影像分析技術辨識丁壩變動結合水理動床模擬分析之研究應用
Other Title
Image Recognition Technology for Groyne Configuration Variation Combined with Hydrodynamic and Mobile-bed Simulation
Journal
台灣水利
Journal Volume
66
Journal Issue
4
Start Page
18
End Page
32
ISSN
0492-1550
Date Issued
2018-12
Author(s)
Abstract
臺灣河川因天然災害（颱洪事件、地震等）及人為因素（採砂頻繁、堰壩建置等），造成河床有明顯沖刷且多處岩盤裸露，加之近年來極端氣候引發降雨強度增加，強降雨所引起之洪水可能增加沖刷強度及造成基腳破壞，進而影響堤內民眾財產生命安全。丁壩為常見保護河岸、高灘地及導引流向之工程，若有規劃良好之丁壩建置則可保護水工構造物安全、防止岸壁淘刷及穩定河道等功能。本研究選擇名竹大橋上游右岸丁壩群為研究區域，利用無人飛行載具（UAV）所拍攝之影像，應用快速影像辨識丁壩變動技術進行丁壩於颱風豪雨前後變動辨識，並採用大尺度粒子影像分析法（LSPIV）進行丁壩周圍河道之表面流場分析，配合二維數值模式模擬成果，比較丁壩區域地形變動及流場量測之差異性。本研究所開發之快速影像辨識丁壩變動技術，經試驗室及現地測試結果可知能有效率的分析出丁壩變動變化。此外，本研究於試驗室測試大尺度粒子影像分析法結果，顯示其能分析出正確表面流場及計算出正確表面流速；且在現地利用無人飛行載具所拍攝影像，進行丁壩局部區域表面流場流速分析亦可得到可靠成果；搭配二維數值模擬之大範圍成果，可完整描述丁壩區域與其上下游水理相關性。
Due to natural disasters (typhoon floods, earthquakes, etc.) and human impacts (such as sand mining and weir/reservoir construction), riverbed erosion and bedrock exposure happened frequently in Taiwan Rivers. In addition, increase in rainfall intensity due to extreme weather may increase the erosion intensity of riverbed and cause the damage of hydraulic structures. The safety of public property inside the embankment could therefore be affected if the embankment or bridge are damaged by floods. Groyne is a general protection work of embankment, floodplain and trajectory bucket. The well designed groyne can have several capabilities such as river channel stabilization and river bank protection. The groyne group at the upstream of the Mingzhu Bridge in the Zhuoshui River is adopted as study area. Unmanned Aerial Vehicle (UAV) is adopted to take aero-photos and videos. Supplemented by artificial recognition in the laboratory and field, the rapid image recognition technique of groynes is developed by this study. It performs capability and accuracy to effectively analyze groyne changing before and after heavy rainfall event. In addition, Large Scale Particle Image Velocimetry (LSPIV) is applied to analyze the flow field around the groynes and main river channel. Based on the laboratory calibration and field measurement comparison by different instruments, the flow velocity of LSPIV analysis can be calculated correctly. A 2-D numerical model is applied to simulate flow field and sediment transportation in the study area. The simulated results of flow velocity and river bed changing are compared to the measurement data and correspond to field investigation performance. Therefore, the flow field and flow velocity around the groynes and main river channel can be reliably measured by using the images taken by UAV and the adapted 2-D numerical model provided comprehensive hydraulic phenomenon for the study integrity.
Subjects
丁壩
無人飛行載具
影像辨識
二維數值模式
流場及流速
Groyne
Unmanned Aerial Vehicle
Image Recognition Technique
2-D numerical model
Flow field and flow velocity
Type
journal article