Study On Harmonic Source Identification Technology In Distribution Grids

With the widely application of the electrical equipment in the power electronic technology, harmonic pollution is more and more serious. Harmonics not only cause a bad influence to the power grid and power equipment, but also severely reduce the efficiency of electric power, it has endanger the safety and economic operation of power grid and electrical equipment. At present, the problem of harmonic analysis and integrated governance has become a subject widely concerned by home and abroad, but the fiest problem of solving the harmonic pollution fundamentally is finding the source of harmonic. Therefore, matters of how to correctly determine the harmonic source location and accurately distinguish their respective harmonic responsibility has been the primary problem of harmonic analysis and comprehensive control, at the same time, it can serve as the basis of harmonic power flow calculation and punishment harmonic overweight user.In this paper, on the basis of the existing distribution network harmonic source identification method, the author put forward a kind of based on partial least-squares regression to estimate the system harmonic impedance and harmonic emission levels by the user. Firstly, this paper analysed the existing distribution network harmonic source identification method qualitatively and quantificationally, and respective advantages and disadvantages are compared. Then, based on the equivalent model of the system side and user side, the research based on single point measurement under the condition of using partial least squares regression method for harmonic source identification; Use the synchronous harmonic current and harmonic voltage of PCC to solve the regression coefficient, after that, the system harmonic impedance and the harmonic emission level can be calculated by regression coefficient; Based on partial least-squares regression method of harmonic source identification with "robust regression estimation method" and "dual regression method" are compared and analyzed, and simulation result by MATLAB/Simulink indicated that in ideal state, using the method of partial least-squares regression for harmonic impedance and the harmonic emission level of the estimated result is reasonable and smaller error. Finally, through to the power network topology analysis, to calculate the bus node branch of harmonic current relative branch injection of harmonic current the sensitive factor of change, and on the basis of the sensitive factors to optimize the harmonic monitoring configuration, also built IEEE 14 bus network model by MATLAB to analyse and verify, reasonable experimental conclusion is given for the multi-point measurement of harmonic source location technology to provide theoretical basis.