Controller Design and Optimization of Magnetic Levitation System (MAGLEV) using Particle Swarm optimization technique and Linear Quadratic Regulator (LQR)

Abstract

Magnetic Levitation System is one of practical examples which faces some nonlinearities behavior. Such systems require special types of controller parameters consideration for accurate results. In this paper, the process of tuning is to determine the system poles and getting them away from the instability region using state feedback (SF) controller methodology. The resulted controllable system parameters are estimated using LQR controller. Since the desired goal is to minimize vital parameters in the system behavior like the steady state error, settling time, raising time of the system and system overshoot, optimization techniques have been used to minimize cost function of the parameters which need to be optimized and reach for more reliable ones for better performance. Particle swarm optimization (PSO) has been used for tuning process. System operation points should be 0.61 A for electric current and 6 mm distance between coil surface and balanced metal ball, results show that using LQR controller will cause about 33% error percentage as steady state error and about 20% overshoot. Using PSO optimization technique for controller parameters will produce less steady state error of 6.5% with 4% overshoot percentage.