA-ST。在cda - st中，利用人口共识来估计最优路线的分布，生成单个建筑和邻居搜索的共识矩阵。SLNS运营商在大型社区进行摧毁和修复行动，以促进多样化。同时，TFR运营商促进小社区的局部改善，以加强集约化。此外，我们还开发了一种停滞触发多样性管理（STDM）策略，以消除冗余的个体，鼓励持续的探索。对比实验证明了其优越性。有效性分析和两个消融实验分别突出了共识信息和多尺度邻域搜索的贡献。此外，通过对京东物流的实际案例研究，进一步验证了cda - st在实际场景中的适用性。

{"title":"A consensus-based EDA with multi-scale neighborhood search for vehicle routing problem with pickup and delivery","authors":"Wei Wang , Yindong Shen","doi":"10.1016/j.swevo.2025.102216","DOIUrl":"10.1016/j.swevo.2025.102216","url":null,"abstract":"<div><div>The vehicle routing problem with pickup and delivery (VRPPD) is concerned with planning optimal routes for a fleet of vehicles to meet the diverse demands of customers. In this problem, scenarios involving either simultaneous pickup and delivery (SPD) or mixed pickup and delivery (MPD) cause fluctuations in vehicle loads. Our theoretical analyses reveal that SPD tightens vehicle capacity constraints, reducing the number of feasible solutions, while MPD expands the feasible region, thus increasing the number of local optima. This necessitates a generic algorithm to balance diversification and intensification and to promote persistent exploration. Therefore, this paper considers multiple neighborhood scales and proposes a consensus-based estimation of distribution algorithm (EDA) incorporating the scalable large neighborhood search (SLNS) and the tour fragment recombination (TFR), abbreviated as CEDA-ST. In the CEDA-ST, population consensus is leveraged to estimate the distribution of optimal routes, generating a consensus matrix for individual construction and neighborhood searches. The SLNS operator conducts destroy-and-repair moves in large neighborhoods to promote diversification. Meanwhile, the TFR operator facilitates local improvements in small neighborhoods to enhance intensification. Furthermore, a stagnation-triggered diversity management (STDM) strategy is developed to eliminate redundant individuals, encouraging persistent exploration. Comparative experiments demonstrate its superiority. An effectiveness analysis and two ablation experiments highlight the contributions of consensus information and multi-scale neighborhood search, respectively. Additionally, a real-world case study on JD Logistics further validates the applicability of CEDA-ST in practical scenarios.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102216"},"PeriodicalIF":8.5,"publicationDate":"2025-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145519886","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}

引用次数: 0

A two-stage ensemble evolutionary algorithm for constrained multi-objective optimization 约束多目标优化的两阶段集成进化算法

IF 8.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Swarm and Evolutionary Computation

Pub Date : 2025-11-07 DOI: 10.1016/j.swevo.2025.102213

Sri Srinivasa Raju Modampuri, Jiahao Fan, Yanan Sun

In constrained multi-objective evolutionary algorithms (CMOEAs), selecting appropriate constraint-handling techniques (CHTs) is challenging without prior knowledge of the problem’s constraint severity or feasible region distribution. Ensemble frameworks that integrate multiple CHTs with distinct populations offer a promising solution but face issues like redundant evaluations and poor exploration–exploitation balance. To address these limitations, we propose a two-stage ensemble-based CMOEA (CMOEA-TENS) that dynamically prioritizes suitable CHTs based on problem characteristics. Specifically, in the first stage, a population dedicated to explore the unconstrained search space drives the evolutionary process, while remaining populations co-evolve by leveraging solutions identified by the exploratory population. In the second stage, an ensemble of distinct populations drives the evolutionary process, each co-evolving with a different CHT focused on feasibility, diversity, or convergence to exploit the feasible regions effectively. Furthermore, we introduce a novel Multi-Armed Bandit (MAB)-based decision-making strategy that, unlike existing static or random selection approaches, adaptively learns and selects the most suitable CHT-based population to drive the evolutionary process based on real-time performance feedback. This dynamic strategy explicitly reduces redundant functional evaluations and ensures better management of exploration–exploitation trade-offs. CMOEA-TENS was evaluated against eleven state-of-the-art algorithms across six popular test suites, encompassing 57 test instances and six real-world problems. The empirical results demonstrate that CMOEA-TENS effectively balances exploration and exploitation while avoiding redundant evaluations by dynamically selecting the most suitable CHT-based population to drive the evolutionary process. Additionally, an ablation study further validates the effectiveness of the designed MAB strategy.

在约束多目标进化算法（cmoea）中，在不知道问题约束严重程度或可行区域分布的情况下，选择合适的约束处理技术（CHTs）是一项挑战。集成多个具有不同人口的cht的集成框架提供了一个有希望的解决方案，但面临诸如冗余评估和不良勘探开发平衡等问题。为了解决这些限制，我们提出了一种基于两阶段集成的CMOEA (CMOEA- tens)，它根据问题特征动态地优先考虑合适的cht。具体来说，在第一阶段，致力于探索无约束搜索空间的种群驱动进化过程，而其余种群通过利用探索种群确定的解决方案共同进化。在第二阶段，不同种群的集合驱动进化过程，每个种群都以不同的CHT共同进化，聚焦于可行性、多样性或收敛性，以有效地利用可行区域。此外，我们引入了一种新的基于Multi-Armed Bandit （MAB）的决策策略，与现有的静态或随机选择方法不同，该策略自适应学习并选择最合适的基于cht的种群来驱动基于实时性能反馈的进化过程。这种动态策略明确地减少了冗余的功能评估，并确保更好地管理勘探开发权衡。CMOEA-TENS在6个流行的测试套件中对11个最先进的算法进行了评估，包括57个测试实例和6个实际问题。实证结果表明，CMOEA-TENS通过动态选择最适合的基于cht的种群来驱动进化过程，有效地平衡了探索和开发，同时避免了冗余评价。此外，消融研究进一步验证了设计的MAB策略的有效性。

