Cooling, Heating and Electric Load Forecasting for Integrated Energy Systems Based on CNN-LSTM
Liu Wenya
Abstract
As “carbon peak and carbon neutrality” have become an important strategic goal of my country's “14th Five-Year Plan”, it is particularly important to accelerate the adjustment and optimization of the energy structure. Forecasting cooling heating and power load of integrated energy system can effectively improve energy utilization efficiency. Aiming at the problem of low prediction accuracy about a single load forecasting method, this paper proposes that a short-term cooling, heating and electric load forecasting method base on CNN-LSTM(Convolution Neural Network-Long Short Term Memory). First, the Pearson correlation coefficient method is used to analyze the correlation between cooling load, heating load and electric load. Then the Z-Score standardization method is used to standardize the data. The standardized data input into a single CNN(Convolution Neural Network) model, a single LSTM(Long Short Term Memory) model, and a CNN-LSTM combined model for cooling heating and electrical load prediction. The results show that CNN-LSTM model has a higher prediction accuracy.