Classical and Intelligence Speed Control Techniques for Separately Excited DC Motor (Published)
| Most industrial process that uses DC Motor requires that DC Motor operates at a desired speed depending on the load and this speed should be sustained during the operational process however, a significant deviation from the desired speed trajectory was observed on the speed characteristic of DC Motor when acted upon by a load. This paper is aim at investigating the best controller for controlling the speed of a Separately Excited DC Motor in which four different controllers were deployed; the classical Proportional Integral (PI) Controller, Fuzzy Logic Controller (FLC), Artificial Neural Network (ANN) Based Controller and the Adaptive Neuro Fuzzy Inference System (ANFIS) Based Controller. The PI controller was designed by tuning its parameters in MATLAB Simulink in which the proportional and the integral gains were obtained for the best performance as 100.83 and 1750.45 respectively whereas the ANN and the ANFIS controllers were trained to mimic the desired plant input and output relationship. The FLC was designed to have single input which is the error signal and single output which is the speed using five membership function which give rise to five fuzzy rules based on Mamdani principle. A transient analysis was carried out on individual controllers using a speed reference of 1600 rpm to 2200 rpm in steps size of 200 rpm and it was observed that the ANFIS controller demonstrated a higher level of performance in tracking the input reference with an average percentage overshoot of 18.25%, a settling time of 1.446 seconds and a steady state error of 0.1%. |
Keywords: ANFIS, ANN, DC motor, Fuzzy Logic Controller, NARMAL-L2, PI controller, Speed