Cable Fault Recognition Method Research Based On The Wavelet Energy Spectrum Of Mode Components

With the continuous development of china’s national economy, the pace of the powersystems construction becomes faster and faster. The power system constituted by generation,transformation, transmission, distribution and consumption，there will be occur variousshort-circuit faults during the transmission, Transient traveling wave signal generated by thefault propagation along the road, travelling wave signal contains a large number of faultinformation, detect these faults information timely and extract the fault feature, and identifyand judge these faults accurately are very important to ensure the stable operation of thepower system.Firstly, the paper analyzes the cable fault causes, types and the reasons andcharacteristics of cable fault signal, introduce the current research situation of the transientsignal recognition. Describe the process of transmission line fault traveling wave, thedispersion characteristics in the transmission process, the application of wavelet transform inthe fault traveling wave detection and the selection of wavelet basis function for faultdetection.Secondly, a simulation model is built; on the basis of this model each mode componenttraveling wave dispersion characteristics and the influence of various factors in the process ofthe fault simulation are analyzed.And then, on the basis of the wavelet theory, wavelet energy spectrum and phase modetransformation theory, a fault recognition algorithm proposed in this paper which based onmode components wavelet energy spectrum, through compared with the fractal dimensionmethod and wavelet modulus maxima method, analyzes the advantages and disadvantages ofvarious methods. The algorithm through extract the three-phase fault transient current signal,phase mode transformation is used to get three separate mode components current signal; Then, using the wavelet transform on the component of three current wavelet decomposition,can get high frequency coefficient after wavelet transform which can reflect the details of theoriginal signal high frequency energy coefficient; Finally, according to the theory of waveletenergy spectrum to deal with detail coefficients, get the mode components current signalenergy, according to the different fault types the mode components carry different energy, canrecognize the fault types. Using the simulation model to test the algorithm in different kindsof short circuit, and construct identification criterionFinally, in order to further determine the algorithm, the recognition algorithm is appliedto cable fault data acquisition based on LabVIEW platform. Through analyze the actual data,the results are consistent with the simulation.The simulation and system experiment results show that the identification criterionwhich constructed through the simulation model and the algorithm can accurately effective torecognize different kinds of fault types, in the experiment system can also be stable operationand the identification process is easier.