Vector control of brushless DC motors using neural network

Vector control is an effective technique for controlling variable speed drives using brushless DC motors. Conventional vector controllers, however, suffer from electrical machine parameter variations because these controllers depend on the parameters. This paper presents the vector control of a brushless DC motor using a neural network. In the proposed method, a neural network is employed as an online estimator of the nonlinear dynamic equations of a brushless DC motor. The neural network based vector controller has the advantage of robustness against machine parameter variations as compared with the conventional vector controller. Computer simulation results verify the useful of the proposed method.