Research On Mathematical Morphology In Power System Protection

The rapid development of power system on capacity, voltage level and scale proposedhigher requirements to the technology of relay. Today, the relay has entered the field ofinformation technology, which is computer technology, communication technology, anddigital signal processing integrated development results. Among these, the mathematicalsignal processing, is to achieve the core technology microprocessor-based protection.Microprocessor-based protection is the use of computer to achieve the measurement, analysisand judgments when it’s a power failure and the operation is based on a number of discrete,quantifiable sequences of digital samples. It can be seen, the core issue ofmicroprocessor-based protection is to find an appropriate discrete algorithms, the result of theoperation to meet the engineering requirements at the same time, calculate thetime-consuming and as short as possible. Mathematical morphology is following thetraditional Fourier analysis and wavelet analysis and then a digital signal processing tools. It isentirely in the time domain waveform of the signal for analysis and processing, does notinvolve the frequency domain analysis, signal processing depends only on local features, andmorphological transformation contains only addition, subtraction and comparison operations.In addition, the calculation of mathematical morphology is simple, fast parallel, delay is small,easy-to-hardware implementation, which has great practical value.This paper has done a mathematical morphology in power system protection appliedresearch work, first introduces mathematical morphology, development and application statusin the field of power system. By the mathematical morphology open and close operations toconstruct the three forms of filters, and a detailed study of mathematical morphology on thepower system transient signals show several noise (high-frequency continuous interference,random background white noise, the peak transient interference) filtering effect. Simulationresults show that, after morphological filtering consistency coefficient of the signalsignificantly reduces, interference signals are varying degrees of smooth cut.This paper presents based on multi-resolution morphological gradient of transmissionline directional protection. The program first used mixed-filter to remove signal noise, andthen transformed with the multi-resolution morphological gradient extraction voltage, current, the polarity of the initial wave, in order to achieve fault direction of the judge. Simulationresults show that, in some special cases, such as near-field fault, transition resistance andsmall initial angle fault, there is still a good stability.The paper also proposes a morphology gradient-based algorithm for fault phase. First, Ianalyzed the fault characteristics of transient current under various fault types. After thephase-modulus transformation, Processing of the modulus by morphological gradienttransform, according to the results to determine fault phase. Simulation results show that thealgorithm can accurately and quickly determine all fault types.