基于RAGA的分布式水文模型参数分类优化方法研究  

Research on Parameter Classification and Optimization Method of Distributed Hydrological Model based on RAGA

作　　者：周祖昊[1] 刘清燕 韦瑞深 刘佳嘉[1] 严军[3] 王鹏翔 贾仰文[1] 王浩[1] ZHOU Zuhao;LIU Qingyan;WEI Ruishen;LIU Jiajia;YAN Jun;WANG Pengxiang;JIA Yangwen;WANG Hao(State Key Laboratory of Simulation and Regulation of Water Cycle,China Institute of Water Resources and Hydropower Research,Beijing 100038,China;Shandong Shuifa Technology Group Co.,Ltd,Jinan 250101,China;North China University of Water Resources and Electric Power,Zhengzhou 450046,China;Wuhan University,Wuhan 430072,China)

机构地区：[1]中国水利水电科学研究院流域水循环模拟与调控国家重点实验室,北京100038 [2]山东水发技术集团有限公司,山东济南250101 [3]华北水利水电大学,河南郑州450046 [4]武汉大学,湖北武汉430072

出　　处：《水文》2023年第3期28-33,共6页Journal of China Hydrology

基　　金：“十三五”国家重点研发计划课题(2016YFC0402405);“十四五”国家重点研发计划课题(2021YFC3000205)。

摘　　要：计算效率低及异参同效(局部优化)是分布式水文模型参数优化研究中存在的主要问题。本文基于RAGA(基于实数编码的加速遗传算法),提出一种分布式水文模型参数分类优化方法,将需要率定的参数根据物理意义分成若干类,逐类进行优化。这种方法可降低待优化参数的维度,一方面可以提高优化计算的速度,另一方面可以在一定程度上逼近全局最优,减少异参同效的问题。本文采用分布式水文模型WEP-L(Water and Energy Processesin Large Scale Basin)模型,针对黄河流域玛曲水文站以上区域1997—2000年逐月流量过程进行参数率定,并对2006—2016年系列进行验证。对比参数不分类优化方法,发现采用参数分类优化方法后,WEP-L模型参数优化的速度提高37%左右,纳什效率系数(NSE)从0.739提高到0.829。说明参数分类优化方法既可以节约时间,又可以保证优化算法的全局性,提高模拟的精度。The main problems in parameter optimization of distributed hydrological model were low computational efficiency and the same effect of different parameters(local optimization).Based on RAGA(Accelerated Genetic Algorithm Based on Real Number Coding),this paper proposed a distributed hydrological model parameter classification optimization method,that is,the parameters needed to be calibrated and were divided into several categories according to their physical meanings,and optimized category by category.This method could reduce the dimensions of the parameters optimization.On the one hand,it could increase the speed of optimization calculation,on the other hand,it could approach the global optimization to a certain extent and reduce the problem of different parameters with the same effect.This paper adopted the distributed hydrological model WEP-L(Water and Energy Processes in Large Scale Basin)model to simulate and calibrate the monthly discharge process of the area above the Maqu hydrological station in the Yellow River Basin from 1997 to 2000,and verified the series from 2006 to 2016.Compared with the parameter optimization methods without classification,it is found that the speed of WEP-L model parameter optimization increased by about 37%,and the Nash efficiency coefficient(NSE)increased from 0.739 to 0.829 after using the parameter classification optimization method.It showed that the parameter classification optimization method can not only save time,but also ensure the globality of the optimization algorithm and improve the accuracy of simulation.

关 键 词：分布式水文模型 参数优化 参数分类 RAGA 

分 类 号：TV213.2[水利工程—水文学及水资源] P338[天文地球—水文科学]