{"title":"A two-stage ensemble evolutionary algorithm for constrained multi-objective optimization","authors":"Sri Srinivasa Raju Modampuri, Jiahao Fan, Yanan Sun","doi":"10.1016/j.swevo.2025.102213","DOIUrl":"10.1016/j.swevo.2025.102213","url":null,"abstract":"<div><div>In constrained multi-objective evolutionary algorithms (CMOEAs), selecting appropriate constraint-handling techniques (CHTs) is challenging without prior knowledge of the problem’s constraint severity or feasible region distribution. Ensemble frameworks that integrate multiple CHTs with distinct populations offer a promising solution but face issues like redundant evaluations and poor exploration–exploitation balance. To address these limitations, we propose a two-stage ensemble-based CMOEA (CMOEA-TENS) that dynamically prioritizes suitable CHTs based on problem characteristics. Specifically, in the first stage, a population dedicated to explore the unconstrained search space drives the evolutionary process, while remaining populations co-evolve by leveraging solutions identified by the exploratory population. In the second stage, an ensemble of distinct populations drives the evolutionary process, each co-evolving with a different CHT focused on feasibility, diversity, or convergence to exploit the feasible regions effectively. Furthermore, we introduce a novel Multi-Armed Bandit (MAB)-based decision-making strategy that, unlike existing static or random selection approaches, adaptively learns and selects the most suitable CHT-based population to drive the evolutionary process based on real-time performance feedback. This dynamic strategy explicitly reduces redundant functional evaluations and ensures better management of exploration–exploitation trade-offs. CMOEA-TENS was evaluated against eleven state-of-the-art algorithms across six popular test suites, encompassing 57 test instances and six real-world problems. The empirical results demonstrate that CMOEA-TENS effectively balances exploration and exploitation while avoiding redundant evaluations by dynamically selecting the most suitable CHT-based population to drive the evolutionary process. Additionally, an ablation study further validates the effectiveness of the designed MAB strategy.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102213"},"PeriodicalIF":8.5,"publicationDate":"2025-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466376","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}

引用次数: 0

A dynamic multi-objective optimization evolutionary algorithm based on multi-modal feature fusion and entropy-driven reinforcement learning 基于多模态特征融合和熵驱动强化学习的动态多目标优化进化算法

IF 8.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Swarm and Evolutionary Computation

Pub Date : 2025-11-06 DOI: 10.1016/j.swevo.2025.102212

Quanxiu Li, Liang Wang, Yongjie Ma

Dynamic Multi-objective Optimization Problems (DMOPs), where objective functions, constraints, or decision variables change over time, present significant challenges to maintaining both convergence and diversity during the optimization process. Effectively identifying and tracking the optimal solution while balancing the convergence and diversity of the solution set remains the core challenge faced by evolutionary algorithms. To address these challenges, we propose a Dynamic Multi-objective Optimization Evolutionary Algorithm (DMOEA) based on multi-modal feature fusion and Reinforcement Learning (RL). Firstly, a multi-modal feature fusion strategy was designed that integrates Pareto front distribution, decision variable variation, crowding distance, and centroid shift to accurately detect environmental changes and classify their severity. Secondly, a distribution entropy-driven RL strategy is used to dynamically identify diversity-oriented decision variables, and tailored uniformization strategy is applied to increase the diversity. Finally, an adaptive bidirectional search strategy is designed that can perform fine-grained searches on non diverse decision variables in two directions to enhance convergence without sacrificing diversity. Extensive experiments on dynamic test functions demonstrate that our method significantly improves population adaptability, diversity maintenance, and convergence accuracy compared to state-of-the-art DMOEA, offering a promising direction for real-time dynamic optimization in complex environments.

动态多目标优化问题（dops）是指目标函数、约束或决策变量随时间变化而变化的问题，这对优化过程中保持收敛性和多样性提出了重大挑战。有效地识别和跟踪最优解，同时平衡解集的收敛性和多样性是进化算法面临的核心挑战。为了解决这些挑战，我们提出了一种基于多模态特征融合和强化学习（RL）的动态多目标优化进化算法（DMOEA）。首先，设计了一种融合Pareto前沿分布、决策变量变化、拥挤距离和质心偏移的多模态特征融合策略，以准确检测环境变化并对其严重程度进行分类；其次，采用分布熵驱动的强化学习策略动态识别面向多样性的决策变量，并采用定制化统一策略增加多样性；最后，设计了一种自适应双向搜索策略，在不牺牲多样性的前提下，在两个方向上对非多样化决策变量进行细粒度搜索，以提高收敛性。大量的动态测试函数实验表明，与目前最先进的DMOEA相比，我们的方法显著提高了种群适应性、多样性维护和收敛精度，为复杂环境下的实时动态优化提供了一个有希望的方向。

