Study On Operational Topology Identification And Optimal Measurement Placement Of Distribution Networks

Accurate real-time operation state perception is the basis to ensure safety,reliability,economy and efficiency of distribution networks,which can be effectively realized by means of state estimation.The increasing access of distributed generations and electric vehicles to networks makes the network structure and operation state more and more complex,leading to the improving requirements of the operation state perception ability.Based on the state estimation theory,operational topology identification and optimal placement of micro-phasor measurement units in distribution networks are studied.The main contributions are as follows:1)The basic distribution network state estimation algorithm based on node voltage is analyzed.The state estimation process is expressed in mathematical terms and measurement equations with state variables of node voltage amplitudes and phase angles are established.Then,the estimation model is built under the criterion of weighted least squares and the Gauss-Newton iterative scheme and normal equations are derived to solve the model.2)A two-stage topology identification method based on power flow matching for radial distribution networks is proposed.Firstly,an improved mixed integer quadratic programming topology identification model is established to generate the preliminary topology.Secondly,a final topology identification model based on spanning tree generation algorithm and power flow matching is proposed,where the set of neighbor trees of the preliminary topology is searched by the spanning tree generation algorithm,and the power flow matching evaluation index is determined to select the best tree as the final topology.Elimination of trees with poor indices effectively reduces computation costs.Finally,IEEE 33-bus test feeders are used to validate the effectiveness of the proposed method.3)A novel method for optimal placement of micro-phasor measurement units in distribution networks for highly accurate state perception is proposed.Firstly,under the circumstance of advanced metering infrastructure meeting the requirement of network observability,the optimal placement problem is formulated by minimizing the mean squared error of state estimation while randomly selecting one operating point as parameters of covariance matrix of state estimation errors.Secondly,the objective function is expressed explicitly through decision variables by executing Cholesky decomposition on the gain matrix so that existing commercial solvers can be adopted to obtain high quality feasible solutions in an efficient manner.Finally,IEEE 33-bus and 69-bus test feeders are used to validate the effectiveness of the proposed method.