Swarm and Evolutionary Computation最新文献

The efficacy of surrogate-assisted multi-objective evolutionary algorithms (SAMOEAs) in addressing expensive multi-objective optimization problems (MOPs) is contingent upon the modeling techniques and model-based infill sampling strategies. In addressing this pivotal aspect, this paper introduces a pioneering methodology known as batch subproblem coevolution with Gaussian process-driven linear models (BSCo-GPLM). Specifically, from a modeling perspective, BSCo-GPLM decomposes the MOP into single-objective subproblems. Following this decomposition, for each subproblem, a Gaussian process-driven linear model (GPLM) is collaboratively trained to prevent overfitting and improve prediction accuracy. Regarding infill sampling, collaborative optimization of all GPLMs yields optimal candidate solutions for each subproblem, organized into coherent clusters. Within each cluster, only the solution with the highest utility is evaluated. Relying on the heightened prediction accuracy of the GPLM model and an efficient batch sampling strategy, BSCo-GPLM exhibits clear superiority over state-of-the-art SAMOEAs in effectively addressing expensive MOPs. The source code of BSCo-GPLM is available at https://github.com/CIAM-Group/EvolutionaryAlgorithm_Codes/tree/main/BSCo-GPLM.

In recent years, an increasing number of prediction-based strategies have shown promising results in handling dynamic multi-objective optimization problems (DMOPs), and prediction models are also considered to be very favorable. Nevertheless, some linear prediction models may not always be effective. In particular, when the motion direction trends of different individuals are not aligned, these models can yield inaccurate prediction results. Inverted generational distance (IGD) is a commonly used metric for evaluating the performance of algorithms. This paper proposes a prediction model based on the IGD metric. Specifically, we assume that the pareto optimal front (POF) of the population at the previous time step is the true POF, and the POF at the current time step is the approximate POF. We cluster the current population with reference to the euclidean distances from uniform points on the true POF to the current POF points, with slight overlap between adjacent clusters, enables a better tradeoff between convergence and diversity in the prediction process. We consider the movement directions of individuals within each cluster separately through different cluster distributions, while balancing the individual movement directions and the overall population movement direction by overlaying cluster coverage areas, thereby helping to avoid the clustered prediction population from getting trapped in local optima. Experimental results and comparisons with other algorithms demonstrate that this strategy exhibits strong competitiveness in handling DMOPs.

Multi-objective optimization problems (MOPs) involve optimizing multiple conflicting objectives simultaneously, resulting in a set of Pareto optimal solutions. Due to the high computational or financial cost associated with evaluating fitness, expensive multi-objective optimization problems (EMOPs) further complicate the optimization process. Surrogate-assisted evolutionary algorithms (SAEAs) have emerged as a promising approach to address EMOPs by substituting costly evaluations with computationally efficient surrogate models. This paper introduces the self-organizing surrogate-assisted non-dominated sorting differential evolution (SSDE), which uses surrogate model based on a self-organizing map (SOM) to approximate the fitness function. SSDE offers advantages such as reduced computational cost, improved accuracy, and the speed of enhanced convergence. The SOM-based surrogate models effectively capture the underlying structure of the Pareto optimal set and Pareto optimal front, leading to superior approximations of the fitness function. Experimental results on benchmark functions and real-world problems, including Model-Free Adaptive Control (MFAC) and the Yagi-Uda Antenna design, demonstrate the competitiveness and efficiency of SSDE compared to other algorithms.

Handling constrained multimodal multi-objective optimization problems (CMMOPs) is a tremendous challenge as it involves the discovery of multiple equivalent constrained Pareto sets (CPSs) with the identical constrained Pareto front (CPF). However, the existing constrained multi-objective evolutionary algorithms are rarely suitable for solving CMMOPs due to the fact that they focus solely on locating CPF and do not intend to search for multiple equivalent CPSs. To address this issue, this paper proposes a framework of clustering-based constrained multimodal multi-objective evolutionary algorithm, termed FCCMMEA. In the proposed FCCMMEA, we adopt a clustering method to separate the population into multiple subpopulations for locating diverse CPSs and maintaining population diversity. Subsequently, each subpopulation evolves independently to produce offspring by an evolutionary algorithm. To balance the convergence and feasibility, we develop a quality evaluation metric in the classification strategy that considers the local convergence quality and constraint violation values, and it divides the populations into superior and inferior populations according to the quality evaluation of individuals. Furthermore, we also employ a diversity maintenance methodology in environmental selection to maintain the diverse population. The proposed FCCMMEA algorithm is compared with seven state-of-the-art competing algorithms on a standard CMMOP test suite, and the experimental results validate that the proposed FCCMMEA enables to find multiple CPSs and is suitable for handling CMMOPs. Also, the proposed FCCMMEA won the first place in the 2023 IEEE Congress on Evolutionary Computation competition on CMMOPs.

This paper tackles the challenges of computation offloading in the cloud–edge paradigm. Although many solutions exist for enhancing the server’s computational and communication efficiency, they mainly focus on reducing latency and often neglect the impact of overlapping multi-request processing on scheduling reliability. Additionally, these approaches do not account for the preemptive characteristics of applications running in the VMs that lead to higher energy consumption. We propose a novel hybrid integer multi-objective dynamic decision-making approach enhanced with the gravity reference point method. This method determines the proportion of computations executed on cloud servers versus those handled locally on edge servers. Our hybrid approach leverages the gravitational potential reference point and crowding degrees to improve the characteristics of whale populations, addressing the limitations of the traditional whale algorithm, which depends on individual whales’ varying foraging behaviors influenced by a random probability number. By evaluating the crowding level around the prey, the foraging behavior of individual whales is adjusted to enhance the algorithm’s convergence speed and optimization accuracy, thereby increasing its reliability. The results show that our hybrid computation offloading model significantly improves time latency by 76.45%, energy efficiency by 63.12%, reliability by 82%, quality of service by 83.78%, distributor throughput by 87.31%, asset availability by 73.05%, and guarantee ratio by 89.72% compared to traditional offloading methods.

