论文部分内容阅读
针对传统月径流预报模型存在的缺陷,建立了相似过程衍生法与概率预报相结合的月径流概率预报模型。运用相似过程衍生法发布确定的预报结果,在定点预报的基础上利用概率预报提供一定置信水平下的预报区间作为模型预报结果。模型结构简单、易于构建且建模过程中无需考虑预报因子的选择问题。将该模型与BP神经网络模型进行对比仿真试验,结果表明该预报模型具有较好的预报精度,且合格率高于BP神经网络模型,可在水库月径流预报中推广应用。
Aiming at the shortcomings of the traditional monthly runoff forecasting model, a monthly runoff probabilistic forecasting model combining the similar process derivation and probability forecasting is established. The method of similar process derivation is used to release the definite forecast results. Based on the fixed point forecast, probabilistic forecasting is used to provide the forecasting interval with a certain confidence level as the model forecasting result. Model structure is simple, easy to build and modeling process without considering the selection of the forecasting factor. Comparing the model with the BP neural network model, the simulation results show that the forecasting model has better forecast accuracy and the qualified rate is higher than that of the BP neural network model, which can be widely applied in the reservoir monthly runoff forecasting.