The Algorithm Study For Fault Diagnosis And Electricity Service Restoration In Distribution System

Distribution network is the final part of the whole power system which faces the customers directly. With the continuous development of social economics in our country, security of power supply is demanded by more and more electric power customers. It is essential that the DAS should be prompted on a large scale. Fault location, isolation and restoration system for distribution network is an important part of DAS which plays a crucial role in the feeder automation system to simultaneously detect and isolate fault section where particular fault occurs and restore power supply as quickly as possible for outage areas. Study of the algorithm for fault diagnosis and electricity service restoration is the heart of this paper.First of all, this paper introduces the model of structure variable dissipated network based on topology and measurement characteristics in the distribution system that is the theory basis for the distribution network analysis with computer. On the basis of the predecessor, this paper lists all kinds of algorithm covering the field of fault diagnosis and then classifies them into species.Based on the analysis of shortcoming inherent in existing matrix algorithm for fault diagnosis, the general ideas about the improved matrix methods is put forward. After precise mathematic inducing, improvements have been done individually according to two algorithm types to form general formula expression, as a result the weakness of old algorithm is effectively overcome.In practice, fault location and isolation systems are all exposed outdoors that are inevitably disturbed because of environment. In order to locate fault areas correctly, the algorithm must perform high fault-tolerance ability. Nowadays Genetic Algorithm is a representative to this type of algorithm. In this paper, a novel and general algorithm-Ant Colony Optimization is introduced to be applied to fault diagnosis in distribution network. There is a series of basic concept involved with Ant Colony Algorithm illustrated in this paper as background knowledge, and then a creative algorithm frame of fault diagnosis based on ACO is presented in detail.In this paper, the program for fault diagnosis based on the new method of ACO is coded on platform of VC++. With the analysis of new method performance and the contrast of Genetic Algorithm, ACO is verified practical and efficient.In the end, this paper makes an attempt at introducing ACO combined with topology theory once more to solve the reconfiguration problem for electricity service restoration outside the faulted zone. By this method, the search space will be limited to the feasible solutions, and the efficiency of calculation will also be enhanced obviously. The proposed algorithm and the fault diagnosis method in the former chapter can be effectively used to calculate the multi-stage optimization problem based on ACO after a distribution network undergoing a specific fault.