A Predictive Model for Coal Consumption and Raw Feed Volume Based on DCCNN-sLSTM

As a high energy consumption industry, cement is facing the double pressure of energy saving, environmental protection and carbon emission indicators. The prediction of coal consumption and raw feed volume in the cement production process can reasonably carry out the raw material ratio and reduce the waste of energy, so as to achieve the purpose of “energy saving and consumption reduction, quality and efficiency improvement”. Therefore, a dual-channel DCCNN-SLSTM model based on convolutional neural network (CNN) and long short-term memory (LSTM) network is proposed to predict coal consumption and raw feed volume simultaneously. First, a dual-channel feature extraction structure of coal consumption and raw feed volume is constructed to predict coal consumption and raw feed volume simultaneously, which effectively solves the problem of parameter redundancy caused by excessive correlation variables in the cement production process and reduces the complexity of the model. Second, the time series data of cement production process are input into CNN model and sLSTM model to extract temporal and spatial characteristics to ensure the prediction accuracy of coal consumption and raw feed volume. Finally, the experimental results of a cement plant in Gansu Province are used to verify the model. The results show that the prediction performance of the proposed model is better than other comparison models, and it has better prediction accuracy and generalization ability.