Based on Probabilistic Neural Network Model for Asynchronous Motor Fault Diagnosis
Nan Wu · Electric Machines and Control Application · 2013
According to the limitation of traditional fault diagnosis method,a diagnosis method based on probabilistic neural network was proposed.An example of asynchronous motor rotor with broken,eccentric,electric residual pressure fault was done.By choosing fault samples to train PNN,and then inputting the diagnosis information to the trained model of PNN,the occurred fault types could be judged from the output results.MATLAB simulation showed that diagnosis method of the motor based on probabilistic neural network could effectively identify motor fault and the fault diagnostic accuracy rate was so high that it could be easily implemented in engineering projection.But as neural network itself was undergoing developing,many problems need to be further studied.