Speed Control of Brushless DC Motor Based on Single Neuron PID and Wavelet Neural Network

The brushless DC motor (BLDCM) is a multi-variable and non-linear system, so it is difficult to get a satisfying result for BLDCM using the conventional linear control method. This paper presents an approach of single neuron PID adaptive control for BLDCM based on wavelet neural network on-line identification. The method uses single neuron PID to construct the adaptive controller of BLDCM. In addition, a wavelet neural network (WINN) is built to construct the on-line reference model of BLDCM, and then identify the output of the motor. The single neuron PID controller achieves on-line regulation of controller parameters by self learning algorithm. And the identification network provides the gradient information needed by the algorithm. In this paper, a TMS320F2812 digital signal processor (DSP) is used to implement this control scheme. And the experimental result shows that the method proposed by this paper can achieve on-line identification and on-line control with high control accuracy, good static and dynamic characteristic and strong robustness.