Intelligent Forecasting Model for Hydrological and Water Resources System

Based on the traditional Bayesian probability theory, the BP neural network with strong non-linear effect is introduced to improve the river flood forecasting in this paper. Considering the uncertainty of hydrological forecasting, this paper combines Bayesian method with BP neural network to study the BP neural network model of the prior distribution and likelihood function of river flood forecasting discharge, to forecast river flood more scientifically. Then, Markov chain Monte simulation method based on adaptive sampling algorithm is used to solve the posterior density of river flood forecasting discharge. Finally, the purpose of probability prediction is achieved by experimental analysis, which also provides a strong basis for hydrological forecast and management decision-making.