Research About The Modeling Technology For The Fault Diagnosis Of Power System Basing On Petri Nets Decomposition

Petri Net, which can describe the asynchronous concurrent phenomenon and use graphsto represent, became a powerful tools for system modeling and behavior analysis.At the sametime,too many nodes result in the "state space explosion" problem.This problem is always thekey factor that bars Petri net’s development.Especially in the face of complex and hugesystem, the system analysis will become very complication.Therefore,one of the Petri netresearcher’s work is to simplify the Petri nets effectively and decompose the Petri net modelintoseveral relatively simple subnet.Fault diagnosis of power system is one of important stage of the electric power systemautomation.Production and propagation of power system’s fault is dynamic behavior whichcharacterized with discrete and concurrent.Petri net can use the graphic representation andrelating matrix to simulation and calculation.The power system’s fault diagnosis basing onPetri net is still in the theory research phase and has a good prospect.Petri net’s union decomposition technology and relative properties are described in thispaper. On the basis of the technique, it puts forward a scheme of decompose that decomposethe net with no loop and multi-output into some single output subnets. And this paper givesthe oretical proving process of correctness of this scheme.In order to validate the practical application of this scheme,this paper applies Petri Net inthe fault diagnosis of power system, and establishes a model for the fault diagnosis of powersystem basing on Petri net.Also,this paper decomposes the complex model into severalrelatively simple subnet models with this scheme.Taking the power system of powerdistribution room of a higher vocational college’s building for training practice in xiangtancity as an example, this paper designs a Petri Net model for it and decompose it. Finally,experiment on the exsiting PetriLab simulation platform, and then，analyze and compare theresolve.Therefore it provides a feasible way for the large or complicated system’s faultdiagnosis and shows the validity and practicability of this decomposition scheme andapplication from another way.