The method of estimating customers harmonic emission level based on bilinear regression

Abstract

In the paper, the statistical methods of customer harmonic emission level are introduced in terms of whether they affect power system running or not, and their advantage and shortages are indicated. The paper principally discusses the method of estimating power customer emission level based on binary linear regression, which make it a condition that harmonic emission is steady at the point of common couple. Combined with system Thevenin equivalent and customer Norton equivalent, according to the least squares method, the voltage that a customers harmonic current at PCC can be estimated in the light of the plural correlation of network parameters. In the end, the correctness of the method is testified by experimental simulation with Matlab.