面向河道水流实时监测的时空图像测速方法及试验研究

Prototype observation of river flow, especially the measurements of velocity, flow field and discharge under high-flow conditions are the scientific foundations for researches on river hydrology and river dynamics. Large-scale image velocimetry (LSPIV) is a non-intrusive technique to measure whole-field velocities at the free-surface of a water body spanning large flow areas (up to hundreds of square meters). Due to the complex field conditions such as poor visibility, uneven distribution and uncertain trackability of flow tracers, the existing methods still suffer from large uncertainties. This project will explore to build a novel operating mode of Space-time Image Velocimetry for real-time river flow monitoring. (1) Propose a target enhancement method based on Spectrum-Polarization Imaging to realize real-time acquizition of high-SNR images under complex imaging conditions. (2) Propose a motion estimation method based on spectra analysis of Space-Time Image to realize accurate reconstruction of high resolution time-averaged velocity field under complex tracing conditions. (3) Propose a velocity calibration method based on scale transformation of object-image plane to realize fast and safe deployment of a GCP-free measuring system under complex testing conditions. (4) Propose a sensitivity analysis method based on the combination of numerical simulation, model experiment and field test to build a full set of uncertainty evaluation system. The study is potential to provide theoretical and technical supports for developing the new type of integrated optical instruments for flow measurement.

河道水流的原型观测，特别是高洪期流速、流场及流量等量测数据的获取是河流水文学、河流动力学等研究的科学基础。大尺度粒子图像测速是一种用于测量大面积（数百平方米）水体表面流速的非接触式全场流速测量技术。受水流示踪物可见性差、时空分布不均和跟随性不明确等复杂现场条件的影响，现有方法仍存在较大的不确定性。本项目拟探索构建面向河道水流实时监测的时空图像测速工作模式：提出基于光谱-偏振成像的水面目标增强方法，实现复杂成像条件下高信噪比图像序列的实时获取；提出基于时空图像频谱分析的运动矢量估计方法，实现复杂示踪条件下高分辨率时均流场的精确重建；提出基于物像尺度变换的河流水面流速标定方法，实现复杂施测条件下免控制点的测量系统快速安全布设；综合采用数值模拟、模型实验和现场比测等方式开展敏感性分析，建立完善的测量不确定度评估体系。本项研究将为新型集成一体化光学测流仪器的研发提供理论依据和技术方法上的支撑。

河道水流的原型观测，特别是高洪期流速、流场及流量等量测数据的获取是河流水文学、河流动力学等研究的科学基础。大尺度粒子图像测速是一种用于测量大面积（数百平方米）水体表面流速的非接触式全场流速测量技术。受水流示踪物可见性差、时空分布不均和跟随性不明确等复杂现场条件的影响，现有方法仍存在较大的不确定性。本项目探索构建面向河道水流实时监测的时空图像测速法：提出了基于光谱-偏振成像的水面目标增强方法，实现复杂成像条件下高信噪比图像序列的实时获取；提出了基于时空图像频谱分析的运动矢量估计方法，实现了复杂示踪条件下高分辨率时均流场的精确重建；提出了基于物像尺度变换的河流水面流速标定方法，实现复杂施测条件下免控制点的测量系统快速安全布设；综合采用数值模拟、模型实验和现场比测等方式开展了敏感性分析，建立了完善的测量不确定度评估体系。本研究累计申请国家发明专利8项，已授权6项；在国内学术期刊和会议发表论文16篇，其中SCI检索6篇、EI检索4篇，获得会议优秀论文4篇。形成系统实验样机4套，在现场得到应用。研究成果“河-气界面环境下全场流速仿复眼成像测量系统”获得2019年江苏省科学技术奖三等奖。研究成果将为新型集成一体化光学测流仪器的研发提供理论依据和技术方法上的支撑。