Particle Swarm Optimization Considering the Concept of Predator-Prey Behavior

Recently, a variety of optimization algorithms has developed as systems get complicated. One of those is called Particle Swarm Optimization (PSO). PSO is an algorithm which takes a cue from nature´s bird flock or fish school and is known to have superior ability in search and fast convergence. However, it might be difficult to find global optimal solutions when it comes to complex higher-dimensional objective functions which have a lot of local optimal solutions. Therefore, we focused on the predator-prey behavior which is one of the most important concepts in nature but has not been taken in PSO yet, in order to improve the ability of PSO. This paper presents a new PSO which takes in the concept of predator-prey behavior, that is, predators chase the center of preys´ swarm, and preys escape from predators, in order to avoid local optimal solutions and find global optimal solutions efficiently. And then, we validate the proposed method through numerical simulations with several benchmark problems comparing to well known PSO method.