| How to locate the harmonic sources and quantitative their harmonic comtributions is an urgently power quality problem at present and the future in social development.In allusion to random distribution and existence of interference currents at non-harmonic source nodes, an improved sparse representation-based method to locate harmonic sources is proposed. Using sparse representation, the measurement equation for harmonic state estimation established by least square method (LSM) is turned into linear optimization problem and the selection of measuring nodes is adjusted, and then the positions of harmonic sources are judged according to estimation results of injected harmonic currents. Utilizing IEEE 123-bus system, the results of harmonic sources location obtained by the proposed method, the LSM-based method and the sparse representation-based method are compared. Comparison results show that the proposed method possesses the advantages of high positioning accuracy and strong ability of counter-interference.In allusion to the condition that the harmonic impedance is difficult to be attained accurately, a harmonic source location method under unknown harmonic impedance is proposed. Through observability analysis and symbolic admittance matrix, the measurement point can be achieved. Taking harmonic voltage as the quantity to be measured, the harmonic current is estimated by independent component analysis (ICA). The sequence of current estimation of harmonic source is solved by identification algorithm. And according to the search of the minimum value of objective function, the position of harmonic source can be located. In IEEE 14-bus system, from the two evaluation indices of variance and correlation coefficient, it is proved that the proposed method can accurately estimate and locate the harmonic source.Identifying harmonic sources with the background harmonic voltage dynamics is an important task. A new method is proposed to estimate the harmonic contributions of multiple harmonic sources by using data clustering analysis. And the total harmonic contribution index is then used to assess the total harmonic contribution of each harmonic source. Combining affinity propagation algorithm with k-means clustering, the harmonic voltage and harmonic current data are clustering according to different background harmonic voltage clusters, which are used to determine the harmonic contribution by partial least-squares regression. In IEEE 13-bus system, the performance of the proposed method is robust and well in terms of the background harmonic voltage fluctuation, unknown harmonic source and measurement noise. |
PDF Full Text Request