Material level detection and optimum control of BBD coal mill

In this paper, based on the noise signal, BBD ball mill material detection method and mill pulverizing system optimization control are presented. The noise of ball mill is decomposed using wavelet packet. The eigenvectors reflecting coal level of mill can be obtained from wavelet packet parameters. Through neural network training, the statistical model of coal level and ball mill eigenvectors is established. Therefore, the accurate material measurement is implemented. In addition, according to detection results, the mill optimal control method is also studied, based on neural network internal model control. This method can solve the control problems of ball mill, such as nonlinear, close coupled variables. Finally, the experiment results demonstrate the good performance of proposed method.