Touch screen-based motor bearing fault diagnosis

A combined method with wavelet packet and BP neural network based on touch screen for motor bearing fault diagnosis is presented. Firstly, this method uses the time-frequency technology of wavelet packet for the feature extraction of motor vibration signals. Secondly, BP neural network is designed based on energy feature vector, and the algorithm is realized with MATLAB software. Finally, diagnostic results are displayed on the touch screen, which is based on three typical running states of motor rotor system. Simulation studies show that the proposed algorithm is reliable, and efficient.