SPEED CONTROL OF DC MOTOR USING SELF-TUNED FUZZY PID CONTROLLER

Abstract

This thesis work describes how the speed of a DC motor can be controlled by a selftuned fuzzy controller. The speed of a DC motor depends on armature voltage, field current and the torque demand. In this thesis, the armature voltage is changed in order to have the required speed output. The armature voltage is firstly controlled by a simple proportional (P) controller and then controlled by a proportional-integral-derivative (PID) controller and finally the values of the proportional, integral and derivative constants are controlled and tuned by a fuzzy logic controller (FLC), hence the name Self-tuned fuzzy PID controller. There are two inputs to the FLC. One input is the difference between the reference speed (the desired speed) and the actual speed available as the output. This difference is known as the speed error (e). The other input is the rate of change of this speed error (de). Both of these are crisp (binary valued) sets. The FLC takes these two crisp sets, convert them into two separate fuzzy sets, takes decision based on a fuzzy inference system comprising of 25 rules in case of our model, then defuzzify the result into three crisp sets of the proportional gain (Kp), integral gain (Ki) and derivative gain (Kd). Basically the PID controller alone can control the speed of the motor, which is first shown in the paper. However, it is difficult to fathom the values of the parameters of the PID controller (Kp, Ki, Kd) that would give the best output (i.e., an output that has small rise time, settling time and fall time, small overshoot, and small or no steady state error).It is not practically feasible to make a trial and error analysis to figure out the values and the calculations regarding the computations of the accurate values can be quite complex, time consuming and not suitable at all in a practical application where a wide range of motor running speed may be desired. The FLC tunes the values of PID parameters based on the fuzzy rules fed into it and gives a satisfactory as well as controlled output wave shape for wide range of motor speed variation which is shown later. All block diagrams and simulations are done utilizing MATLAB Simulink, and the Fuzzy Logic Toolbox of MATLAB is used to design our proposed model of FLC.