Short-Term Load Prediction Using CNN-LSTM Network and Sparrow Search Algorithm for Combined Cooling Heating and Power Systems

In combined cooling heating and power (CCHP) systems, the interaction and impact between cooling, heating and electrical energy play a vital role, and accurate load forecasting is the key to economic dispatch and efficient operation of CCHP systems. In this paper, based on convolutional neural networks-long short-term memory network (CNN-LSTM) model and sparrow search algorithm (SSA), the short-term multiple load prediction of CCHP system is studied. First, the spatio-temporal features of the load data are extracted using CNN and LSTM respectively, and the CNN-LSTM network model is constructed. Subsequently, the SSA is used to optimize the learning parameters of the CNN-LSTM network model, namely, learning rate, training times and number of neurons of the LSTM. Simulation results show that compared with other existing prediction techniques, the proposed method can achieve better prediction accuracy in cooling, heating and electrical energy, and mean absolute error and mean absolute percentage error are reduced, which testify the effectiveness and feasibility of the proposed method.