The Research And Application Of Knowledge Representation And Reasoning Based On Improved Fuzzy Petri Nets

With the rapid development of science and technology,knowledge has become the driving force of economic development and enterprise success.As a kind of knowledgebased system,the expert system is characterized by using the captured expert knowledge to help nonexperts solve specific problems reaching the level of experts.A lot of research and practice have proved that knowledge representation not only has great impact on the performance of expert systems.Accurate and efficient knowledge representation and reasoning is the key to the success of expert system.Fuzzy Petri nets(FPNs)are a useful instrument for modelling knowledge-based systems to conduct knowledge representation and reasoning.It is not only deals with uncertain,imprecise and fuzzy information in various types of expert systems,but also represents fuzzy production rules and formulates fuzzy reasoning.However,it is sometimes inaccurate and inefficient to apply FPNs directly in knowledge-based systems due to the rise in complexity of systems.To overcome the drawbacks of traditional FPNs and improve its performance from the perspective of knowledge representation and reasoning,this thesis explores two modified FPN models and applied them in expert systems with uncertainty.Specifically,the research·proposes a cloud reasoning Petri net(CRPN)model by using interval cloud and interval cloud hybrid averaging operator for knowledge representation and reasoning in fuzzy environment.A fault diagnosis case and comparison with some previous methods are presented to elaborate the application and advantages of the proposed CRPN model.·develops a grey reasoning Petri nets(GRPNs),for knowledge representation and reasoning in the large group environment,by combining grey numbers and Bonferroni mean operator.The proposed GRPN model is conducted to evaluate risks in transportation system and results are compared with existing studies.