PID Multi-objective Optimization Design Based on Pareto Optimality

It is hard to consider the system requirements for both time and frequency domain performances with the current PID optimization methods.To solve this problem,a PID multi-objective optimization design method is proposed: taking dynamic performances as the optimization objectives,and frequency performances as constraints,a multi-objective optimization algorithm based on Pareto optimality is applied to the optimal design.The algorithm employs a new calculation method of crowding distance and double elitism mechanisms,which improves the efficiency of the evolution and more effectively increases the diversity of the Pareto-optimal solutions.And the satisfactory solution is selected in Pareto-optimal solutions according to the requirements of the present system.Simulation results and its application indicate the effectiveness of the proposed algorithm.