Stochastic Triad Topology Based Particle Swarm Optimization for Global Numerical Optimization

Particle swarm optimization (PSO) has exhibited well-known feasibility in problem optimization. However, its performance still encounters challenges when confronted with complicated problems many local areas. In PSO, the interaction among particles and utilization of communication information play crucial roles improving learning effectiveness diversity particles. To promote particles, this paper proposes a stochastic triad topology to allow each particle communicate two random ones via their personal best positions. Then, unlike existing studies that employ positions updated neighboring position direct update, adopts one mean three associated as guiding exemplars update particle. further an archive is maintained store obsolete then used interact topology. enhance chance escaping from regions, restart strategy probabilistically triggered introduce initialized solutions archive. alleviate sensitivity parameters, dynamic adjustment strategies are designed dynamically adjust parameter settings during evolution. Integrating above mechanism, topology-based PSO (STTPSO) developed effectively search complex solution space. With techniques, largely promoted thus STTPSO expected explore exploit space appropriately find high-quality solutions. Extensive experiments conducted on commonly CEC 2017 benchmark set different dimension sizes substantiate proposed achieves highly competitive or even much better than state-of-the-art representative variants.