Application of Nuclear magnetic resonance Ultra-short-term load forecasting model based on digital twin

Abstract

With the increasing complexity of power system and the improvement of demand for power consumption quality, in order to further improve the accuracy of load forecasting, this paper proposes the ELMAN neural network and the modified model based on digital twin for ultra-short-term load forecasting. Firstly, the ELMAN neural network ultra-short-term load prediction model was built, and the neural network model was optimized by immune particle swarm optimization. Secondly, according to the information of digital twin platform, the most similar days are found and the load prediction results are revised. Finally, the power load of a city in China is taken as an example and verified in two scenarios: winter and summer. The results show that the model proposed in this paper can effectively improve the accuracy of load prediction.