Water area detection and early warning method based on graph neural network

Abstract

The invention relates to the technical field of water area detection and relates to a water area detection early warning method based on a graph neural network. The method comprises the following steps of: (1) data acquisition: performing water surface cruising in a to-be-detected water area by adopting an overwater robot, and acquiring water quality and environmental data of the to-be-detected water area; (2) constructing a graph neural network model of the to-be-measured water area; and (3) water quality condition prediction: learning and predicting the water quality condition of the to-be-detected water area by utilizing the graph neural network model of the to-be-detected water area. A pollution source in the to-be-detected water area is positioned by adopting the overwater robot in the prior art, and accurate prediction of the water quality of water areas with different shapes and sizes is completed by utilizing spatial position information of the pollution source and adopting a method of constructing the graph neural network model and learning the graph neural network model. Compared with an existing detection and early warning method of a fixed monitoring station, the methodhas better universality.

Description

Technical field [0001] The invention relates to the technical field of water area detection, in particular to a water area detection and early warning method based on graph neural network. Background technique [0002] As an indispensable precious resource, water pollution is closely related to the lives and health of the people. [0003] The level of intelligence in water pollution detection is low. Water quality monitoring equipment collects water quality data from discrete sites, and can only collect data from nearby waters. It cannot locate the source of water pollution and cannot make an accurate assessment and prediction of water quality. [0004] The water environment is open and complex, the water quality factors show nonlinear changes, and the interaction between the various influencing factors is complex, and there are many coupling factors. How to construct a high-precision prediction and early warning model of water quality parameters has very important theoretical value...

AI Technical Summary

Problems solved by technology

[0006] Existing detection and early warning methods for water areas are mainly based on traditional prediction methods such as multivariate discriminant analysis, support vector machine, fuzzy comprehensive evaluation, and neural network. Accurate prediction of water quality conditions for a period of time in the future

Images