Trend Predicting of Distribution Network Operation States Based on PMU

Abstract

Trend predicting of distribution network operation states is an important basis for understanding the safe operation of distribution network. In view of the impacts of measurement data and prediction accuracy, a novel trend predicting method of distribution network operation states is proposed in this paper. It is based on an improved multi-dimensional grey-neural network hybrid coordination prediction model and a branch current forward and backward substitution power flow model considering Phasor Measurement Unit (PMU). This trend predicting method analyzes the errors existing in the historical measurement data and establishes the distribution network measurement database as the input of the trend prediction. Improved multi-dimensional grey-neural network hybrid coordination prediction model is designed to achieve high-precision power prediction in the short term. The result of the power prediction is used as the input of the branch current forward and backward substitution power flow model utilizing PMU to calculate the system trend in the future and to implement the trend predicting of distribution network operation states. The feasibility and effectiveness of the proposed method are verified by analyzing the predicted and true values of IEEE33 system.