Font Size: a A A

Study Of An Intelligent Type Fault Locaion In Distribution Systems

Posted on:2005-08-10Degree:MasterType:Thesis
Country:ChinaCandidate:B YanFull Text:PDF
GTID:2132360122488527Subject:Power system and its automation
Abstract/Summary:PDF Full Text Request
The study subject of this paper is Study Of An Intelligent Type Fault Location In Distribution Systems. The results of fault location in distribution systems are an important reference to restore faults rapidly. So study of the high accurate fault locating has a significant and practical worth.The study of fault location in distribution systems has being carried on for about ten years. Because of undistinguished fault features in single-phase-to-ground fault, it is being a knotty point. At present, this fault location is not yet to be applied in the practice. With the development of the distributed automation in distribution systems, this study of fault locating of the feeder becomes the more urgency needed.Based on the deep analysis into the fault location in distribution systems, the paper concludes all kinds of methods and pays more attention to essential problem.It is a breakthrough to extract key information, which directly reflect fault distance from fathomable electrical measurements with appropriate mathematics tool. The paper investigates applications of neural network and wavelet transform in power system, which provides necessary and reliable base of applications of fault location in distribution system.According to the characteristic formed of transient current component (TCC) of single-phase-to-ground fault in distribution systems (DS), this paper presents TCC of healthy phases in fault feeder as basis of fault location, which is mainly determined by fault distance. The wavelet measure sequence (WMS) of feature band can be extracted through wavelet transform. A modified algorithm to wavelet measure for eliminating the influence of the fault angle change is presented based on deep analysis factors influenced on WMS. On the basis of the above, it is built that the WMS based neural network model to realize mapping relation between WMS and fault location. A large number of simulation and test through EMTP/ATP prove the approach proposed that can accurate fault location to feeders.The paper analyzes the same and different characteristic between other types of fault and single-phase-to-ground fault, regards the reactive transient current as subject investigated, and presents new method of fault location based on WMS and NN model. A large number of simulation and test through EMTP/ATP prove the approach proposed that can accurate fault location to feeders.The study achievements presented in this paper, an intelligent type fault location in distribution systems, is creative and broken. Of course, we should go on ahead to find problem and improve the method on the job in order to accelerate the research findings in to reality. It is the direction which we should strive for.
Keywords/Search Tags:Distribution Systems, Feeder, Fault Location, Wavelet Measure Sequence (WMS), Feature Band, Neural Network (NN)
PDF Full Text Request
Related items