In the domain of just-in-time permutation flowshop scheduling, most studies typically assume that all jobs either have their own soft due date or none of them do. However, in practice, scheduling a combination of hard and soft due date jobs, particularly with the context of emergency order insertion, remains a significant research topic. This paper addresses a constrained permutation flowshop scheduling problem with a mix of hard and soft due date jobs under total weighted tardiness criterion (CPFSP-TWT). We establish a mathematical model and propose an effective Two-Stage Iterated Greedy (ETSIG) algorithm tailored to the problem's characteristics, incorporating a two-stage constructive heuristic to generate a high-quality initial solution. We introduce problem-specific acceleration mechanisms based on position-bound considerations to enhance operational efficiency. We propose three knowledge-based repair strategies for handling infeasible solutions, along with a dynamic self-adjustment mechanism. Additionally, three efficient local search procedures integrate several specific perturbation operators to balance algorithmic exploitation and exploration abilities. Experimental evaluations affirm ETSIG's superiority over five state-of-the-art metaheuristics from closely related literature, establishing its efficacy in addressing CPFSP-TWT.

In recent years, the distributed hybrid flowshop scheduling problem (DHFSP) has garnered widespread attention due to the continuous emergence of practical challenges. The production model, characterized by multiple varieties and small batches, is widely observed in the industrial sector. Additionally, in various real-world scenarios, batches often undergo repeated processes across multiple stages. This paper addresses the research gap by introducing the reentrant nature of batches and the heterogeneity of factories into the DHFSP, resulting in a novel problem referred to as the distributed reentrant heterogeneous hybrid flowshop batch scheduling problem (DRHHFBSP). To tackle this problem, we propose a mixed-integer linear programming (MILP) model. Given that this problem falls into the NP-hard category, an iterative construction-local search-reconstruction algorithm (ICLSRA) is designed. Specifically designed by incorporating construction, local search, and reconstruction processes that have different roles, this algorithm strikes a balance between local and global search. Comparative analysis with the MILP model and state-of-the-art algorithms demonstrates the superiority of ICLSRA in achieving efficient solutions for the DRHHFBSP.

Differential Evolution (DE), as a population-based meta-heuristic global optimization technique, has shown excellent performance in handling optimization problems in continuous spaces. Despite its effectiveness, the DE algorithm suffers from shortcomings such as complexity of parameter selection and limitations of the mutation strategy. Therefore, this paper presents a new strategy for generating trial vectors based on a hierarchical archive, which integrates promising information during evolution with current populations to obtain a good perception of the objective landscape. Moreover, to mitigate mis-scaling by scale factor, an adaptive parameter generation mechanism with hierarchical selection (APSH) is proposed. Furthermore, a novel population diversity metric technique and a restart mechanism based on wavelet functions is introduced in this paper. Comparative experiments were conducted to evaluate the performance of the proposed algorithm using 100 benchmark functions from the CEC2013, CEC2014, CEC2017, and CEC2022 test suites. The results demonstrate that the HAPI-DE algorithm outperforms or is on par with 6 recent powerful DE variants. Additionally, HAPI-DE was utilized in parameter extraction for the photovoltaic model STP6-120/36. The findings suggest that our algorithm, HAPI-DE, demonstrates competitiveness when compared to the 6 other DE variants.

A multimodal multiobjective optimization problem can have multiple equivalent Pareto Sets (PSs). Since the number of PSs may vary in different problems, if the population is restricted to a fixed size, the number of solutions found for each PS will inevitably fluctuate widely, which is undesirable for decision makers. To address the issue, this paper proposes a clustering-assisted adaptive evolutionary algorithm based on decomposition (CA-MMEA/D), whose search process can be roughly divided into two stages. In the first stage, an initial exploration of decision space is carried out, and then solutions with good convergence are used for clustering to estimate the number and location of multiple PSs. In the second stage, new search strategies are developed on the basis of clustering, which can take advantage of unimodal search methods. Experimental studies show that the proposed algorithm outperforms some state-of-the-art algorithms, and CA-MMEA/D can keep the number of solutions found for each PS at a relatively stable level for different problems, thus making it easier for decision makers to choose the desired solutions. The research in this paper provides new ideas for the design of decomposition-based multimodal multiobjective algorithms.

Drones are increasingly utilized for transportation reconnaissance due to their expansive field of view, cost-effectiveness, and agility. They can pre-reconnaissance the traveling routes of the ground vehicles carrying valuable goods to ensure the safety of vehicles and their goods. This rises a novel routing problem for the Vehicles and their Pre-Reconnaissance Drones, which is an integrated point and arc routing problem with temporal and spatial synchronization. A mixed integer linear programming model is developed to formulate the problem with complex synchronization constraints. The Variable Neighborhood Search algorithm integrated with the Temporal & Spatial synchronization-based Greedy search and simulated annealing strategy is designed to solve the model. A practical case based on real urban data from Beijing, China, and random instances with different sizes are tested and compared with the proposed algorithm. Computational results indicated that the proposed algorithm can solve the problem efficiently and outperform the simulated annealing algorithm and the greedy algorithm.