A Multilayer and Distributed Alarm Processing and Fault Diagnosis System Based on Multiagent

Blackouts caused by accidents can not be avoided completely while power grid is becoming larger and more complex. So how to analysis faults and find out the reason and the location as fast as possible, then take exact measures to reduce outage scope and duration, is a very important issue. Local alarm processing and fault diagnosis system can deal with local faults with local information. But there are a series of partial faults in wide area when large blackouts happen. So it is necessary to establish alarm processing and fault diagnosis system in multilayer structure. So multilayer and distributed alarm processing and fault diagnosis system are established, including local systems and regional systems. In order to make the systems more flexible and intelligent, multiagent architecture is adopted. Moreover, we design a specific FDA (fault diagnosis agent) on the basis of BDI (belief-desire-intention) model, and introduce SSM (simulating society multiagent) cooperation method. Knowledge-based method, the way specific responsibility Agents in charge of logic analysis make good use of knowledge, is also discussed in detail. Calculations based on the methods introduced in this paper show the fast and accurate results and the prospect of application in the prevention and disposal of serious blackout of large-scale power grid