基于BP神经网络的贝叶斯概率水文预报模型

本文在贝叶斯概率水文预报系统(BFS)框架之上，研究了双牌水库水文预报的不确定性，建立了流量先验分布及似然函数的BP神经网络模型，并通过Markov链Monte Carlo(MCMC)方法求解得到流量后验分布及其统计参数。通过对双牌水库历史洪水的研究结果表明，基于BP神经网络的BFS不仅显著提高了预报精度，而且为防洪决策提供了更多的信息，使得预报人员在决策中能考虑预报的不确定性，定量的估计各种决策的风险和后果。

Bayesian probabilistic forecasting model based on BP ANN

Based on the Bayesian Forecasting System(BFS) framework,a new prior density and likelihood function model using BP artificial neural network(ANN) is developed to study the hydrologic uncertainty of the Shuangpai Reservoir,China.The Markov chain Monte Carlo method is applied to solve the posterior distribution and statistics of reservoir stage.The study result of the floods in history shows that Bayesian probabilistic forecasting model based on BP ANN not only remarkably improves the forecasting precision but also offers more information for flood control,which makes it possible for decision makers to consider the uncertainty of hydrologic forecasting during decision-making and estimate the risks of different decisions quantitatively.