Water quality data prediction method based on improved Re-GCN model

The invention discloses a water quality data prediction method based on an improved Re-GCN model, and the method comprises the following steps: obtaining and preprocessing water quality data, and constructing a water quality data time sequence knowledge graph training set and a test set for training and testing; performing hyper-parameter tuning by using grid search, inputting the training set into a pre-constructed improved Re-GCN model for iterative training, calculating a loss function, and performing global optimization on the basis of the loss function until iteration is finished; importing the test set into the trained model, evaluating the model through an evaluation index until a result meets an expected requirement, and outputting a prediction model; and using the prediction model to predict water quality data in a period of time in the future. Through the improved Re-GCN model, the water quality data prediction efficiency and accuracy are improved. 本发明公开基于改进的Re-GCN模型的水质数据预测方法，包括如下步骤：获取水质数据并进行预处理，构建用