A new framework for power system fault diagnosis

With the ever increasingly enlarge and complicate development of the power systems, operators office would be flooded with extraordinarily abundant alarm information of the relevant relay equipments which can hardly be processed whenever the fault occurs. Fault diagnosis system can locate the fault equipments and the improperly executed relays according to the structure of power grid and the fault information, and the fault location is the key issue among these function. Therefore it is an essential part of maintaining the stable operation of the power system to study on and develop a completed fault diagnosis method for power system, and the discrimination of fault elements is the premise of self-healing grid. In addition, knowledge is the basement of modern artificial intelligence. The establishment of its representation is the premise of intelligent algorithm. This paper expressed the thought that making fault diagnosis through 3-layer framework to reduce the complexity of the algorithm. The middle layer-knowledge layer is mainly discussed in this paper and it contains topology knowledge base and protection knowledge base. The frame is realized in UNIX system and it is shown by a case that the framework can be diagnosed correctly and effectively.