PARTICLE SWARM OPTIMIZATION BASED FIXED-STRUCTURE H_∞ LOOP SHAPING CONTROL OF MIMO SYSTEM

摘要

H_∞ loop shaping is a sensible method for designing a robust controller; however, the controller designed by conventional H_∞ loop shaping is complicated with a high order. It is not easy to implement this controller in practice. To overcome this problem, in this paper, a new design technique of a fixed-structure robust controller for MIMO system is proposed. The performance and robust stability conditions of the designed system satisfying the H_∞ loop shaping are formulated as the objective function in the optimization problem. Particle Swarm Optimization (PSO) method is adopted to solve this problem and to achieve the control parameters of the proposed controller. The performance and robustness of the designed controllers, centralized and decentralized PID controllers, are investigated in a MIMO system (a chemical process) in comparison with the conventional H_∞ loop shaping control. Simulation results show that the robustness and performance of the proposed centralized controller are almost identical to those of the controller designed by H_∞ loop shaping method. However, because of the complicated controller in the conventional design, the proposed approach offers a significant improvement in practical control viewpoints by simplifying the controller structure, reducing the controller order and still retaining the robust performance. Simulation results also demonstrate that the proposed approach is numerically efficient and leads to performance comparable to that of previously published methods.