Flood peak evolution forecasting method based on rainfall and water level feature similarity analysis

The invention relates to a flood peak evolution forecasting method based on rainfall and water level feature similarity analysis, and solves the problems that the current flood peak evolution forecasting is mainly based on hydrodynamics and hydrological model calculation, has high requirements on computing power and an algorithm and cannot well meet the requirement of rapid simulation calculation in actual flood control and disaster reduction work. The method comprises the following steps: firstly, collecting and processing watershed hydro-meteorological data; dividing a watershed section and constructing a watershed flood peak transmission topological relation; 2, analyzing and calculating historical hydrological and meteorological data characteristics of each section to form a database; and 3, searching a historical flood session similar to the current session of flood, and finally, carrying out flood peak evolution forecasting section by section based on similarity analysis of historical rainfall and flood