Flood forecasting method based on data excavation-similarity theory

Abstract

The invention discloses a flood forecasting method based on the data excavation-similarity theory. The method comprises steps of carrying out data excavation on history data of a forecasting station; classifying flood of history periods; analyzing the type to which the to-be-forecasted flood belongs; carrying out similarity comparison on collected flow data of the to-be-forecasted flood at the initial period and the flood of history periods in the belonged type; and searching the flood of the history periods which has most similar characteristics with to the to-be-forecasted flood to serve as a flood forecasting result. According to the invention, a disadvantage that the current flood forecasting method is not suitable for drought / half-drought regions is overcome; and based on data excavation and similarity, the precision of the flood forecasting results of regions of this type is effectively improved.

Description

technical field [0001] The invention relates to a flood forecasting method based on data mining-similarity theory, in particular to a method for performing related flood forecasting by analyzing the similarity between historical floods of sites and characteristics of floods to be forecasted. Background technique [0002] Since the Horton runoff and the Nash unit line were put forward, the theory and technology of flood forecasting have been greatly developed. At present, commonly used flood forecasting methods can be summarized into two categories, one is based on data-driven flood forecasting models (such as correlation analysis, linear regression, etc.), and the other is flood forecasting models driven by physical processes (such as Xin'anjiang model, TOPMODEL wait). Both types of models have been developed rapidly and played an important role in actual production. [0003] However, existing hydrological models are more suitable for flood simulation and forecasting in hu...

AI Technical Summary

Problems solved by technology

[0003] However, existing hydrological models are more suitable for flood simulation and forecasting in humid areas. In arid / semi-arid areas with more complex hydrological processes, the forecast accuracy is often low, and it is difficult to meet the needs of flood control and disaster reduction in such areas.