Cloud Computing-Based Graph Convolutional Network Power Consumption Prediction Method

With the continuous increase of electricity consumption data in smart grids, the data storage and data analysis capabilities of traditional single-node data mining algorithms can no longer meet the requirements of electricity consumption forecasting. This paper designs a electricity consumption forecasting method based on cloud computing and graph convolutional network. The method first proposes a GCN-based electricity consumption forecasting model, then builds a Hadoop platform, uses MapReduce to parallelize and iteratively trains the GCN model on the platform, and then uses the trained model to predict electricity consumption. In addition, the prediction accuracy of this method is verified through experiments.