{"title":"A dynamic multi-objective optimization evolutionary algorithm based on multi-modal feature fusion and entropy-driven reinforcement learning","authors":"Quanxiu Li, Liang Wang, Yongjie Ma","doi":"10.1016/j.swevo.2025.102212","DOIUrl":"10.1016/j.swevo.2025.102212","url":null,"abstract":"<div><div>Dynamic Multi-objective Optimization Problems (DMOPs), where objective functions, constraints, or decision variables change over time, present significant challenges to maintaining both convergence and diversity during the optimization process. Effectively identifying and tracking the optimal solution while balancing the convergence and diversity of the solution set remains the core challenge faced by evolutionary algorithms. To address these challenges, we propose a Dynamic Multi-objective Optimization Evolutionary Algorithm (DMOEA) based on multi-modal feature fusion and Reinforcement Learning (RL). Firstly, a multi-modal feature fusion strategy was designed that integrates Pareto front distribution, decision variable variation, crowding distance, and centroid shift to accurately detect environmental changes and classify their severity. Secondly, a distribution entropy-driven RL strategy is used to dynamically identify diversity-oriented decision variables, and tailored uniformization strategy is applied to increase the diversity. Finally, an adaptive bidirectional search strategy is designed that can perform fine-grained searches on non diverse decision variables in two directions to enhance convergence without sacrificing diversity. Extensive experiments on dynamic test functions demonstrate that our method significantly improves population adaptability, diversity maintenance, and convergence accuracy compared to state-of-the-art DMOEA, offering a promising direction for real-time dynamic optimization in complex environments.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102212"},"PeriodicalIF":8.5,"publicationDate":"2025-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466375","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}

引用次数: 0

Stable matching of shared manufacturing resources based on IMPSO algorithm under hesitant fuzzy information 犹豫模糊信息下基于IMPSO算法的共享制造资源稳定匹配

IF 8.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Swarm and Evolutionary Computation

Pub Date : 2025-11-05 DOI: 10.1016/j.swevo.2025.102203

Peng Liu , Xiaochang He , Dongli Cao , Na Yi

The unceasing infiltration of the sharing economy into the field of manufacturing has contributed significantly to the emergence and development of shared manufacturing. Considering the psychological behavioral characteristics such as hesitation, reference dependence, and loss aversion exhibited by resource suppliers and demanders, this paper integrates the bidirectional projection method and TODIM (Tomada de Decisāo Interativa Multicritério) to propose a novel satisfaction calculation framework, which can significantly enhance the expression effect of preference differences. Incorporating the pursuit of matching fairness, this paper innovatively introduces the one-to-many stable matching constraint into the matching model and designs an Improved Particle Swarm Optimization (IMPSO) algorithm to solve the model. Numerical experimental results demonstrate that: (1) In satisfaction calculation, the method proposed in this paper can more effectively amplify preference differences between supply and demand parties towards different matching objects; (2) Regarding matching effect, the proposed method not only ensures the robustness of matching results but also exhibits a certain sensitivity to changes in decision-makers' loss preferences; (3) At the algorithm performance level, the IMPSO algorithm not only overcomes the single-solution limitation of the Gale-Shapley (G-S) algorithm but also significantly enhances convergence efficiency and speed through the introduction of innovative mechanisms such as penalty and reward strategies. These improvements enable it to exhibit better performance in algorithm comparison experiments of different scales, providing effective decision support for resource optimization in shared manufacturing platforms.

共享经济不断向制造业领域渗透，为共享制造的产生和发展做出了重要贡献。考虑到资源供需双方表现出的犹豫、参考依赖、损失厌恶等心理行为特征，本文将双向投影法与TODIM （Tomada de Decisāo Interativa multicritacririo）相结合，提出了一种新的满意度计算框架，该框架可以显著增强偏好差异的表达效果。结合对匹配公平性的追求，在匹配模型中创新性地引入一对多稳定匹配约束，并设计了一种改进的粒子群优化算法（IMPSO）来求解该模型。数值实验结果表明：(1)在满意度计算中，本文提出的方法能更有效地放大供需双方对不同匹配对象的偏好差异；(2)在匹配效果上，该方法既保证了匹配结果的鲁棒性，又对决策者损失偏好的变化具有一定的敏感性；(3)在算法性能层面，IMPSO算法不仅克服了Gale-Shapley （G-S）算法的单解限制，而且通过引入奖惩策略等创新机制，显著提高了收敛效率和收敛速度。这些改进使其在不同尺度的算法对比实验中表现出更好的性能，为共享制造平台的资源优化提供有效的决策支持。

{"title":"Stable matching of shared manufacturing resources based on IMPSO algorithm under hesitant fuzzy information","authors":"Peng Liu , Xiaochang He , Dongli Cao , Na Yi","doi":"10.1016/j.swevo.2025.102203","DOIUrl":"10.1016/j.swevo.2025.102203","url":null,"abstract":"<div><div>The unceasing infiltration of the sharing economy into the field of manufacturing has contributed significantly to the emergence and development of shared manufacturing. Considering the psychological behavioral characteristics such as hesitation, reference dependence, and loss aversion exhibited by resource suppliers and demanders, this paper integrates the bidirectional projection method and TODIM (Tomada de Decisāo Interativa Multicritério) to propose a novel satisfaction calculation framework, which can significantly enhance the expression effect of preference differences. Incorporating the pursuit of matching fairness, this paper innovatively introduces the one-to-many stable matching constraint into the matching model and designs an Improved Particle Swarm Optimization (IMPSO) algorithm to solve the model. Numerical experimental results demonstrate that: (1) In satisfaction calculation, the method proposed in this paper can more effectively amplify preference differences between supply and demand parties towards different matching objects; (2) Regarding matching effect, the proposed method not only ensures the robustness of matching results but also exhibits a certain sensitivity to changes in decision-makers' loss preferences; (3) At the algorithm performance level, the IMPSO algorithm not only overcomes the single-solution limitation of the Gale-Shapley (G-S) algorithm but also significantly enhances convergence efficiency and speed through the introduction of innovative mechanisms such as penalty and reward strategies. These improvements enable it to exhibit better performance in algorithm comparison experiments of different scales, providing effective decision support for resource optimization in shared manufacturing platforms.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102203"},"PeriodicalIF":8.5,"publicationDate":"2025-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466372","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}

