本文首先采用数据驱动模型获取长江中游沿程入汇流量,再依据相关关系法对中游短期水位进行预测预报,并针对监利、武汉关、九江等重要站点预测水位影响因素较多的特性,采取多元回归模型去综合考虑多因素与变量的线性关系,构建精度更佳的水位预测模型。 In this paper, the data-driven model was used to get the inflow of the Yangtze River in the middle reaches of the Yangtze River. Then the short-term water level in the middle reach was predicted and predicted according to the correlation method. According to the characteristics of water level prediction in important sites like Jianli, Wuhan and Jiujiang, Regression model to comprehensive consideration of the linear relationship between variables and variables to build a better accuracy of the water level prediction model.