Research On Discrete Event Dynamic Systems Modeling Based On Fuzzy Neural Petri Net

Different from the continuous variable dynamic systems, discrete event dynamic system's specific discrete characteristics make continuous variable dynamic systems cannot use sophisticated research methods to study. Petri net is a powerful tool to study the discrete event dynamic system. Through system modeling we can analyze the intrinsic nature and operation performance of the system, can be designed a superior performance system to meet the actual requirements. You can also be improved for the existing systems and provide theoretical support, and guide people for production.In this paper, Petri net and its extension model a more in-depth study, and by examples of relevant research results are analyzed to verify the theoretical results to the actual production of the guiding value. The content of this paper have been summarized as follows.Firstly, described the basic concepts of Petri net, basic nature and analysis methods, and introduced the detailed definition of fuzzy Petri net and fuzzy neural Petri net-related idea, and pointed out their advantages in discrete event dynamic system modeling study.Secondly, researched fuzzy reasoning representation and optimization algorithm in fuzzy Petri net, and according to propos general fuzzy production rule representation describes the optimization algorithm by use the form of a matrix. The examples of the application in fault detection method is simple and convenient, its parallel computing features of the system can quickly find the failures. The research results not only for fault diagnosis and effective, but also can be applied to other fuzzy Petri net-based discrete event dynamic systems modeling and analysis.Thirdly, in fuzzy Petri net cited artificial neural network theory. Fuzzy Petri net have access to the learning ability of artificial neural by using neural network theory, and through the network training it will be able to converge in a relatively short time to allow the error to the target range. This paper describes two kinds of models, and conducted research and experimentation, the experimental results to achieve the desired objectives. Additionally, through the fuzzy neural Petri net can also predict the state of the system, according to forecasts to take appropriate measures to deal with.Finally, the Timed Petri Net-period optimization idea was proposed. We can use the minimum-maximum algebra method to constructing the functional relationship and then using dynamic programming approach to optimize the cycle. Comprehensive results of previous studies, expounded modeling steps and analysis method for discrete event dynamic systems based on fuzzy neural Petri net theory. For the logistics system in the enterprise logistics automated warehouse system, using the steps above discussed in detail the modeling process and gives the modeling results. To a certain extent, it verifies advantage of the model based on fuzzy neural Petri net theory.The research of the above discrete event dynamic systems modeling and performance optimization in the production practice has obvious significance.