Dynamic state estimation method of distribution network based on partition of AMI total measurement points

Abstract

Aiming at the problems of slow calculation speed and low estimation accuracy of traditional three-phase unbalance dynamic state estimation of distribution network，a dynamic state estimation method based on partition of advanced metering infrastructure （AMI） total measurement points is proposed. Taking the AMI total measurement points as the partition nodes of distribution network，the partition objective function integrating three indexes is put forward to partition the distribution network，which can completely decouple the sub-regions and reduce the system scale. The multi-scale measurement data is fused through the proposed data fusion framework. Based on the measurement cycle of the remote terminal unit，the AMI measurement data with a long measurement cycle is fused and the system state at non-AMI measurement time is followed. A high-precision ensemble Kalman filtering algorithm based on sub-region data fusion is proposed and the covariance expansion method is used to improve the divergence problem of filter. The simulative results show that the proposed method can effectively improve the calculation speed and estimation accuracy of distribution network dynamic state estimation.