Fuzzy control of impact machines using genetic algorithms
Motofumi SASAKI, T. Uchida, Taroh Sasaki, Kunio KOIZUMI · 2002
A control design method for impact machines is proposed using fuzzy inference and genetic algorithms. The proposed fuzzy PID controller with output constraints employs only position measurements with no use of velocity measurements, and has the characteristics of optimal selection of fuzzification and defuzzification scaling factors, automatic selection of fuzzy rules, acquisition of smooth control input, high-speed, high-accuracy and robustness with respect to plant perturbations. By computer simulation, it is demonstrated that the proposed method is effective.