Application And Research Of Using MBD In Fault Diagnosis For Electric Distribution Network

In modern power system, the electric distribution network structure is becoming more and more complex, and the probability of its failure is increasesing. In recent years, many scholars use the theory and method of artificial intelligence in fault diagnosis of the electric distribution network in order to find out the cause of the malfunction in a relatively short period of time when the electric distribution network breaks down or an accident happens, Most of the existing diagnosis systems used in electric distribution network are expert systems based on experiences, which can not diagnose the faults beyond experiences and are very difficult for system transplanting and maintaining. In order to overcome the defect of the traditional expert system, the method of model-based diagnosis (MBD) is applied to the electric distribution network.Firstly, this paper analyzes the current studies on the fault diagnosis of the electric distribution network, the background and the significance of the subject. The basic ideas, the basic processes, the common classification and the common diagnosis strategies of the MBD were introduced.Secondly, the concept and the related theories of minimal conflict sets were introduced. The search process of the minimal conflict sets is divided into two steps in this paper. The first step is to search potential minimal conflict sets off line, and the second step is to search the minimal conflict sets from the potential minimal conflict sets online. To the problem of searching potential minimal conflict sets, Relation Guiding Algorithm (RGA) and two conventional search algorithms were introduced, and a test comparing the three algorithms was made. The test indicated that RGA algorithm performance is superior to the other two algorithms in its real-time property. At the same time, the on-line diagnosis strategy was discussed.Thirdly, to the problem of the complexity and the low efficiency in the process of computing the minimal hitting sets and the program programmed difficultly when MBD is employed to diagnose the electric distribution network, this paper proposes an algorithm for the computation of minimal hitting sets based on the binary code logical operations. The algorithm maps the minimal hitting sets problem to 0/1 integer programming problem, achieving the goal of finding all the minimal hitting sets in a short period. Compared with other algorithms which compute minimum hitting sets, the algorithm has good real time property, and an example of actual electric distribution network is adopted to explain the process of computing minimal hitting setsIn the end, considering the structure of the electric distribution network and the demands in the fault diagnosis of the electric distribution network, this paper presents a scheme in which the theories of MBD is applied to the distribution system for fault diagnosis, and an example of actual electric distribution network was given. In this example, the uniform models were constructed and the system description of the electric distribution network was given. In addition, the diagnosis experiment was done with a set of observations of the electric distribution network. According to the result of this experiment, it was verified that the scheme was feasible and effective.