Position regulation of robots via genetic p/fuzzy logic/pi control

摘要

A novel multi-mode genetic approach is proposed for the position regulation of robots via combined proportional, fuzzy logic and proportional plus integral (P/FL/PI) controls. To achieve a fast dynamic response with a low overshoot and small steady state error, the P, FL and PI controls are respectively used according to different regulation deviations. The parameters in the fuzzy logic control and the switching logic among these three different control strategies are simultaneously optimized via genetic algorithms (GAs), which thus guarantee the optimal self-tuning of the multi-mode controller with a high-level processor. Finally, the conventional PID, FL control, expert PID and the GAs based P/FL/PI control strategies are implemented on a robot arm for comparison. The experiment results demonstrate that our control method does provide the best dynamic response with a superior steady state behavior.