Water level prediction method based on deep learning mesh drainage basin submerging model

The invention relates to a water level prediction method based on a deep learning net-shaped drainage basin inundation model, and the method comprises the following steps: S1, obtaining rainwater condition data of each measurement point in a drainage basin, and carrying out the preprocessing of the rainwater condition data; s2, carrying out feature extraction and superposition operation on the preprocessed rainwater condition data; s3, according to the extracted and superposed features, carrying out training fitting based on an LSTM model; s4, searching an optimal parameter of the LSTM model trained in the step S3 by using PSO to obtain a deep learning-based mesh drainage basin submerging model; and S5, performing drainage basin water level prediction based on the deep learning net-shaped drainage basin inundation model. According to the method, inheritance of monitoring data of different measuring points can be realized, computing power resource waste caused by repeated calculation is effectively avoided, an