Research On Topology Errors Identification And Parameters Estimation In Power System

The structure and parameters in the power grid are the backbone of the advanced application analysis of the power system,and it is also the bridge between the original data and the advanced application of the EMS on state estimation.It can be seen that the structure and parameters of the power grid are very important in the power system.In the analysis of this paper,the structure of the network is mainly for the geometric topology formed by considering the constraints of the system components and the constraints of the connection;the parameters of the network include the equipment parameters about the power components and the equivalent parameters on external network;the structure and parameters are collectively referred to herein as network model data.In practice,due to the intervention of human or natural factors,resulting in deviation or inaccurate calculation of the network model data and generating bad data.It is undoubted that directly using the network model data containing bad data will affect the accuracy of the network analysis.In order to identify and correct the bad data in the model data,this paper has carried out the following work:(1)Propose a multiple topology error identification method based on measured residual-branch topological bias modelThe relationship between topology deviation and measurement residual error is established by using the model of measurement residual and branch topology deviation.Then the theory of spatial projection collinear is introduced: measurement residual vector is collinear with its projection from that vector space located in the residual-branch incidence matrix corresponding to the topological error branch.Therefore,by using the residual vector and its projection vector close to zero as a criterion,the elements in the suspect branch set are successively reduced to discern the topology error,thereby realizing the identification of multiple topology errors.In addition,two criteria for the consistency of the switch state and the information of one side of the information to the other side are added to analyze the various types of errors that may occur in the branch,to some extent,which effectively distinguished the parameters or topology errors.The single topology error,multiple topology errors,parameters and topology errors constructed in the paper have three scenarios at the same time,and its results verify the effectiveness of the proposed algorithm.(2)Propose a parameters estimation method based on Ward equivalenceIn this method,the parallel structure of single ring network branch is constructed by Ward equivalent technology,and then the distribution relationship of the parallel branch power according to conjugate of the branch admittance is constructed.Finally,the linear measurement estimation equationes including the parameters located in the branch to be estimated is constructed.The method considers the parameters in both the transformer branch and the power line,which has wide applicability in the power grid.The power distribution relationship of parallel branches can effectively show the strong correlation between impedance and power.In addition,when the equivalent branch parameters of parallel branches change,the generated transfer power can improve the robustness of the proposed method.The test example of IEEE14 bus system construction shows that the effectiveness of the parameter estimation method of the ring network branch based on Ward equivalentce has great advantages compared with the extended state estimation and the two-sided information method.(3)Propose a black-box parameters estimation method using tie-line power mutationThe proposed method,based on the simplified Ward equivalent model of the external network,aims to address the following issues: unstable parameter estimation of the existing black box based static equivalence method;and low accuracy or difficult implementation.Correspondingly,the main features of proposed method,which makes it as a solution for these existing issues above,are summarized as follows: it is easy to implement as it only involves the associated measurements of boundary nodes in two adjacent time instants that before and after a designed large disturbance of the internal grid;it is robust to measurement errors,thus guaranteeing more accurate equivalent modeling;it can be applied to both two-port and three-port equivalent networks.At the same time,the theoretical method has been developed into practical software and put into operation in Guangdong power grid.The software mainly realizes the estimation of the static equivalent model of Hong Kong power grid.The equivalent model of Hong Kong power grid is combined with the model of Guangdong power grid to state estimate for the integrated model,which can provide a complete foundation model and data for the integrated analysis and calculation in Guangdong dispatching center.