Mid-term runoff forecasting method considering multiple time sequence process factors

Abstract

The invention discloses a mid-term runoff forecasting method considering multiple sequential process factors, and belongs to the technical field of hydrological forecasting. The method comprises the following steps: firstly, determining the types of prediction alternative factors according to the data obtainable condition and prediction experience of a research basin, selecting the prediction alternative factor with the maximum correlation coefficient from each type of prediction alternative factors as a prediction factor, taking the prediction alternative factor as an input to establish a BPneural network runoff prediction model, and calculating the root-mean-square error of a prediction result; secondly, sequentially increasing the number of time-interval-by-time factors, respectively establishing BP neural network runoff forecasting models for forecasting, calculating a root-mean-square error of a forecasting result, and comparing the root-mean-square error with the forecasting result in the last step; thirdly, performing multiple combinations on the forecasting factors; and finally, determining the optimal combination of the forecasting factors suitable for the research basinaccording to the principle of minimum root-mean-square error, and establishing a runoff forecasting model. The process is simple, the forecasting method is easy to operate, the runoff forecasting precision can be effectively improved, and the purpose of improving the forecasting effect is achieved.

Description

technical field [0001] The invention belongs to the technical field of hydrological forecasting, and relates to a medium-term runoff forecasting method considering multiple time series process factors. Background technique [0002] Mid-term runoff forecasting can help management decision-makers to grasp the situation of runoff in advance, and it is an important basis for reservoir scheduling, power generation plan formulation, and rational allocation of regional water resources. In recent years, with the rapid development of meteorological and hydrological forecasting, the development of medium-term runoff forecasting is also very rapid. However, due to the uncertainty of hydrological phenomena and the difficulty in determining predictors, medium-term runoff forecasting still has the problem of low forecasting accuracy. How to find a suitable method to improve the accuracy of medium-term runoff forecast is still of great concern. [0003] The method of forecasting using da...

AI Technical Summary

Problems solved by technology

However, runoff is the final manifestation of the superposition of different processes of rainfall and other predictors. The difference in the process is likely to have an impact on the forecasted runoff. Only non-process factors that reflect the cumulative amount cannot reflect the difference in the process, and it may be difficult to meet the accuracy requirements of runoff forecasting.

Images