Abstract:Multimodal optimization problems (MMOPs) aim to locate multiple global optima simultaneously, which requires a high diversity of solutions. As a diversity preservation strategy customized for evolutionary algorithms (EAs), niching has been extensively incorporated into many algorithms to deal with MMOPs. However, many existing niching techniques require pre-defined parameters to control the size or number of niches, making the algorithm sensitive to parameters. Taking individuals as nodes, the inverse distance of two individuals as edges, and the historical information of recent generations as node attributes, we can construct a population as a network, whose purpose is to divide niches through community detection without relying on sensitive parameters. Inspired by this idea, this paper proposes a network community-based differential evolution for multimodal optimization problems (NetCDEMMOPs). In NetCDEMMOPs , three strategies are proposed and work together. Firstly, the community detection-based niching (CDN) strategy constructs the population as an attribute network and automatically divides it into multiple niches without sensitive parameters. Secondly, the community elites-based updating (CEU) strategy allocates more evolutionary resources to elite individuals of communities, then predicts and updates these elites based on historical information. Finally, the poor individual remolding (PIR) strategy remolds the last-ranking individuals and guides them toward promising positions. After extensive experimentation with the widely adopted CEC'2013 test suit, the experimental results reveal the superiority of NetCDEMMOPs over many successful or recent algorithms for MMOPs.

Adaptive Clustering-Based Differential Evolution for Multimodal Optimization

Multimodal Multiobjective Differential Evolution Algorithm Based on Enhanced Decision Space Search

Differential Evolution with Adaptive Subpopulation Size for Multimodal Optimization Problems

A Cluster-Based Differential Evolution with Self-Adaptive Strategy for Multimodal Optimization

Adaptive Estimation Distribution Distributed Differential Evolution for Multimodal Optimization Problems

Differential Evolution for Multimodal Optimization with Species by Nearest-Better Clustering.

Niching Community Based Differential Evolution for Multimodal Optimization Problems

Two-stage differential evolution algorithm using dynamic niche radius for multimodal optimization

A differential evolution with adaptive neighborhood mutation and local search for multi-modal optimization

Minimum Spanning Tree Niching-Based Differential Evolution with Knowledge-Driven Update Strategy for Multimodal Optimization Problems

Dual-Strategy Differential Evolution with Affinity Propagation Clustering for Multimodal Optimization Problems.

Learning Multimodal Parameters: A Bare-Bones Niching Differential Evolution Approach

A network community-based differential evolution for multimodal optimization problems

A Niching-Based Reproduction and Preselection-Based Multiobjective Differential Evolution for Multimodal Multiobjective Optimization.

A density clustering based niching genetic algorithm for multimodal optimization

Adaptive memetic differential evolution with niching competition and supporting archive strategies for multimodal optimization

A Reachability-Distance Based Differential Evolution with Individual Transfer for Multimodal Optimization Problems

Hypercube-Based Crowding Differential Evolution with Neighborhood Mutation for Multimodal Optimization

Optimal Peaks Detected-Based Differential Evolution for Multimodal Optimization Problems.

Pseudo Multi-Population Differential Evolution for Multimodal Optimization.

Differential evolution using improved crowding distance for multimodal multiobjective optimization