Research On The Short-circuit Fault Recognition In Power System Based On The Electronic Sensor

On-line detection and fault diagnosis in power system is the base of protective relaying. With the development of electronic technology, computer technology, communication technology and network technology, people demand higher quality of power supply, monitor technology of high voltage and current in distribution and supply network is improving. Now, widely used traditional electromagnetic mutual-inductor have the problems of magnetic saturation,ferroresonance,small dynamic range,narrow bandwidth, which are unable to satisfy the development of power system. So, to research the new type detection instrument, and discuss the exact fault diagnosis method has the important meaning.This thesis is based on electronic transducer which has been widely researched,analyzed and validated the feasibility and precision of current sensor unit--Rogowski coil, which was used for the acquisition of transient current in power system. The whole structure of electronic has been introduced, and utilized the output feature of Rogowski coil, designed the short-circuit fault recognition system.As for the feature extraction of short-circuit fault, the advantages of wavelet transform to the traditional Fourier transform have been analyzed, and chose the db4 wavelet that is compactly supported with favorable orthogonality, extracted wavelet coefficient as the vectors of fault characteristic, and designed this arithmetic on the DSP platform. At last, utilized local effect of radial base function, a short circuit fault recognition based on the RBF neuron network has been designed, and chose the digital neuron network chip ZISC78 as the hardware platform.On the platform of Matlab with Power system blockset, a transmission line of 10kV was simulated. After data acquisition, feature extraction, and fault recognition, satisfied results of this short-circuit recognition system has been validated.