Measurements of flow velocity by Using UAV-LSPIV in different flow conditions

• Main Authors: Yu-Jen Chuang, 莊昱仁
• Other Authors: 詹勳全
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/cgi-bin/gs32/gsweb.cgi/login?o=dnclcdr&s=id=%22107NCHU5080046%22.&searchmode=basic

碩士 === 國立中興大學 === 水土保持學系所 === 107 === Previously, LSPIV analysis was affected by poor mobility of the vehicle, which made it limited in image capture. In recent years, with the advancement of technology, the development of unmanned aerial vehicles (UAV) as a photographic equipment for field survey has been developed. UAV has the characteristics of small volume and high mobility, which can penetrate deep into the areas or river sections that are difficult to arrive. Therefore, UAV are be chosen as the photographic vehicle for LSPIV flow velocity analysis in this study. In LSPIV analysis, imagery of tracking particle is an important factor affecting the accuracy of its analysis. Due to the changes of UAV’s flight height or shooting angle, the resolution and deformation of particles in the image will be different. Therefore, this study simulates the change of different shooting conditions through UAV, analyzes the flow rate with LSPIV under different flow conditions, and compares it with the ADCP measurement results. In addition, this study introduces the concept of outliers in calculations to determine the appropriate IA size for different flow conditions and to improve the defect of previous judgments on IA size. In the comparison of LSPIV flow velocity analysis, the difference between the using artificial particles or not is 30%, which is the biggest difference among all the shooting factors. The increase in shooting height and angle will result in a decrease in image resolution and an increase in the amount of deformation, resulting in an increase of 2% to 10% in the analysis result compared with the measured value. Through the interpretation of IA and the comparison between various factors, it is hoped that the analysis efficiency of relevant LSPIV outdoor experiments will be improved.