Based On The Evaluation And Diagnosis Of Complex Electromechanical System Fault State Study On Fuzzy Fault Petri Net Improvement

Complex electromechanical system failure has the characteristics of multi-level, multi-variable, nonlinear, multi correlation and uncertain. The issue of failure diagnosis has become a hot and difficult research subject recently. So how to build an effective fault models to realize failure condition evaluation and diagnosis have become an urgent problem.To improve the safety and reliability of complex electromechanical systems, a modeling and reasoning method of improved fuzzy fault Petri net (IFFPN) is proposed based on fuzzy Petri net and fault Petri net. In order to describe machinery failure information more intuitively and clearly, the coloring rules of place, token and transition is defined, combined with the fuzzy production rules in the IFFPN model. Doing functional and structural analysis of the failure system, then, the FMEA method for IFFPN is used to analyze the failure data, finally, building IFFPN failure model based on the modeling rules. The forward reasoning of IFFPN model reflect the failure flooding characteristics. The forward deduction algorithm based on the intelligent judgment of firing matrix is applied in realizing machinery failure condition evaluation. The backward reasoning of IFFPN model is the process of machinery failure diagnosis, and thereafter, the backward reasoning matrix is proposed, combined with a minimal cut sets diagnosis based on the incidence matrix of Petri net, to realize automatic diagnosis with any place in the model. The minimal cut sets priority can be acquired according to the machinery failure frequency, which could contribute to the efficiency of the failure diagnosis.In the end, take a kind of CNC machine tools as an example, the IFFPN fault model of frequent fault sub-systems are built. Further, the feed system of CNC machine tool is performed for the realization process of the failure condition evaluation and diagnosis. Then, the correctness, effectiveness and feasibility are evaluated and verified by Pearson Correlation.