Research on Load Forecasting Method of Distribution Transformer based on Deep Learning

Electrical load forecasting is of great significance, which can guarantee the system stability under large disturbances and optimize the distribution of energy resources in the smart grid. Load forecasting of distribution is a very active research field and an important aspect, as the load power of distribution transformer is closely related to the health and reliable of the power grid. Deep Learning technology have shown to be very powerful in load forecasting of distribution. Based on the deep learning technology, this paper uses the historical load data of distribution transformer to establish an appropriate prediction model to predict the future load of transformer. In this paper, the long short-term memory(LSTM) model is used to predict the distribution transformer load. The idea of integrated learning is used to improve the accuracy of LSTM network in load forecasting. The input dimension of the model is enriched to solve the problem that the model cannot be built due to the lack of historical data. Finally, the experimental results show the superiority of the proposed method.