Research And Its Application On Dynamic Fuzzy Petri Net

Petri net theory is a powerful tool for the research of discrete event dynamic system. As an important branch of Petri net, dynamic fuzzy Petri net can concisely describe the states and the variation process of the DEDS, which is proposed by Professor FanzhangLi from SooChow University. The thesis carries out the research about the theory of DFPN as well as its appliacation on fault diagnose.Firstly, this paper discusses the basic theory of dynamic fuzzy Petri net (DFPN), DFS and DFL theory is also introduced. By means of some examples, this paper discovers the importance of DFPN theory in some reserching field.Secondly, DFPN has graphic interface, but it could not give enough proof for parameter tuning and setting. By useing some methods in fields of artificial intelligence, the problem of parameter optimization can be well sovled. After intoducing evolutionary computation into DFPN parameter optimization, this article enlarges the applicable scope of DFPN model or rules.every transition has independent certainty factor and threshold value, which impoves the self-adaptation for DFPN in evolutionary learning, as well as its generalization ability.Finally, based on DFPN, this article proposes some design method for fault diagnose.through two sets of simulation test, the availabiltiy of the software is verificated.All in all, features of the paper reprents as follow:1. Discusses the basic theory of dynamic fuzzy Petri net.2. Introduce theory of DFPN into optimism of genetic algorithm and fault diagnose.3. Confirm the validity of DFPN by simulation test.Although we got some summings-up, the work what we do is basic research. There are some questions that need be deeply researched. For example, the robustness of DFPN, the stability of multi-mode DFPN etc.