Research Of Measure Placement Optimization And Topology Error Identification In Distribution System

With the development of distribution system and the expanding application ofAdvanced Meteing Infrastructure, the condition for the field application of distributionsystem state estimation is becoming mature gradually. For the distribution system in thefuture, the primary task is to grasp its real-time operation status and forecast its operationtrend, and this needs accurate state estimation result. However due to history andeconomic reason, the number of measurements and sensors in distribution system islimited. This has great influence on the precision of state estimation results. For one hand,due to the lack of measurement, a large number of pseudo-measurements are used in stateestimation, which adversely affects the accuracy and quality of state estimation result.For another hand, the topology of distribution system changes frequently. Due to the lackof sensors on most beakers, some errors may exist in the topology got by the topologyprocessor, which may result the state estimation result unbelievable. According to thesituation above, one way to improve the precision of distribution state estimation result isadding more measurements in distribution system and identifying the errors in topology.So in this thesis, how to optimize measurements position to maximumly improve thequality of distribution state estimation result is studied, and how to identify the topologyerror to provide a credible topology for state estimation is studied as well.Firstly, different distribution system state estimation methods are briefly introducedand compared in the paper. And the branch current-based distribution system stateestimation algorithm is introduced in detail.Secondly, based on the branch current-based distribution state estimation algorithm,the influence characteristic of different type measurements for improving the accuracy ofthe state estimation results in distribution system is studied through numerical tests. Andthe reasons for these influence characteristic are preliminary analyzed and discussed.Then, based on branch current error covariance matrix and branch power errorcovariance matrix, optimization method for measurement placement is proposed.At last, in this paper, the concept of optimal matching loop power is proposed.Based on optimal matching loop power, an efficient identification method is developed,which can find out several most possible topologies. Then, state estimator is used tochoose the most reasonable topology with the minimal measurement residuals.