System and method for predicting typhoon surge based on artificial intelligence

Abstract

The application discloses an intelligent method for predicating surge when a typhoon occurs at a land stream hydrometric station (target reading station), and information of typhoon high tide surge and high tide level when the typhoon occurs at the target reading station after 12 hours can be intelligently predicted. The method includes prediction parameters selected for predicting the typhoon surge and comprehensively influencing the march route and intensity of the typhoon, and parameters influencing height of storm tide of the typhoon, wherein the prediction parameters consist of real-time typhoon observation data, longitude, latitude, central pressure of the typhoon, the maximum wind speed, move speed, move direction of the last typhoon observation data (before 6 hours) before the real-time observation time, information of upstream inflow of a river (observation surge information 12 hours before the target prediction time of an upstream reading station located in an upstream area of the river target reading station), and historic surge information of the target reading station (including surge information of two high tides in 24 hours before the target prediction time). The historic typhoon surge information of the target reading station is simulated through a mathematical model used for predicting typhoon surge and established through an artificial neural network, then the real-time typhoon surge information as well as the historic typhoon surge information of the upstream reading station and the target reading station are input, so that the typhoon surge information of the target reading station at the target observation time can be predicted.

Description

technical field [0001] This application relates to typhoon water increase prediction technology in the field of hydrology. Background technique [0002] Existing typhoon flood forecasting techniques mainly include empirical statistical methods and numerical calculation methods. Empirical statistical method is based on historical data to establish the relationship between predictors and various hydrometeorological elements, the main means is the use of regression and statistical correlation methods. The numerical calculation forecasting method is to determine the future state of the ocean based on the solution of fluid dynamics and thermodynamic equations. The numerical calculation method can reveal and analyze the physical essence of the ocean and atmospheric changes, and requires extensive knowledge of mathematics, physics, and oceans. Both the empirical pre-statistical method and the numerical calculation forecasting method need to consider a large number of meteorologica...

AI Technical Summary

Problems solved by technology

However, empirical forecasting has high requirements on the personal professional quality of the forecaster, and is easily affected by subjective factors.

Therefore, there is still a gap in the use of artificial intelligence methods to predict typhoon water increase.

Images