Research Of Transmission-Line Fault Diagnosis Based On Kernel Fisher Discriminant Analysis

Transmission line is the transmission hub of the entire power system,carrying the role of transmission of electricity,so the requirements of it must be highly reliable.However,the most vulnerable part of the whole power system is also the transmission line.And the reasons causing the breakdown of the transmission line are various and random,including many non-linear fault problems,which is difficult to diagnosis timely.How to ensure the safe,efficient and accurate operation of the transmission line,and the rapid diagnosis of failure are becoming key issues which researchers focus on.At present,it is difficult to deal with this nonlinear problem with the application of a single artificial neural network algorithm.Therefore,this paper introduces a new research method,that is,the combination of kernel Fisher discrimination method and artificial neural network,which can not only maximize the advantages of artificial neural network,but also efficiently deal with some non-linear fault information.This research program can bring great convenience to the industrial production and scientific research exploration,which has great practical significance and practical value.In this paper,we first analyze some of the fault conditions and the corresponding fault mechanism on the transmission line.Then we select the appropriate kernel function,input the sample information and the data to be processed,and project it from its own space to the high Dimensional space inside.After further data processing,the optimal kernel Fisher characteristic vector solution is obtained.Finally,the feature type of the fault is determined by the detection parameter,and the fault type in the transmission line is further determined.Aiming at the complexity of information processing in short circuit fault and the real-time and poor accuracy of single diagnosis method,a fault diagnosis model of transmission line based on KFDA is constructed.Finally,the fault detection of transmission line based on kernel Fisher is carried out on MATLAB analysis.In this paper,a large number of sample data are collected to prove the feasibility and accuracy of the designed transmission line diagnosis model.The results show that the method based on kernel Fisher discrimination and neural network fault diagnosis method can distinguish the fault type of transmission line properly,which is almost the same as expected.