A New Inertia Weight Adjustment Strategy of Particle Swarm Optimization

• 1、School of Modern Post,Beijing University of Posts and Telecommunications,Beijing 100876,China

Abstract：Aiming at the premature convergence of particle swarm optimization (PSO) and the shortcomings of search accuracy of existing inertia weight adjustment strategies, this paper introduces the arctangent function of trigonometric function by being transformed and improved into particle swarm optimization and proposes a new inertia weight adjustment strategy based on Arctangent function (ATWPSO). To verify the performance of ATWPSO, the standard PSO, linear weight PSO, linear differential weight PSO, and three kinds of PSO with adaptive inertia weight are selected to be tested by 10 functions of 4 types such as unimodal functions, simple multimodal functions, hybrid functions and composition functions by different dimensions which were proposed on Congress on Evolutionary Computation 2017 (CEC 2017). The experimental results show that the ATWPSO strategy proposed in this paper has better performance in solving accuracy and standard deviation, and that especially in high-dimensional problems, it is easier to obtain higher search accuracy.

Keywords： Particle Swarm Optimization Inertia Weight Adjustment Strategy Based on Arctangent Function Convergence