A System Based On Trouble Call Information For Locating Faults On Distribution Network

A new distribution network model for fault location reasoning is put forward, in which, a distribution network is regarded as a graph, switch devices and consumers are regarded as nodes while the feeder lines are regarded as edges. Based on the proposed model, an algorithm of distribution network hierarchizing is discussed.The method to describe a distribution network by object-oriented data structure of right child and left sibling binary tree is presented. Compared with other data structures, such as adjacent matrix and adjacent table, the proposed approach has the advantages of simpler in description, more flexible in searching process and more savings in space.Uncertainties in fault location reasoning are discussed. A subjective Bayesian method based fault locating scheme is presented to meet the requirement of uncertain fault location reasoning in distribution system. Certainty factors are introduced to describe the uncertainty in trouble calls together with the way to determine the value of such parameters. Based on human experiences of fault location judgment, important parameters such as likelihood of necessity factor and likelihood of sufficiency factor in subjective Bayesian methods are detailed. A new concept of threshold value is introduced, based on which, the mechanism of fault location reasoning is set up. A universal program is developed for locating faults on complicated distribution networks of different kinds. Simulation results show that the bad effects of mistaken information in a few trouble calls can be eliminated and consequently, a reliable reasoning result of fault location can be obtained. Based on the developed program, an integrated fault locating system using information from the trouble call system, and displaying by Geography Information System (GIS) is designed and realized. Design considerations to guarantee the accuracy and reliability of fault locating are also discussed.