引用次数: 0

Optimal predictive load frequency control with multi-objective PID-based search algorithm 基于多目标pid搜索算法的最优预测负荷频率控制

IF 8.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Swarm and Evolutionary Computation

Pub Date : 2025-11-05 DOI: 10.1016/j.swevo.2025.102214

Yang Yang , Yuchao Gao , Jinran Wu , Shangce Gao

Maintaining frequency stability in modern interconnected power systems (IPSs) has become increasingly challenging due to growing system complexity, the integration of variable energy sources, and frequent load fluctuations, particularly in manufacturing and industrial environments where power quality is critical. To address this challenge, we propose an optimal predictive load frequency control (LFC) framework that combines real-time forecasting with multi-objective controller optimization. An online extreme learning machine (ELM) predicts short-term load deviations, enabling proactive regulation. The control structure adopts a cascaded fractional-order design, integrating FOPI and FOPID controllers. These controllers are optimized using a multi-objective PID-based search algorithm (MOPSA), which simultaneously minimizes the integral of time-weighted absolute error (ITAE), enhances the damping ratio, and reduces the cost of energy storage operation. Fast-response energy storage devices (ESDs) are further coordinated to buffer transient imbalances. Simulation results demonstrate that the proposed FOPI–FOPID controllers with ESDs reduce ITAE by 90% (from 464.99 to 44.65) and shorten frequency settling time by 81% (from 83.03 s to 15.49 s), significantly outperforming benchmark methods such as DSA:FOPI–FOPD. These findings confirm the proposed framework’s ability to deliver precise control in dynamic multi-region IPS environments.

在现代互联电力系统（ips）中，由于系统复杂性的增加，可变能源的集成以及频繁的负载波动，特别是在电力质量至关重要的制造业和工业环境中，保持频率稳定性变得越来越具有挑战性。为了解决这一挑战，我们提出了一种将实时预测与多目标控制器优化相结合的最优预测负载频率控制（LFC）框架。在线极限学习机（ELM）预测短期负荷偏差，实现主动调节。控制结构采用级联分数阶设计，将FOPI和FOPID控制器集成在一起。采用基于pid的多目标搜索算法（MOPSA）对控制器进行优化，使时间加权绝对误差（ITAE）积分最小化，提高阻尼比，降低储能运行成本。快速响应储能装置（ESDs）进一步协调以缓冲瞬态不平衡。仿真结果表明，基于ESDs的FOPI-FOPID控制器的ITAE降低了90%（从464.99降低到44.65），频率稳定时间缩短了81%（从83.03 s降低到15.49 s），显著优于DSA: FOPI-FOPD等基准方法。这些发现证实了所提出的框架在动态多区域IPS环境中提供精确控制的能力。

{"title":"Optimal predictive load frequency control with multi-objective PID-based search algorithm","authors":"Yang Yang , Yuchao Gao , Jinran Wu , Shangce Gao","doi":"10.1016/j.swevo.2025.102214","DOIUrl":"10.1016/j.swevo.2025.102214","url":null,"abstract":"<div><div>Maintaining frequency stability in modern interconnected power systems (IPSs) has become increasingly challenging due to growing system complexity, the integration of variable energy sources, and frequent load fluctuations, particularly in manufacturing and industrial environments where power quality is critical. To address this challenge, we propose an optimal predictive load frequency control (LFC) framework that combines real-time forecasting with multi-objective controller optimization. An online extreme learning machine (ELM) predicts short-term load deviations, enabling proactive regulation. The control structure adopts a cascaded fractional-order design, integrating FOPI and FOPID controllers. These controllers are optimized using a multi-objective PID-based search algorithm (MOPSA), which simultaneously minimizes the integral of time-weighted absolute error (ITAE), enhances the damping ratio, and reduces the cost of energy storage operation. Fast-response energy storage devices (ESDs) are further coordinated to buffer transient imbalances. Simulation results demonstrate that the proposed FOPI–FOPID controllers with ESDs reduce ITAE by 90% (from 464.99 to 44.65) and shorten frequency settling time by 81% (from 83.03 s to 15.49 s), significantly outperforming benchmark methods such as DSA:FOPI–FOPD. These findings confirm the proposed framework’s ability to deliver precise control in dynamic multi-region IPS environments.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102214"},"PeriodicalIF":8.5,"publicationDate":"2025-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466374","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}

引用次数: 0

A hybrid algorithm incorporating sequencing flexibility for integrated process planning and scheduling problem 综合工艺规划与调度问题的一种具有排序灵活性的混合算法

IF 8.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Swarm and Evolutionary Computation

Pub Date : 2025-11-05 DOI: 10.1016/j.swevo.2025.102201

Zixin Peng , Yi Kang , Xinyu Li , Liang Gao , Qihao Liu , Chunjiang Zhang

Integrated process planning and scheduling (IPPS) is important for enhancing the efficiency of manufacturing systems. However, simultaneously considering both process planning and scheduling significantly increases the complexity of IPPS, making it an even more challenging NP-hard problem. Although there are many advanced algorithms, their integration with the characteristics of the problem could be improved, especially in decoding method and neighborhood structure. This paper mainly considers the sequencing flexibility of IPPS and proposes an improved hybrid genetic and tabu search algorithm. Firstly, a newly active decoding method incorporating sequencing flexibility is proposed. Each operation is scheduled at its earliest feasible time within the constraints of the network graph, rather than being restricted by a fixed encoding scheme. This reduces unnecessary constraints and improves solution quality. Secondly, two neighborhood structures considering sequencing flexibility are designed: N8-F and NK-F during the local search phase. The improved neighborhood structures expand the solution space while ensuring the feasibility of neighboring solutions. In the experimental part, in addition to 37 benchmark problems, 50 complex instances are scientifically designed due to the simplicity of existing problems. The results show that the proposed algorithm is more efficient in obtaining the optimal solution for the existing problems, and it also outperforms other comparative algorithms on the newly designed instances.

