Conv-Transform-based flood forecasting method

Abstract

The invention belongs to the technical field of data-driven water flow forecasting, and discloses a Conv-Transformer-based flood forecasting method, which comprises the following steps of: firstly, collecting hydrological data of a researched large-scale watershed, and then inputting the collected hydrological historical data into a model after data preprocessing; secondly, performing data cleaning, data conversion, data set division and the like on the hydrological historical data; thirdly, constructing a flood forecasting model based on Transform, performing relative position coding on the model by using a long-short-term memory network based on convolution operation to extract spatial information, and improving the learning ability of the model on long-term dependency information; a self-attention mechanism in the Transform module can capture dynamic space-time correlation among hydrological elements by capturing internal correlation of a hydrological sequence, and a multi-attention mechanism in the Transform module enables the model to learn long-term and short-term hydrological historical information at the same time; inputting test data to test the performance of the forecasting model, judging whether the network performance meets requirements or not, and if not, adjusting parameters until an ideal forecasting result is achieved; and finally, the model is analyzed through an evaluation standard, and flood forecasting is completed. The method has the beneficial effects that the flood peak precision and the flood trend can be effectively forecasted, and the method is an effective tool for real-time forecasting of large-scale basin flood.

Description

technical field [0001] The invention relates to the technical field of data-driven flood forecasting, in particular to a Conv-Transformer-based flood forecasting method. Background technique [0002] Flood forecasting is one of a series of important non-engineering measures to prevent flood disasters. Timely and effective flood warning and forecasting can help humans effectively prevent floods and reduce flood losses. It is an important disaster prevention and mitigation application. [0003] At present, flood forecasting generally adopts two methods: hydrological model based on runoff process and data-driven intelligent model, and the two models complement each other in actual forecasting. Data-driven modeling basically does not consider the physical mechanism of the hydrological process, and is a black-box method with the goal of establishing the optimal mathematical relationship between input and output data. Floods caused by prolonged rainfall need to consider long-term...

AI Technical Summary

Problems solved by technology

However, the linear transformation of absolute coding is easy to lose the position information, and it is difficult to control the relative distance information of various hydrological features, which is not conducive to the extraction of hydrological spatial information.

Images