A setting method of initial conditions in particle swarm optimization for positioning of a linear stage

Abstract

This paper considers the auto-tuning of a feedforward controller in order to improve the performance of stage positioning. Particle swarm optimization (PSO) technique is employed for the auto-tuning of the controller. In the proposed method, firstly, the dimensionality of the particle in PSO is reduced. Then, the initial placement of each particle is determined by considering the nominal value of the stage. By using these two approaches, the tuning time can be shortened. After the auto-tuning, the settling time of the positioning is improved. Effectiveness of proposed methods is shown by simulation.