汉口水文站AI流量模型构建及应用研究  

Research on the Construction and Application of AI Discharge Model for Hankou Hydrological Station

作　　者：周波 邓山 陆鹏程 毛北平 魏猛 ZHOU Bo;DENG Shan;LU Pengcheng;MAO Beiping;WEI Meng(Hydrology Bureau of Yangtze River Water Resources Commission,Wuhan 430010,China;Department of Hydrology,Ministry of Water Resources,Beijing 100053,China;Middle Yangtze River Hydrology and Water Resources Survey Bureau,Hydrology Bureau of Yangtze River Water Resources Commission,Wuhan 430014,China)

机构地区：[1]长江水利委员会水文局,湖北武汉430010 [2]水利部水文司,北京100053 [3]长江水利委员会水文局长江中游水文水资源勘测局,湖北武汉430014

出　　处：《水文》2024年第5期1-9,105,共10页Journal of China Hydrology

基　　金：国家重点研发计划项目(2022YFC3204502);长江委科技人才创新团队项目(水文水利智能感知)。

摘　　要：为进一步提升水文站实时推流的稳定性及精度,在开展测站特性分析的基础上,提出一种基于人工智能算法的河流流量在线推求新方法。经汉口水文站实践应用表明,无论是建模精度还是泛化精度,均达到了一类精度站水平,且具有较高稳定性。研究成果具有较高的推广价值,可为水文测站业务的智能化升级提供科学参考。To further improve the stability and accuracy of real-time flow prediction at hydrological stations,a new method for online river flow prediction based on artificial intelligence algorithms was proposed based on the analysis of station characteristics.The practical application of Hankou Hydrological Station has shown that both modeling accuracy and generalization accuracy have reached the level of a first-class accuracy station with high stability.The research results are well worth spreading and can provide scientific references for the intelligent upgrading of hydrological station services.

关 键 词：数字孪生 人工智能 河流流量 精度评价 

分 类 号：P332[天文地球—水文科学]