A Dual-Channel Particle Swarm Optimization Algorithm Based on Adaptive Balance Search

TL;DR

The paper tackles the persistent challenge of balancing exploration and exploitation in PSO by revealing two problematic P and G behaviors that undermine generalization. It introduces DCPSO-ABS, a dual-channel PSO with adaptive balance search, featuring a DC framework (non-G-channel and G-channel) and a cell-division initialization, plus an ABS strategy (PDG, adaptive channel selector, R&P box, adaptive transformation switch, resetter) to empower P with voluntary exploration and to regulate G's guidance. The approach decouples P and G and adaptively allocates channel usage across iterations, achieving superior generalization on 57 benchmark functions with demonstrated stability and scalability across different function evaluation budgets and dimensions. Results indicate that controlling P and G dynamics via DC framework and ABS yields robust performance across unimodal, multimodal, and complex landscapes, offering a practical pathway to improved PSO applicability in diverse problems.

Abstract

The balance between exploration (Er) and exploitation (Ei) determines the generalization performance of the particle swarm optimization (PSO) algorithm on different problems. Although the insufficient balance caused by global best being located near a local minimum has been widely researched, few scholars have systematically paid attention to two behaviors about personal best position (P) and global best position (G) existing in PSO. 1) P's uncontrollable-exploitation and involuntary-exploration guidance behavior. 2) G's full-time and global guidance behavior, each of which negatively affects the balance of Er and Ei. With regards to this, we firstly discuss the two behaviors, unveiling the mechanisms by which they affect the balance, and further pinpoint three key points for better balancing Er and Ei: eliminating the coupling between P and G, empowering P with controllable-exploitation and voluntary-exploration guidance behavior, controlling G's full-time and global guidance behavior. Then, we present a dual-channel PSO algorithm based on adaptive balance search (DCPSO-ABS). This algorithm entails a dual-channel framework to mitigate the interaction of P and G, aiding in regulating the behaviors of P and G, and meanwhile an adaptive balance search strategy for empowering P with voluntary-exploration and controllable-exploitation guidance behavior as well as adaptively controlling G's full-time and global guidance behavior. Finally, three kinds of experiments on 57 benchmark functions are designed to demonstrate that our proposed algorithm has stronger generalization performance than selected state-of-the-art algorithms.

Figures (9)

• Figure 1: nth particle's potential search regions in two-dimensional space at $k$th and$(k+1)$th iterations. (a) No $\boldsymbol{P}_n$ and $\boldsymbol{G}$ updated at $k$th iteration. (b) Only $\boldsymbol{P}_n$ updated at $k$th iteration. (c) $\boldsymbol{G}$ updated by itself at $k$th iteration. (d) No $\boldsymbol{P}_n$ updated but $\boldsymbol{G}$ updated by others at $k$th iteration. (e) $\boldsymbol{P}_n$ updated by itself and $\boldsymbol{G}$ updated by others at $k$th iteration.
• Figure 2: Architecture of DCPSO-ABS algorithm.
• Figure 3: Potential search regions under non-G-channel (blue) and G-channel (red). (a) Influenced by P. (b) Influenced by Q.
• Figure 4: Performance of DCPSO-ABS with various values of $M$.
• Figure 5: The number that each channel is employed changes with the number of iterations (up). The combination of two channels employed by the whole swarm on different iterations (down).