Implementation of Particle Swarm Optimization Algorithm Inspired by the Social Behaviour of Birds
Neha Tyagi, Deepshikha Bhargava, Anil Ahlawat · 2024
Particle swarm optimization (PSO) is an evolutionary computation technique based on the social behaviour of flocking bird. The PSO method was first proposed by Kennedy and Eberhart in 1995, and has since developed into a generic optimization framework, well suited for application in virtually any industry. In this article, we seek to give an overarching introduction to PSO starting from basic understanding to core structure visit. This paper provides a complete introduction to PSO, protecting its fundamental ideas, which includes particle illustration, fitness assessment, and key hyperparameters including inertia weight, cognitive weight, and social weight. We use the mechanics of PSO, that specialize in how particle replace their positions based totally on personal best (pbest) and Global best (gbest) positions to stability exploration and exploitation. Significant advancements and modifications in PSO are reviewed, highlighting their effect on robustness and performance. These techniques consist of more desirable mastering strategies, fuzzy common sense for dynamic parameter tuning, mutation strategies, Lévy Flight, and competition-primarily based learning. The effectiveness of these innovations is represented via packages in energy structures, image processing, and function optimization.