Complex subway station foundation pit deformation space-time prediction method

The invention discloses a complex subway station foundation pit deformation space-time prediction method, which comprises the following steps: S1, obtaining foundation pit deformation data, and preprocessing the foundation pit deformation data to obtain a training set and a test set; and S2, constructing a neural network model of GCN-GRU according to the training set and the test set, obtaining a prediction result of the neural network model, and completing deformation prediction of the complex subway station foundation pit site. The invention provides a complex subway station foundation pit deformation space-time prediction method, and solves the problems that a previous complex subway station foundation pit deformation prediction model can only excavate time information or a future development trend of deformation data of a single foundation pit deformation measuring point, and cannot fully utilize a complex dependency relationship of a plurality of foundation pit deformation measuring ...