Research And Application Based On Colored Petri Net Modeling And Fault Diagnosis Method

The research foundation of Petri theory is relatively mature, so it is widely used in various engineering research. Especially in the face of large and complex systems analysis and processing, more and more experts or scholars use Petri net methods to develop and research the project. These studies can be summarized into two categories, one is the optimization of the system, and the other is to diagnose the faults of the system. In series of research objects, the focus things of the research still are concentrated in the power grid and industrial automation system. Due to these problems have same characteristics, like nonlinear, discrete, uncertainty and so on.In the research process of automated storage and retrieval system (AS/RS), we found the system also has distinct features like the nonlinear, discrete and uncertain, according these features together with the large number of theoretical research, this paper boldly applied theory of the Petri net into the modeling and fault diagnosis based on automated storage and retrieval system. Firstly, this paper contrasted the modeling of transport car which set up by colored Petri net(CPN) and Petri net(PN), get the result with the modeling by CPN can be optimized,then according to the operation process in two areas storage of goods in the automated warehouse system, establishing the Petri net model, after introducing the professional colored Petri nets simulation software, we analyzed the model of state space, It is proved that the system model has the property of deadlock free, activity, boundedness and fairness, meanwhile there is no logical error in the design of the system. Then, according to the condition that AS/RS runs abnormally establishing fault Petri net, combined with examples to get the main system faults by fuzzy Petri net model. After that,according the fuzzy production rules to get the fault reason, put forward matrix description method for reverse inference verification, giving rise to find fault reasons accurately, at the same time, getting the probability of the occurrence of quantitative fault causes. Finally, by changing input conditions for different degrees of various fault simulations, Visual Object Net++is used to verify the reliability and accuracy of the proposed diagnostic method, which increases warehouse management efficiency.