Research and application of fuzzy decision based on multi-agent system

The aim is to study the characteristics of dual-net and dual-protection configuration of 750 kV substation and take into consideration the uncertainty of the substation, such as maloperation and rejection miss trip of protection and breaker. This paper proposed a diagnosis fault method with redundancy based on fuzzy Petri net in a framework of multi-agent system. The reasoning model is divided into global agent layer, local agent layer, and connected agent layer according to multi-agent hierarchical method, and the fault region and suspicious fault component are determined by the former two layer based on case information. The restriction and rule are formed by topological structure and protection configuration of grid. This algorithm adopts the information entropy to determine the credibility of initial information, the model of fuzzy Petri is established in the third layer and divided into main network and redundant network to carry out the sub-network, and then the diagnosis result with minimum uncertainty is obtained by fuzzy reasoning. The simulation results show this model has higher accuracy and better fault tolerance ability.