New approach of blind source separation and its application

Started off with the view of independence among the blind acoustic sources, a new approach of bland source separation (BSS) is put forward. This method, based on the joint probability statistics of blind source signals, is used to obtain the optimum rotation angle according to the direction derivative entropy of joint probability distribution minimum. Compared with the approach of fast independence component analysis (ICA) and neural networks, iteration computation is unnecessary for the proposed method, and the performance of the method is verified by the application to separating the mixing signals of bearing test rigs.