Research On State Estimation For Active Distribution Network Considering Harmonic

With more distribution genentaiton(DG) integrated into distribution networks, today's distribution networks are evolving from radial convential distribution networks with single source to active distribution networks(ADN) with multiple sources and meshes. Thus, the research about ADN distribution management systems(DMS) is attracting more and more attention. As a significant part of DMS, distribution network state estimation provides the basic model for other advanced functions. However, high penetration of DG leads to a large amout of plug-in power electronic devices which will provide plenty of harmonic. Besides, the nonlinear load will also produce a great deal of harmonic. The harmonic component will be measured as the part of measurement because the measuring instrument is designed based on fundamental frequency. As a result, the precision of state estimation is low due to the bad measurement caused by the harmonic component. Hence, aiming at the problems metioned above, this paper analyzes the influence of harmonic on measurement and state estimation. A method for reducing the effect of harmonic on state estimation is proposed in this paper. The main work of this paper is as follows:In terms of fundamental power flow, the loop-analysis-based power flow algorithm is introduced for the weakly meshed active distribution network while a strict method for processing PV nodes is proposed. As for the basic harmonic power flow, the uncertainty of harmonic spectrum is considered. Thus, corresponding Latin hypercube sampling is proposed for independent variables, dependent variables and discrete data, respectively. The validity of proposed method is verified and the influence of harmonic on power flow is analyzed.The models of harmonic source, voltage and current measurement channels are developed. Then, the whole measurement system is constructed according to the principle of electronic energy meters. The measurement and state estimation errors are analyzed. Last, the effect on state estimation caused by harmonic is demonstrated by simulation considering different scenarios including different length of channel cable, harmonic order and content.The active distribution network is divided into two parts: monitoring area and non-monitoring area according to the location of harmonic monitoring devices. The state variables and constraints are added into the state estimation for monitoring area in order to eliminate the harmonic errors. Thus, an anti-harmonic-error state estimation based on the harmonic monitoring is proposed. Last, the validity of proposed method is verified.