集成工艺规划与调度（IPPS）是提高制造系统效率的重要手段。然而，同时考虑过程规划和调度会大大增加IPPS的复杂性，使其成为一个更具挑战性的NP-hard问题。虽然有许多先进的算法，但它们与问题特点的结合还有待改进，特别是在解码方法和邻域结构方面。本文主要考虑了IPPS的排序灵活性，提出了一种改进的遗传和禁忌搜索混合算法。首先，提出了一种具有序列灵活性的主动解码方法。在网络图的约束下，将每个操作调度到最早可行的时间，而不受固定编码方案的限制。这减少了不必要的约束并提高了解决方案的质量。其次，在局部搜索阶段设计了考虑排序灵活性的N8-F和NK-F两个邻域结构；改进的邻域结构在保证邻域解可行性的同时扩大了解空间。在实验部分，由于现有问题的简单性，除了37个基准问题外，还科学设计了50个复杂实例。结果表明，本文提出的算法对现有问题的最优解求解效率更高，并且在新设计的实例上优于其他比较算法。

{"title":"A hybrid algorithm incorporating sequencing flexibility for integrated process planning and scheduling problem","authors":"Zixin Peng , Yi Kang , Xinyu Li , Liang Gao , Qihao Liu , Chunjiang Zhang","doi":"10.1016/j.swevo.2025.102201","DOIUrl":"10.1016/j.swevo.2025.102201","url":null,"abstract":"<div><div>Integrated process planning and scheduling (IPPS) is important for enhancing the efficiency of manufacturing systems. However, simultaneously considering both process planning and scheduling significantly increases the complexity of IPPS, making it an even more challenging NP-hard problem. Although there are many advanced algorithms, their integration with the characteristics of the problem could be improved, especially in decoding method and neighborhood structure. This paper mainly considers the sequencing flexibility of IPPS and proposes an improved hybrid genetic and tabu search algorithm. Firstly, a newly active decoding method incorporating sequencing flexibility is proposed. Each operation is scheduled at its earliest feasible time within the constraints of the network graph, rather than being restricted by a fixed encoding scheme. This reduces unnecessary constraints and improves solution quality. Secondly, two neighborhood structures considering sequencing flexibility are designed: N8-F and NK-F during the local search phase. The improved neighborhood structures expand the solution space while ensuring the feasibility of neighboring solutions. In the experimental part, in addition to 37 benchmark problems, 50 complex instances are scientifically designed due to the simplicity of existing problems. The results show that the proposed algorithm is more efficient in obtaining the optimal solution for the existing problems, and it also outperforms other comparative algorithms on the newly designed instances.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102201"},"PeriodicalIF":8.5,"publicationDate":"2025-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466373","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}

引用次数: 0

Dynamic strategy-based hybrid genetic algorithm for solving multi-task decision-making problems for heterogeneous surface vessels 基于动态策略的混合遗传算法求解异构水面舰艇多任务决策问题

IF 8.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Swarm and Evolutionary Computation

Pub Date : 2025-11-03 DOI: 10.1016/j.swevo.2025.102206

Yutong Li , Yufeng Chen , Rui Zhou , Zhiwu Li

The rapid advancement of marine robotics, especially marine unmanned surface vessels, has revolutionized their use in maritime missions ranging from military operations to civilian tasks. This development is critical for sustainable marine assignments, decreasing costs, enhancing surface vessel collaboration flexibility, and improving safety. The control of multiple vessel formations for efficiently performing specific tasks is a significant achievement in a complex marine environment. For huge task domains and a heterogeneous fleet, a significant optimization challenge is to enhance the efficiency of cooperative heterogeneous surface vessels to accomplish multiple tasks and satisfy task demands under limited temporal intervals and vessel operation capabilities. In this paper, a mathematical framework of the heterogeneous surface vessel multi-task decision-making problem is proposed. Several dynamic algorithms are adopted to optimize the population distribution of the genetic algorithm. A 2-opt algorithm is adopted to construct the hybrid genetic algorithm. A novel dynamic hybrid genetic algorithm containing the dynamic algorithm and the 2-opt algorithm is developed, which can improve local search ability and enhance computational efficiency. Compared with several common evolutionary algorithms, hybrid evolutionary algorithms and other latest state-of-the-art (SOTA) algorithms, the dynamic hybrid genetic algorithm can achieve a better route sequence and shorter sailing time for this multi-task decision-making problem.

