| abstract |
The invention discloses a method for determining the relationship between water level and flow of a river channel based on the linkage learning of multiple hydrological stations. The method includes: step 1, collecting the observation data of flow and water level of each hydrological station in the upper and lower reaches of the river where the river is located, and the influence factors of the sections where the hydrological stations are located. , construct a sample set; step 2, use the sample set to train the neural network model; step 3, input the influencing factors of the target river section into the trained neural network model, and obtain the parameter pairs A and B corresponding to the target section of the river, using The water level-discharge relationship formula constructs the water level-discharge relationship of the target section of the river. On the premise that there is no actual measurement and observation data of the target section of the river channel, the present invention realizes the efficient determination of the relationship between the water level and flow of the target channel section. |