数字孪生永宁江洪水预报模型构建及系统应用

Construction of flood forecast model for digital twin Yongning River and system application

摘要:

以永宁江流域作为试点，构建数字孪生流域，建设具有“四预”功能的“2+N”业务应用体系，综合保障区域防洪安全。考虑到流域内地势起伏大、用地类型复杂、横跨多个行政区块等客观原因，为提高预报精度，基于 Mike 系列模型，建立永宁江洪水预报模型。采用 2013 年“菲特”、2015 年“苏迪罗”和 2019 年“利奇马”等台风洪水作为计算场次，根据 2013—2019 年水文资料选取 5 场洪水进行率定，结果表明各站模拟与实测过程趋势线基本一致，其中峰值水位误差：下路站为 1.31%，西江闸下站为 2.23%，永宁江闸上站为 -0.98%，南洋站为 0.04%。 4 站的洪峰水位平均误差为 0.58%，洪峰滞时平均误差为 0.8 h，主要预报站点的确定性系数在 0.7 以上，计算时间在 30 s以内。将预报模型、工程调度信息与数字孪生流域系统相耦合，总体展示永宁江流域的工情和防汛情况，实现永宁江流域防汛的智慧决策，为调度指挥提供有力的数值分析保障。

Abstract:

Taking the Yongning River Basin as a pilot，this paper aims to construct a digital twin river basin and build a “2 + N ” business application system with “four pre” functions，so as to comprehensively guarantee the regional flood control safety. Considering the objective reasons such as rugged terrain，complex land use types，and multiple administrative regions in the river basin，a flood prediction model for Yongning River is established to improve the prediction accuracy based on series of the Mike models. Adopted the 2013 “Fitter”，2015 “Surodi” and 2019 “Lekima” floods as calculated sessions，based on five flood rates selected from the hydrological data from 2013 to 2019，the results show that the simulation of each station is basically consistent with the measured process line trend. The errors of peak water level are 1.31% at the Xialu station，2.23% below the Xijiang River Gate，-0.98% on the Yongning River Gate，and 0.04% at the Nanyang station. The average flood peak error of the four stations is 0.58%，and the average error of flood peak delay time is 0.8 hours. The deterministic coefficient of the main forecast stations is above 0.7， and the calculation time is within 30 seconds. Moreover，the forecast model，information of project scheduling and the digital twin basin system are integrated，and the industrial situation and flood control of Yongning River are displayed as a whole，so as to realize the intelligent decision-making of flood control for the Yongning River Basin and provide a powerful numerical analysis guarantee for scheduling and command.