海洋机器人技术的快速发展，特别是海洋无人水面舰艇，已经彻底改变了它们在从军事行动到民用任务的海上任务中的应用。这一发展对于可持续的海上作业、降低成本、增强水面船舶协作灵活性和提高安全性至关重要。在复杂的海洋环境中，控制多个船舶编队以有效执行特定任务是一项重要的成就。对于庞大的任务域和异构舰队，如何在有限的时间间隔和船舶操作能力下，提高协同异构水面舰艇完成多任务和满足任务需求的效率是一个重要的优化挑战。本文提出了异构水面舰艇多任务决策问题的数学框架。采用了几种动态算法来优化遗传算法的种群分布。采用2-opt算法构造混合遗传算法。提出了一种包含动态算法和2-opt算法的动态混合遗传算法，提高了局部搜索能力，提高了计算效率。与几种常见的进化算法、混合进化算法和其他SOTA算法相比，动态混合遗传算法可以获得更好的路线序列和更短的航行时间。

{"title":"Dynamic strategy-based hybrid genetic algorithm for solving multi-task decision-making problems for heterogeneous surface vessels","authors":"Yutong Li , Yufeng Chen , Rui Zhou , Zhiwu Li","doi":"10.1016/j.swevo.2025.102206","DOIUrl":"10.1016/j.swevo.2025.102206","url":null,"abstract":"<div><div>The rapid advancement of marine robotics, especially marine unmanned surface vessels, has revolutionized their use in maritime missions ranging from military operations to civilian tasks. This development is critical for sustainable marine assignments, decreasing costs, enhancing surface vessel collaboration flexibility, and improving safety. The control of multiple vessel formations for efficiently performing specific tasks is a significant achievement in a complex marine environment. For huge task domains and a heterogeneous fleet, a significant optimization challenge is to enhance the efficiency of cooperative heterogeneous surface vessels to accomplish multiple tasks and satisfy task demands under limited temporal intervals and vessel operation capabilities. In this paper, a mathematical framework of the heterogeneous surface vessel multi-task decision-making problem is proposed. Several dynamic algorithms are adopted to optimize the population distribution of the genetic algorithm. A 2-opt algorithm is adopted to construct the hybrid genetic algorithm. A novel dynamic hybrid genetic algorithm containing the dynamic algorithm and the 2-opt algorithm is developed, which can improve local search ability and enhance computational efficiency. Compared with several common evolutionary algorithms, hybrid evolutionary algorithms and other latest state-of-the-art (SOTA) algorithms, the dynamic hybrid genetic algorithm can achieve a better route sequence and shorter sailing time for this multi-task decision-making problem.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102206"},"PeriodicalIF":8.5,"publicationDate":"2025-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466371","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}

引用次数: 0

GMO: A general multimodal optimization framework applicable to various global metaheuristic algorithms GMO：一个通用的多模态优化框架，适用于各种全局元启发式算法

IF 8.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Swarm and Evolutionary Computation

Pub Date : 2025-11-03 DOI: 10.1016/j.swevo.2025.102207

Hui Ren, Luli Gao, Zhibin Su, Yujian Jiang

This paper proposes a general multimodal optimization (GMO) framework that enables metaheuristic algorithms (MAs), originally designed for single-convergence global optimization, to effectively solve multimodal optimization (MMO) problems without altering their structures. Most existing MMO algorithms incorporate mechanisms tailored to individual MAs, leading to high specificity and limited transferability. Furthermore, the No Free Lunch theorem indicates that no single algorithm performs best across all problems. Inspired by this, GMO provides a broadly applicable and flexible approach, capable of integrating with diverse MAs to address MMO tasks efficiently. It comprises three core components: (1) a multi-subpopulation competitive (MPC) strategy that enhances exploration capability by maintaining population diversity through competition between dominant individuals; (2) an archive elite refinement (AER) strategy that improves exploitation capability by re-optimizing convergent individuals, and by archiving the convergent information, it enables the reuse of these converged individuals to increase the likelihood of discovering more global optima; (3) a fitness landscape reconstruction (FLC) strategy that prevents repeated access to the same peak by dynamically updating the search landscape based on the archive, thereby improving search efficiency. Finally, nine MAs were integrated with GMO and tested against several MMO algorithms. Results show that GMO can effectively integrate with various MAs to achieve MMO, without requiring algorithm-specific modifications. It demonstrates broad applicability and flexibility, enabling efficient MMO methods tailored to specific problems. Moreover, performance comparisons with other MMO algorithms reveal that GMO-MAs achieve improved performance on the CEC2013 multimodal benchmark functions and are capable of obtaining global optima with higher accuracy.

本文提出了一种通用的多模态优化（GMO）框架，使最初设计用于单收敛全局优化的元启发式算法（MAs）能够在不改变其结构的情况下有效地解决多模态优化（MMO）问题。大多数现有的MMO算法都包含针对单个MAs的机制，导致高特异性和有限的可移植性。此外，没有免费的午餐定理表明，没有一个算法在所有问题中表现最好。受此启发，GMO提供了一种广泛适用且灵活的方法，能够与不同的MAs集成以有效地解决MMO任务。它包括三个核心部分：(1)多亚种群竞争策略，通过优势个体之间的竞争来保持种群多样性，从而提高探索能力；(2)归档精英细化（AER）策略，通过对收敛个体进行再优化来提高开发能力，并通过对收敛信息进行归档，使这些收敛个体能够重用，从而增加发现更多全局最优解的可能性；(3)适应度景观重建（FLC）策略，通过基于存档动态更新搜索景观，防止重复访问同一峰值，从而提高搜索效率。最后，将9个MAs与GMO集成，并针对几种MMO算法进行了测试。结果表明，GMO可以有效地与各种MAs集成以实现MMO，而无需对算法进行特定修改。它展示了广泛的适用性和灵活性，使有效的MMO方法能够针对特定问题进行定制。此外，与其他MMO算法的性能比较表明，GMO-MAs在CEC2013多模态基准函数上取得了更好的性能，能够以更高的精度获得全局最优解。

