WEIGHT OPTIMIZATION AND STRUCTURE-SPECIFIED ROBUST H_∞ LOOP-SHAPING CONTROL OF A PNEUMATIC SERVO SYSTEM USING GENETIC ALGORITHM

摘要

H_∞ loop-shaping (HLS) is a feasible method for designing a robust controller; however, the controller designed by this method is usually complicated with high order. To overcome this problem, we propose a new design technique, weight selection and structure-specified HLS control using genetic algorithms (GAs), to design a robust controller. In the proposed technique, both the performance weight and structures-specified robust loop-shaping controller-are simultaneously determined by GA. The performance and robust stability conditions of the designed system satisfying the HLS are formulated as the objective function in the optimization problem. The designed controller contains simple structure with lower order and still retains the robustness and performance specification. As shown in the simulation results, the robustness and performance of the proposed controller are almost identical to those of the controller designed by HLS method and the proposed technique leads to performance comparable to that of the reduced-order controller. Experiments on the pneumatic actuators of a pneumatically actuated robot are performed to verify the effectiveness of the proposed technique.