Prediction of Distribution Network Operation Trend Based on the Secondary Modal Decomposition and LSTM-MFO Algorithm

Abstract

Distribution network operation trend prediction is a key technology to analyze the network security operation status and potential hidden dangers of distribution side dynamically, how to accurately depict distribution network operation status and trend change is an important work to ensure the safe and stable operation of distribution network. In this paper, a trend prediction strategy for distribution network operation is proposed. First, fusion integration Ensemble Empirical Mode Decomposition (EEMD) and Variational Mode Decom (VMD) establish a secondary modal decomposition model of distribution network operating parameters, extract relatively stable subsequents and trend sequences, in order to reduce the impact of disordered components in high frequency sequences on predictive accuracy; Moth Flame Optimization (MFO) optimizes the LSTM parameters and uses the optimized Long-term Memory Networks (LSTM) to predict the subsethics of the operating parameters to further improve the accuracy of the distribution network operation trend prediction, and finally, the validity of the method proposed in this paper is verified in the actual power grid in a province of China. This method can accurately depict the changing trend of distribution network operation and perceive the security risks that distribution network may awareness.