Study Of Distribution Network Faultlocation Based On The Ant Colony Algorithms

With the development of economy and society, electricity users are more and more concerned for the power supply reliability, therefore, it is an inevitable trend to popularize the distribution automation system. Rapid and accurate fault location is the basic premise for isolating the faulty section and restoring the service in the un-faulted region, which is an important part of distribution system automation. Ant Colony Algorithm is a recently developed bionic optimization algorithm which has the features of distributed computing, rapid positive feedback, and heuristic search. So, the major subject of this thesis is to discuss the fault location of distribution network employing Ant Colony Algorithm.SCADA system can implement the fault locating with a certain algorithm in accordance with the fault information collected from FTU (feeder terminal unit). Because of the effect of various environmental factors, the FTU upload fault information is inevitably distorted. The traditional Matrix Algorithm has a high computing speed, but it is not error-tolerant; thus, a wrong judgement will be made when the fault information is distorted. Ant Colony Algorithm (ACA) is an optimum-searching algorithm and has good error-tolerance ability. So, when the fault information has a little distortion, with ACA, the faulty point can still be located accurately. However, when locating the fault in distribution network with ACA, the search efficiency is low because of lacking initial information, and, its pheromone releasing and updating mechanism still need to be improved.In this paper, the Ant Colony Algorithm for distribution network fault location is carefully studied, and accordingly several efficient improvement strategies are presented. First, a method for initial pheromone creating based on some useful information contained in fault information from FTU is proposed and the search efficiency of the algorithm is improved. Second, the pheromone releasing mechanism for distribution fault location with ACA is improved and the possibility of a local optimum is reduced. Thirdly, according to the basic idea of the optimal - worst ant colony algorithms, the negative pheromone updating mechanism is introduced, which further improves the search efficiency of the algorithm.An algorithm procedure is programmed with Visual Basic. The effectiveness of the improvement strategies is verified by calculation results. Finally, the prospect of ACA application in the distribution network fault location is discussed.