摘要
径流中长期预测是水利水电工程设计、施工和运行管理的重要依据,而预报因子的选取和预报模型的构建一直是径流中长期预测的两大难点。本文以宜昌站1951—2013年汛期流量资料为例,根据国家气候中心74项环流特征量,分别构建了基于相关系数法和随机森林算法的径流预测遥相关模型,并且对比了依据不同的相关系数和因子选择数目对预测精度的影响。结果表明,相较于传统的相关系数法,基于随机森林算法的预测模型的稳健性与预测精度均可大大提高。研究内容可为三峡水库中长期径流预报及中长期调度计划制定提供技术支撑。
The medium and long-term runoff forecast is an important basis for water conservancy and hydropower engineering design, construction and operation management. Predictors-selection and model construction are two key issues in the mid-long term runoff fore- casting. In this paper, taking the flow data of Yichang station during the flood season from 1951 to 2013 for an example, based on the 74 circulation characteristics of the National Climate Center, the runoff prediction models based on the correlation coefficient method and the stochastic forest algorithm are constructed respectively. And the influence of different correlation coefficient and factor selection number on prediction accuracy is compared. The results show that the robustness and prediction accuracy of the prediction model based on stochastic forest algorithm can be greatly improved compared with the traditional correlation coefficient method. The study can pro- vide technical support for the long-term runoff forecast and the long-term scheduling plan of the Three Gorges Reservoir.
作者
董前进
张旭
DONG Qianjin ZHANG Xu(State Key Laboratory of Water Resources and Hydropower Engineering Science, Wuhan University, Wuhan 430072, China Key Laboratory of Poyang Lake Wetland and Watershed Research, Ministry of Education, Jiangxi Normal University, Nanchang 330022, China Hubei Provincial Collaborative Innovation Center for Water Resources Security, Wuhan University, Wuhan 430072, China)
出处
《华北水利水电大学学报（自然科学版）》
2016年第6期38-42,共5页
Journal of North China University of Water Resources and Electric Power：Natural Science Edition
基金
国家自然科学基金资助项目(51439007
51190094)
江西省重大生态安全问题监控协同创新中心资助项目(JXS-EW-00)
关键词
中长期径流预测
遥相关
随机森林算法
多元线性模型
宜昌站
mid-long term runoff prediction
hydro-climate teleconnection
random forest algorithm
multi-liner regression
Yichang station
