Research On Dynamic State Estimation Of Asymmetrical Harmonic In Distribution System

With the development of new energy and power electronic technology,the increased types of single phase energy saving nonlinear load and the injecting of distributed generation in distribution network,asymmetrical harmonic pollution is serious in distribution network.Estimating the harmonic state of the distribution network accurately in real time has become an important guarantee for the effective control of harmonic pollution and improving the power quality.In this paper,the harmonic transmission characteristics of transformer and transmission line are given by theoretical analysis and simulation examples.When harmonic is asymmetry,single harmonic does not have one phase sequence characteristics.Three-phase asymmetrical harmonic can be decomposed into positive sequence,negative sequence and zero sequence components.The transmission of positive and negative sequence components is not affected by the transformer connection group while the transmission of the zero sequence component is affected by the transformer connection group.Due to the influence of the distributed capacitance along the transmission line,the harmonic current is amplified in the transmission process.Harmonic voltage injection has a maximum value at the harmonic source.On the basis of analysis on the transmission characteristics of asymmetrical harmonic,the three-phase measurement equation is established respectively with and without considering the phase sequence characteristics of harmonic,then a three-phase model for asymmetrical harmonic estimation is proposed.Based on the analysis of the three phase model asymmetric harmonic state estimation result,the symmetric component method is used to decompose the asymmetric harmonic into positive,negative and zero sequence components.Then respectively establishing the corresponding order network and measurement equation and a symmetric component model for asymmetric harmonic estimation is proposed.For these two different models,singular value decomposition algorithm is used to estimate the static harmonic state of asymmetry distribution network and the IEEE13 node system is used as a simulation example.The estimation result shows that the three phase network model without considering the phase sequence characteristics of harmonic has more accurate estimation results than considering harmonic phase sequence characteristics.By comparing the estimated results of the three-phase model and the symmetrical component model,the harmonic state estimation based on the symmetrical component model has higher accuracy.According to the continuity and inherent inertia of the load fluctuation of power system,and based on the current state and the estimated value of the upper state,an improved Kalman filtering algorithm is proposed to solve the dynamic harmonic state estimation model based on three phase model and the symmetrical component method.Example analysis of IEEE13 nodes with different structures show that the improved Kalman filtering algorithm proposed in this paper can accurately track the harmonic state of the system both the system is in steady state and dynamic state and the dynamic harmonic state estimation model based on the symmetrical component method has higher estimation precision.