Sensorless speed control of separately excited DC motor using neuro-fuzzy controller

A neuro-fuzzy controller designed for sensorless speed control of DC motor is presented in this paper. Artificial Neural network is used to solve the problem of tuning a fuzzy logic controller. The neuro-fuzzy controller uses neural network learning technique for tuning membership functions and setting the rule base from the simple data provided to estimate the DC motor speed. The speed of DC motor is estimated based on armature current and terminal voltage sensors to overcome mechanical and physical problems associated with traditional speed sensor. The neuro-fuzzy controller is designed and trained as a model adaptive reference system method. The DC drive circuit is designed, evaluated and modelled by MatlabSimulink in the forward and reverse motoring operation modes, respectively. The DC drive system is simulated at different speed variation in steady state and dynamic operating conditions. The simulation results illustrate the effectiveness of the controller. The speed response has fast dynamic response and acceptable agreement between the actual, estimated and desired speed.