Load Prediction of Composite Curing Equipment Cluster Based on Combined EEMD-WOA-LSTM Model

For groups utilizing composite curing equipment, load forecasting is crucial in ensuring productivity, optimizing energy consumption, and reducing costs. This paper aims to enhance the accuracy of load forecasting through the utilization of artificial intelligence algorithms. Neural networks with long shortterm memory (LSTM) are integrated into a combined model for short-term load forecasting, which is based on whale bionic optimization. The whale bionic algorithm is applied to improve parameter optimization accuracy and address the issue of local optimization within long-short-term memory neural networks. Multiple feature components are extracted from the source signal using empirical mode decomposition. The bionically optimized combined model utilizes each feature component as input for prediction, ultimately resulting in improved load prediction results. The enhanced combination model using whale bionics demonstrates reduced prediction error and greater accuracy when compared to the prediction results of RNN model, LSTM model, WOA-LSTM model, and EEMD-ARMA model.