Abstract：To characterize the contributions of harmonic sources in a power system,the method of assessing harmonic emission level was proposed,based on an improved complex regression. Fourier analysis was used to process the data of voltage and current measured at the point of common coupling( PCC). The system harmonic impedance has been carried out by improved complex least squares algorithm,and the harmonic emission level on either side of PCC can be assessed consequently. The simulation results show the validity and improvement effect of the method. The difference between the ordinary complex regression and the improved complex regression is that,the former minimizes the sum of squares of the absolute residual as its optimization objective,however,the improved method optimizes the measurement sample and constructs a new objective based on the relative residuals. Therefore,this improved method can remove the weighted effect of least squares,and reduce the influence of outliers and is suitable for the engineering application.

Keyword：relative residual;complex least squares method;harmonic impedance;harmonic emission level;