Short-term load prediction method and system based on ARIMA and CNN-LSTM combined model

The invention discloses a short-term load prediction method and system based on an ARIMA and CNN-LSTM combined model. The method comprises the following steps: collecting historical load data and meteorological data of a park, carrying out data preprocessing on original information, randomly dividing the processed data into a training set and a test set, and setting a proportion to obtain a training data sequence; constructing a CNN-LSTM prediction model, inputting the training data sequence into CNN convolutional layers, capturing local and global features in data through multiple convolutional layers and a pooling layer, and transmitting the local and global features to LSTM for capturing a time dependency relationship in two-dimensional time sequence data; constructing an ARIMA prediction model, extracting a residual error from the training data sequence, and inputting the residual error into an LSTM for correction; and combining prediction values of the CNN-LSTM prediction model and the ARIMA prediction m