Application of Gray Wolf Optimization Algorithm in Urban Electricity Load Forecasting Model

Abstract A combined prediction model based on long short-term memory neural network (LSTM) and convolutional (CNN) is proposed in order to increase the accuracy of load. To address issue that gray wolf optimization (GWO) search process prone falling into local optimum. An improved grey algorithm (IGWO) update convergence factor using lower incomplete gamma function improve global performance. The Dropout technique used generalization ability model; layers are by increasing initialization weights; trained an adaptive moment estimation (Adam) optimizer, test data input model, finally, optimized for prediction. high method demonstrated experimentally.