基于最小二乘法及信息熵的短期水文预报组合模型的比较  

Comparison of Short-term Hydrological Combination Forecasting Models Based on Least-squares Algorithm and Entropy Theory

作　　者：王婷婷[1,2] 周建中[1,2] 张勇传[1,2] 陈璐[1,2] 吴江[1,2] 

机构地区：[1]华中科技大学水电与数字化工程学院,湖北武汉430074 [2]华中科技大学数字流域科学与技术湖北省重点实验室,湖北武汉430074

出　　处：《水电能源科学》2015年第10期13-17,共5页Water Resources and Power

基　　金：国家自然科学基金重点项目(51239004)

摘　　要：为探讨组合预报中平均加权法的实用性,以三峡流域宜昌站为例,建立了基于最小二乘法和信息熵理论的两种径流组合预报模型,并将组合预报结果与原单一模型的预报结果做了对比,选取具有代表性的预报精度评定指标检验预报精度。结果表明,两种组合预报模型均显著改善了预报精度评定指标,提高了短期水文预报精度,突破了传统单一水文预报模型的局限性,实际应用时可根据预报精度评定的侧重点选择合适的组合预报模型。In order to verify the performance of combination forecast,Three Gorges Reservoir is selected as research object.We established two combination forecasting models based on least-squares algorithm and entropy theory to predict the discharge at Yichang gauging station.Besides,typical precision evaluation indexes are chosen to check whether the prediction accuracy of the hydrological forecasting models is improved or not through comparison of combination forecasting results with the original single model.The results show that both of combination forecasting models can improve the prediction accuracy of short-term hydrological forecasting,and break the limitations of the traditional single forecast models.It can choose the appropriate combination forecast model in the practical application according to the priorities of forecast accuracy evaluation.

关 键 词：组合预报 最小二乘法 信息熵理论 短期径流预报 

分 类 号：TV124[水利工程—水文学及水资源]