Optimization methods based on metaheuristics are proposed as a class of global optimization methods, by which the global minimum can be obtained without being trapped in local minima. Particle swarm optimization (PSO), which is one of these methods, is known for its high searching ability and easy implementation. However, it might be difficult to find the global optimum for optimization problems with a number of decision variables and multiple local optima. In this paper, we propose three types of new PSO methods to overcome this difficulty. One is a model with nonlinear dissipative term introduced by Fujita and colleagues  to prohibit the search point's velocity from being zero. The others are models with the nonlinear dissipative term with the pbest or the gbest information to disturb the search around them.
ASJC Scopus subject areas
- コンピュータ ネットワークおよび通信