Partition Method of Power System State Estimation Parallel Computing Based on Substations

For distributed and parallel power systems state estimation, a partition method is proposed with the goal of minimizing the total computation time. Firstly, the structure of the gain matrix in the Weighted Least Squares (WLS) is analyzed. Secondly, in the processing of substations data based on sparse matrix technique, the calculating path is analyzed; According to the path set of calculation and the location of substations, a time-consuming model of distributed and parallel state estimation is established. Taking into account the balance of computational burden, time consumption and resource allocation, the method presented in this paper can give a better partition scheme, and simulation results turn out it is practical and easy to implement.