Simulation Research on Rolling Element Bearing Fault Signal Extraction Based on Blind Source Separation
Yong Xiang Zhang, Jie Ping Zhu, Shuai Zhang · Advanced materials research · 2014
In order to extract the fault information from rolling element bearing, combined with Kurtosis criteria and Hessian matrix. An improved rolling element bearing fault signal extraction method is proposed. Kurtosis is the cost function. The method is according to the construction principles of blind source separation (BSS), and it uses an analytically derived Hessian matrix in the maximization process of the cost function used. Then the impact signal is extracted successfully. The effectiveness of the method is demonstrated on simulated signal.