Research On State Warning And Fault Diagnosis Of Mechanical Equipment Based On PNN-model

With the mechanical equipment becoming larger, more continuous and more automatic, the technology of fault diagnosis has been demanded higher. The article analyzed the present situation of mechanical equipment's state incipient warning and fault diagnosis in the field of fault diagnosis, and researched the problem using artificial neural network. The thesis has several main contents:(1) The present situation and the main technology method were discoursed within fault diagnosis domain, and several defects were put forward. According that, the main content and significance of this article was proposed.(2) According to the theory of artificial neural network, several neural network models' structure, operational mechanism, the choice of parameters and their merit and defect were analyzed. Probabilistic neural network was researched emphatically. And different improvements were made based on different application. Then the technology method of the problem was fixed.(3) In order to solve the flaw of mechanical equipment's state incipient warning technology, an incipient warning model of equipment running condition was constructed using probabilistic neural network based on equipment's historical data. The correctness and reasonability of the model were proved by simulation.(4) Based on the incipient warning model, self-adaption alarming line was set up according to the probabilistic. The simulation proved that the alarming line could adjust adaptive in line with the running condition of equipment.(5) Several typical faults of gear box were diagnosed by error back propagation neural network, radial basis function neural network and probabilistic neural network. Then a comparative analysis was made between the three network models.