Artificial neural network based permanent magnet DC motor drives

A novel scheme for the speed control of a permanent magnet (PM) DC motor drive incorporating an artificial neural network (ANN) is proposed. The drive system includes an ANN speed controller, microprocessor based DC-DC power converter and a laboratory PM DC motor. A multilayer artificial neural network structure with a feedback loop is designed in order to precisely operate the control circuit for the DC-DC power converter. The complete drive system is simulated and implemented in real-time. Both the simulation and experimental results prove the inherent capability of the ANN which makes it possible to maintain desired speed control in the presence of parameter variations and load disturbances. The performances of the ANN based PM DC motor drive system are compared with the simulated results of the conventionally controlled drive system. This clearly indicates the better performance of the ANN based PM DC motor drive system, particularly in the case of parameter and load variations.