Using Probabilistic Neural Network to Diagnose the Fault of Self-propelled Gun Engine

Aim\ To research Probabilistic Neural Network(PNN) model, and its application to fault diagnosis. Methods\ Based on probability statistics theory and Bayes Classification Rule, PNN model, network structure, algorithm, and their characteristic is analyzed, and applied to fault diagnosis. An optimization method to estimate smoothing parameters is established. Results\ It is very good to diagnose the oil and gas fault in Self propelled Gun Engine(SPGE) by using PNN. Conclusion\ Using PNN will get better effect in the field of pattern recognition and fault diagnosis.