{"title":"GMO: A general multimodal optimization framework applicable to various global metaheuristic algorithms","authors":"Hui Ren, Luli Gao, Zhibin Su, Yujian Jiang","doi":"10.1016/j.swevo.2025.102207","DOIUrl":"10.1016/j.swevo.2025.102207","url":null,"abstract":"<div><div>This paper proposes a general multimodal optimization (GMO) framework that enables metaheuristic algorithms (MAs), originally designed for single-convergence global optimization, to effectively solve multimodal optimization (MMO) problems without altering their structures. Most existing MMO algorithms incorporate mechanisms tailored to individual MAs, leading to high specificity and limited transferability. Furthermore, the No Free Lunch theorem indicates that no single algorithm performs best across all problems. Inspired by this, GMO provides a broadly applicable and flexible approach, capable of integrating with diverse MAs to address MMO tasks efficiently. It comprises three core components: (1) a multi-subpopulation competitive (MPC) strategy that enhances exploration capability by maintaining population diversity through competition between dominant individuals; (2) an archive elite refinement (AER) strategy that improves exploitation capability by re-optimizing convergent individuals, and by archiving the convergent information, it enables the reuse of these converged individuals to increase the likelihood of discovering more global optima; (3) a fitness landscape reconstruction (FLC) strategy that prevents repeated access to the same peak by dynamically updating the search landscape based on the archive, thereby improving search efficiency. Finally, nine MAs were integrated with GMO and tested against several MMO algorithms. Results show that GMO can effectively integrate with various MAs to achieve MMO, without requiring algorithm-specific modifications. It demonstrates broad applicability and flexibility, enabling efficient MMO methods tailored to specific problems. Moreover, performance comparisons with other MMO algorithms reveal that GMO-MAs achieve improved performance on the CEC2013 multimodal benchmark functions and are capable of obtaining global optima with higher accuracy.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102207"},"PeriodicalIF":8.5,"publicationDate":"2025-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466369","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}

引用次数: 0

A framework for enhancing simulation reliability in hard-sphere packing: Surrogate model comparison, multi-objective optimization, and multi-criteria decision making for robust parameter identification 增强硬球填料仿真可靠性的框架：代理模型比较、多目标优化和鲁棒参数识别的多准则决策

IF 8.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Swarm and Evolutionary Computation

Pub Date : 2025-11-02 DOI: 10.1016/j.swevo.2025.102205

Xiangling Ma , Weichen Zhao , Jianjun Liang , Gaofeng Yin , Weiheng Ou

Simulations of high-fidelity hard-sphere packings are useful for studying particulate systems; however, the cost of computing and the inherent sensitivity in the simulations can be prohibitive. We develop a robust framework that leverages surrogate-assisted modeling, simulators with varying fidelity, multi-objective optimization, and multi-criteria decision-making (MCDM) to improve the reliability of the simulation while maintaining the efficiency of the simulation. XGBoost surrogate models created for predicting failure probabilities for FBA and LS simulations, using TPE algorithm for hyperparameter tuning in Optuna, were able to make accurate predictions, yielding an R² of 0.9270 ± 0.0038 and 0.9535 ± 0.0013 for FBA and LS respectively, while also maintaining a low MAE and RMSE. The use of AGE-MOEA lead to the generation of clearly defined Pareto fronts whose L shaped trade-off was between minimizing failure probability for each method (FBA and LS). The use of TOPSIS identified compromise solutions to practical parameter selection and use. The analysis of the decision space outlined the significance of scaling factors and particle counts for stability, while also demonstrating the degree of freedom for particle ratios and polydispersity, as the assurance of stability takes precedence over reliability. The framework we have proposed is a significant step in evolving the state of the art in multi-dimensional understanding through the integrated use of surrogate modeling, systematic and multi-fidelity evaluations, and systematic decision-making strategies based on performance and reliability. Together, we offer reliable and computationally efficient strategies for optimizing the parameter selection that is associated with particulate simulations.

高保真硬球填料的模拟对研究颗粒系统是有用的；然而，在模拟中计算的成本和固有的敏感性可能是令人望而却步的。我们开发了一个强大的框架，利用代理辅助建模，具有不同保真度的模拟器，多目标优化和多标准决策（MCDM）来提高仿真的可靠性，同时保持仿真的效率。XGBoost代理模型用于预测FBA和LS模拟的故障概率，在Optuna中使用TPE算法进行超参数调优，能够做出准确的预测，FBA和LS的R²分别为0.9270±0.0038和0.9535±0.0013，同时保持较低的MAE和RMSE。使用AGE-MOEA可以生成明确定义的Pareto前沿，其L形权衡是在每种方法（FBA和LS）的最小失效概率之间进行的。TOPSIS的使用确定了实际参数选择和使用的折衷解决方案。决策空间的分析概述了尺度因子和粒子计数对稳定性的重要性，同时也展示了粒子比率和多分散性的自由度，因为稳定性的保证优先于可靠性。我们提出的框架是通过综合使用代理建模、系统和多保真度评估以及基于性能和可靠性的系统决策策略，在多维理解方面发展最新技术的重要一步。总之，我们提供可靠和计算效率的策略，以优化与颗粒模拟相关的参数选择。

