Integrated Learning Particle Swarm Optimization Algorithm Based on Clustering

The particle swarm optimization algorithm is widely recognized for its few adjustable parameters and high flexibility. However, it still has some limitations. This paper proposes an integrated learning particle swarm optimization algorithm based on clustering, referred to as ILPSO-C. Using the adaptive learning rate clustering strategy, the population is dynamically divided into multiple subpopulations, each of which adjusts its learning rate based on the problem characteristics. The algorithm's adaptability is further improved through an adaptive parameter tuning strategy. A multimodal learning interaction mechanism introduces information sharing between subpopulations. Finally, a stalled activation strategy encourages the algorithm to escape local optima. The performance of ILPSO-C is evaluated on the CEC 2017 and CEC 2022 benchmark suites and the economic dispatch problem of power systems. Experimental results show that ILPSO-C outperforms other algorithms.