A water velocity prediction method based on multi-feature data and multi-model

Abstract

The invention discloses a water flow velocity prediction method based on multi-feature data and multi-model, comprising: step 1, collecting the roughness of the river bed surface, the size of the slope of the water surface, the regularity of the shape of the section, the size of the sediment concentration of the water flow, and the distance of the measuring point The distance of the bed surface is used as the first feature; step 2, the change of the river level is collected as the second feature; step 3, the Euclidean distance of the corresponding feature points of two different river trajectories is calculated; step 4 is the first The feature, the second feature and the third feature are assigned corresponding weights, and the weight value of the second feature is greater than the weight value of the first feature and the third feature; step 5, the first feature, the second feature, the third feature and their The corresponding weight values ​​are input into the flow rate prediction model, and the water flow rate is output. The present invention fully considers the change of river level and the influence of the indirect effect between rivers on the water flow velocity, comprehensively considers the characteristic data of multiple dimensions, and inputs different data into the corresponding optimal algorithm for calculation, which improves the flow velocity prediction Model precision and accuracy.

Description

technical field [0001] The invention relates to the field of fluid velocity prediction, in particular to a water flow velocity prediction method based on multi-feature data and multi-models. Background technique [0002] The traditional water flow velocity prediction method mostly collects a small amount of data such as the roughness of the current bed surface of the river bed, the size of the water surface slope, the regularity of the cross-section shape, the sediment content of the water flow, and the distance from the measuring point to the bed surface, etc. The data is stored in the database MySQL. , and then use the trained single model to statistically analyze the data to realize the prediction of water flow speed. However, this method has a single source of data and a single model, and does not take into account the changes in channel levels and the effects of indirect interactions between channels on water flow velocity. Therefore, the existing models have low predic...

AI Technical Summary

Problems solved by technology

However, this method has a single source of data, a single model, and does not consider the changes in the level of the river and the indirect effects between the rivers on the flow velocity, so the existing models have low prediction accuracy and low accuracy.

Images