Robust Analysis And Design Of PID Controllers

This paper first analyzes performance criteria of a control system, and proposes a controller design method that incorporates disturbance attenuation and robust performance. The robust PID controller design amounts to an optimization problem with a non-convex constraint. Genetic algorithm is used to solve the problem due to its global search ability, simulation examples show that the method is effective and have general sense. However, genetic algorithm has complex search processes and bad realtime performance. So a simple robustness measure is defined and plotted. It is easy to judge the robustness of system by the robust diagram. Use this idea some well-known PID tuning formulae are analyzed and a few interesting conclusions are drawn. In the end, new PID tuning formulae are derived in this paper for typical industrial processes. Simulation results show that these formulae are effective and the resulting PID settings have good compromise between disturbance attenuation and robustness.