GA-CNN-LSTM Based Temperature and Humidity Prediction Method

In order to avoid the significant impact of the “three highs” climate on the operation of power system equipment, this paper introduces genetic algorithm (GA) into the hybrid prediction model structure of convolutional neural network (CNN) and multi-layer long short-term memory (LSTM) network, and proposes GA-CNN-LSTM model to achieve the prediction of temperature and humidity. Genetic algorithm is used to guide the selection of hyperparameters for the prediction model to ensure that temperature and humidity predictions are made with the optimal combination of hyperparameters. The experimental section compares the prediction performance of the proposed method with back propagation(BP) neural network, traditional LSTM model, and CNN-LSTM model. The research results show that the GA-CNN-LSTM method exhibits the highest accuracy in predicting temperature and humidity.