Study On Fault Diagnosis And Fault Location And Electricity Service Restoration In Distribution System

With the increasing demand for electricity of our country, gradually ageing distribution system will not meet the power supply reliability and quality to the national industrial economic and the People's Daily life. Once the faults occurred, rapid fault diagnosis and shorten outage time is significant, which could improve supply reliability and safety and decrease economic losses. Currently, fault diagnosis and service restoration, which are the most essential and significant functions in distribution automation system, have become one of the most imperative task needed by the management.For distribution network fault diagnosis, this paper introduced rough sets theory, using the decision table reduction to describe the fault diagnosis problem. And a attribution reduction method based on priority of property is raised, Aiming at the decision table reduction. Combined with the attribution value decrease, it gets the reduced decision table and extracts sententious diagnosis rules rapidly. In fault diagnosis ,as for the mispair because of lack of pivotal information, rule match method united with Euclid distance and grey theory is maked use of. Rough set based diagnosis method can provide satisfactory results even lack of the essential information.In This paper,a data fusion technology is introduced to make good use of the steady-state and transient-state data of single-phase earth fault current , and a multi-criterion fusion faulty line detection method is put forward. A mass of MATLAB simulation results indicated that this method had integrated many available faulty line detection methods and can exert the strong points of each one, amplify the range of right evaluation for faulty line. It has more strong flexibility and error tolerant capability than the simplex method, and made more improvement in heightening the accuracy of faulty line detection. It ensured the dependability of faulty line detection and has wally perspective in the future.Fault location on line for distribution network is necessary with the improvement of demand for the reliability of the power supply. In this paper, a new approach developed for ground-fault localization on ungrounded and high resistance grounded system is described. In the approach, a improved feature extraction method based on convolution type of wavelet packet energy moment (CWP-EM) is presented. Compared with feature extraction method based on wavelet packet energy (WPE), the new one can better extract energy distribution feature in frequency bands. It extracts CWP-EM and constructs CWP-EM vectors of signals , Then takes the vectors as fault samples to train three-layer WNN(Wavelet neural network) which is based on immune particle swarm optimization algorithm(IPSO),finally realizes intelligent fault diagnosis. The MATLAB simulation shows that the proposed method could make full use of single-phase to ground faults information and Fault location of this method is more accurate and reliable.At last , heuristic research for obtain appropriate restoration schemes is applied in this paper. The method can reduce the number of combination switch action of segmentation or contact and quickly obtain the initial feasible scheme. Then each of the schemes would be assessed by the fuzzy evaluation technique from load transfer,feeder margin, minimal switch and maximal voltage drop. For this reason, the initial feasible scheme can restore load as much as possible with minimal switch operation.