Intelligent Diagnosis Model Based on Optimized Probabilistic Neural Networks

Abstract

Existing intelligent fault diagnosis models for equipment are insufficient in time-consuming and complication, making it hard to apply to practice. A novel intelligent diagnosis model has been carried out in this paper to improve this issue. Firstly, the process that experts realize the reasoning diagnosis by experience is analyzed to design an intelligent analysis flow. Based on the probabilistic neural network, the fault knowledge learning and reasoning from a large number of samples are carried out. Then the fault knowledge is mapped into a high-dimensional spatial distribution to realize the optimization of the probabilistic neural network. Finally, the fault bearing data is used to verify model performance.