{"title":"A framework for enhancing simulation reliability in hard-sphere packing: Surrogate model comparison, multi-objective optimization, and multi-criteria decision making for robust parameter identification","authors":"Xiangling Ma , Weichen Zhao , Jianjun Liang , Gaofeng Yin , Weiheng Ou","doi":"10.1016/j.swevo.2025.102205","DOIUrl":"10.1016/j.swevo.2025.102205","url":null,"abstract":"<div><div>Simulations of high-fidelity hard-sphere packings are useful for studying particulate systems; however, the cost of computing and the inherent sensitivity in the simulations can be prohibitive. We develop a robust framework that leverages surrogate-assisted modeling, simulators with varying fidelity, multi-objective optimization, and multi-criteria decision-making (MCDM) to improve the reliability of the simulation while maintaining the efficiency of the simulation. XGBoost surrogate models created for predicting failure probabilities for FBA and LS simulations, using TPE algorithm for hyperparameter tuning in Optuna, were able to make accurate predictions, yielding an R² of 0.9270 ± 0.0038 and 0.9535 ± 0.0013 for FBA and LS respectively, while also maintaining a low MAE and RMSE. The use of AGE-MOEA lead to the generation of clearly defined Pareto fronts whose L shaped trade-off was between minimizing failure probability for each method (FBA and LS). The use of TOPSIS identified compromise solutions to practical parameter selection and use. The analysis of the decision space outlined the significance of scaling factors and particle counts for stability, while also demonstrating the degree of freedom for particle ratios and polydispersity, as the assurance of stability takes precedence over reliability. The framework we have proposed is a significant step in evolving the state of the art in multi-dimensional understanding through the integrated use of surrogate modeling, systematic and multi-fidelity evaluations, and systematic decision-making strategies based on performance and reliability. Together, we offer reliable and computationally efficient strategies for optimizing the parameter selection that is associated with particulate simulations.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102205"},"PeriodicalIF":8.5,"publicationDate":"2025-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466368","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}

引用次数: 0

A new dual-population evolutionary algorithm leveraging objective-constraint relationships for constrained multi-objective optimization 基于目标约束关系的约束多目标优化双种群进化算法

IF 8.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Swarm and Evolutionary Computation

Pub Date : 2025-11-01 DOI: 10.1016/j.swevo.2025.102152

Jialu Ye , Chaogui Tan , Yizhang Xia , Zhanglu Hou , Yuan Liu , Juan Zou

Population co-evolution strategies are widely used to handle constrained multi-objective optimization problems (CMOPs). However, existing coevolutionary algorithms oversimplify population collaboration and are rigid in evolving the auxiliary population, ignoring the importance of maintaining population diversity, especially for CMOPs with irregular feasible regions. To tackle this issue, we propose a new dual-population evolutionary algorithm, denoted as AACMO, which attempts to mine the relational features between the unconstrained Pareto front (UPF) and the constrained Pareto front (CPF) to better adapt to CMOPs with various constraints. Specifically, the method is categorized into two phases: a learning phase and an evolving phase. First, the main population (MP) aims to quickly learn the relationships between the UPF and the CPF by leveraging the advantages of different operators, namely, genetic algorithm operators and differential evolution operators, during the learning phase. Second, two auxiliary strategies (positive co-evolution and inverse co-evolution) are designed to collaborate with the MP during the evolving phase. These strategies are capable of employing a more effective auxiliary co-evolution method to assist the MP in maintaining diversity, based on the degree of alignment between the learned UPF and CPF at the end of the learning phase. The effectiveness of AACMO is confirmed through comparative analysis with eleven state-of-the-art algorithms across four CMOP benchmark suites.

种群协同进化策略被广泛应用于求解约束多目标优化问题。然而，现有的协同进化算法过于简化种群协作，对辅助种群的进化过于死板，忽视了保持种群多样性的重要性，特别是对于具有不规则可行区域的cops。为了解决这一问题，我们提出了一种新的双种群进化算法，称为AACMO，该算法试图挖掘无约束帕累托前沿（UPF）和约束帕累托前沿（CPF）之间的关系特征，以更好地适应具有各种约束的cops。具体来说，该方法分为两个阶段：学习阶段和发展阶段。首先，主种群（MP）的目标是在学习阶段利用不同算子的优势，即遗传算法算子和差分进化算子，快速学习UPF和CPF之间的关系。其次，设计了两种辅助策略（正协同进化和逆协同进化），在进化阶段与mps协同工作。这些策略能够采用一种更有效的辅助协同进化方法，根据学习到的UPF和CPF在学习阶段结束时的一致性程度，来帮助MP维持多样性。通过与四个CMOP基准套件中的11种最先进算法的比较分析，证实了AACMO的有效性。

{"title":"A new dual-population evolutionary algorithm leveraging objective-constraint relationships for constrained multi-objective optimization","authors":"Jialu Ye , Chaogui Tan , Yizhang Xia , Zhanglu Hou , Yuan Liu , Juan Zou","doi":"10.1016/j.swevo.2025.102152","DOIUrl":"10.1016/j.swevo.2025.102152","url":null,"abstract":"<div><div>Population co-evolution strategies are widely used to handle constrained multi-objective optimization problems (CMOPs). However, existing coevolutionary algorithms oversimplify population collaboration and are rigid in evolving the auxiliary population, ignoring the importance of maintaining population diversity, especially for CMOPs with irregular feasible regions. To tackle this issue, we propose a new dual-population evolutionary algorithm, denoted as AACMO, which attempts to mine the relational features between the unconstrained Pareto front (UPF) and the constrained Pareto front (CPF) to better adapt to CMOPs with various constraints. Specifically, the method is categorized into two phases: a learning phase and an evolving phase. First, the main population (MP) aims to quickly learn the relationships between the UPF and the CPF by leveraging the advantages of different operators, namely, genetic algorithm operators and differential evolution operators, during the learning phase. Second, two auxiliary strategies (positive co-evolution and inverse co-evolution) are designed to collaborate with the MP during the evolving phase. These strategies are capable of employing a more effective auxiliary co-evolution method to assist the MP in maintaining diversity, based on the degree of alignment between the learned UPF and CPF at the end of the learning phase. The effectiveness of AACMO is confirmed through comparative analysis with eleven state-of-the-art algorithms across four CMOP benchmark suites.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"99 ","pages":"Article 102152"},"PeriodicalIF":8.5,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466370","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}

引用次数: 0