Probabilistic neural Network with statistical feature for fault diagnosis of permanent magnet motor
Lei Dong1, 4, Weimin Li1, Weiguo Zhao2, Yunfei Chen3
1School of Mechanical Engineering, Hebei University of Technology, Tianjin 300130, China
2School of Water Conservancy and Electric Power, Hebei University of Engineering, Handan 056038, China
3School of Control Science and Engineering, Hebei University of Technology, Tianjin 300130, China
4Tianjin Navigation Instruments Research Institute, Tianjin 300131, China
Permanent magnet motors are very important components in commercially available equipments and industrial applications due to high reliability and robust performance, and it is important to take an appropriate and effective approach to diagnose fault for them. The implementation of probabilistic neural network (PNN) with the statistical features for permanent magnet motor is developed in this paper, and the statistical features are determined according to the stator current characteristics of motor to effectively reduce dimensionality of sample space. The experimental results demonstrate that, compared with RBF network, the proposed method is more effective in identifying various types of faults.