Research On Power Network Topology Recognition Method Based On Measurement Data

Power network topology identification is an important module of energy management system and distribution management system.It provides network structure data for power flow calculation,state estimation,fault location and reactive power optimization.It is the basis of power system analysis software.With the development of smart grid,the network topology of power system is becoming more and more complex.At the same time,with the increasing number of monitoring devices in the network,more and more power system data can be obtained.How to use smart grid monitoring data to analyze the network topology structure is the focus of this paper.Firstly,the development and research status of power grid topology identification at home and abroad are introduced,and the theoretical and practical basis of power grid topology identification is explored.The essence of power network topology analysis is the identification of network connectivity,which is essentially a mathematical problem developed on the basis of graph theory.Therefore,the knowledge of graph theory applied in power network topology identification is briefly introduced.The topological structure of power grid is analyzed,especially the admittance matrix of power grid.It is concluded that the admittance matrix is a sparse matrix.On this basis,the sparse technology is introduced into the identification of power grid topology.Secondly,according to the characteristics of sparse connection in large transmission network,a topology identification method based on sparse restoration is proposed.Based on the AC power flow model,a DC power flow model suitable for transmission network topology restoration is deduced.The sparse restoration problem is equivalent to the l₁ norm problem,and the l₁ norm problem is solved by the basis tracking algorithm.Considering that the elements near the diagonal line of admittance matrix are more likely to be non-zero elements,the elements near the diagonal line are weighted to improve the recovery rate of the elements near the diagonal line by using the basis tracking algorithm.In the verification process,the base tracking algorithm and the weighted l₁ norm algorithm are applied to the restoration of the IEEE-30 and the IEEE-300 node topology network,and the validity of this method is verified.Finally,considering the complex structure and various operation modes of distribution network,a new method of distribution network topology identification under uncertain prior information is proposed.Based on Bayesian network theory,the distribution network topology is initialized by the prior connection probability provided by experts;the voltage data is selected as the characteristic data of the topology identification,and the method of balanced three-phase voltage solution under unbalanced three-phase load is deduced;the bad data of measurement is detected,and the autovariance and covariance are obtained when they exceed each other.When setting the threshold value,the data at that time are eliminated,and the structure description length measure is proposed as the evaluation index of the stable state of the distribution network structure,and the simulated annealing algorithm is used to solve the distribution network topology.Finally,the effectiveness of this method is verified based on the IEEE-33 bus system and a practical distribution network.