A Study On Self-Tuning PID Controller Based Fuzzy Inferences

In this dissertation, based on fuzzy inferences, a self-tuning PID is developed. The principle of PID control and Fuzzy control is described. Both their developing history and trend are discussed. A self-tuning PID controller based on fuzzy inferences is designed by joining the advantages of first one to a second one.A model of fuzzy rules is built based on the operator's experiences, then it is changed into an inquiry form. According to system's real responding situation, fuzzy reasoning and decision are used to regulate PID controller's parameters online. Neither identification nor exactly mathematical model of controlled object is needed, it only needs to measure the controlled system's desired output and real output and compute the difference of them, then get the result of self-tuning factor by fuzzy inference and decision. After attaining the self-tuning factor, we can tune all PID controller parameters by the self-tuning formulas. It is obvious that this method is simple and feasible.Designing a controller for a two-order plus time delay system , Simulation results show that the algorithm has better performance than other methods.An algorithm of self-tuning PID parameters based on specified phase and gain margins is proposed. Simulation results of high order process show that the algorithm has high tracking speed.