基于视觉辅助的隧道变形监测系统研究

发布时间：2018-11-10 23:08

【摘要】：本论文受973课题(课题编号:2013CB035405)资助,承担"硬岩掘进过程状态感知原理"中的掘进过程隧道变形在线测量的研究任务。随着我国经济的发展,城市交通、水利工程、市政建设对于TBM(Tunnel Boring Machine,隧道掘进机)的需求越来越大。我国已经成为对TBM需求量最大的国家,需求量占全球的50%。TBM的施工安全是TBM掘进过程中主要考虑的因素之一,而隧道的变形是影响施工安全的主要因素。过大的隧道变形会造成隧道塌方、TBM卡机等工程事故。因此隧道变形监测系统在TBM掘进过程中具有十分重要的意义。本文提出了基于视觉辅助的隧道变形监测系统以实现隧道变形的自动、精确和及时监测。结合TBM施工情况和监测目标,提出了莹光漆标识待测点,相机检测目标,云台驱动激光位移传感器瞄准目标的方法,并初步研究了双目视觉提高测量效率以及三维点云的拼接方法。论文首先介绍了 TBM的概况和TBM掘进过程中隧道变形所引发的问题,分类介绍了目前国内外隧道变形监测的方法和存在的问题,从而引出论文的研究背景和意义。其次,提出了人工标志点及实验装置的设计方案。对实验装置建立运动学模型,对相机建立视觉模型。将相机捕获的图像进行去噪、阈值分割、腐蚀膨胀等图像处理方法提取目标中心像素坐标。采用基于图像的视觉伺服控制使激光位移传感器的激光束瞄准目标。其次,提出了双目视觉测量目标点间相对距离、激光位移传感器测量三维坐标,并使用半定规划的方法,提高监测系统的测量效率。分析了各参数对测量精度的影响,选择合适的参数用于实际测量中。初步研究了不同时刻测量得到的三维点云的拼接方法。最后,通过模拟实验验证测量系统的精度和测量效率。
[Abstract]:This paper is supported by 973 subjects (subject number: 2013CB035405), which is responsible for the online measurement of tunneling deformation in "principle of state perception in hard rock tunneling process". With the development of economy in our country, the demand for TBM (Tunnel Boring Machine, tunneling machine in urban traffic, water conservancy engineering and municipal construction is increasing. China has become the country with the largest demand for TBM. The construction safety of 50%.TBM, which accounts for the global demand, is one of the main factors to be considered in the process of TBM tunneling, and the tunnel deformation is the main factor affecting the construction safety. Excessive tunnel deformation will cause tunnel collapse, TBM card, and other engineering accidents. Therefore, the tunnel deformation monitoring system is of great significance in the process of TBM tunneling. In this paper, a visual aided tunnel deformation monitoring system is proposed to realize the automatic, accurate and timely monitoring of tunnel deformation. Combined with the construction situation of TBM and the monitoring target, this paper puts forward the methods of marking the spot to be measured, detecting the target by camera and aiming at the target by laser displacement sensor driven by cloud head. And the binocular vision to improve the measurement efficiency and three-dimensional point cloud splicing method is studied preliminarily. Firstly, this paper introduces the general situation of TBM and the problems caused by tunnel deformation in the process of TBM tunneling, and introduces the methods and existing problems of tunnel deformation monitoring at home and abroad, which leads to the research background and significance of the paper. Secondly, the design scheme of manual mark point and experimental device is put forward. The kinematics model of the experimental device and the visual model of the camera are established. Image processing methods such as denoising, threshold segmentation and corrosion expansion are used to extract the central pixel coordinates of the target. The laser beam of the laser displacement sensor is aimed at the target by using the visual servo control based on the image. Secondly, a binocular vision method is proposed to measure the relative distance between the target points and the laser displacement sensor to measure the three-dimensional coordinates, and the semi-definite programming method is used to improve the measuring efficiency of the monitoring system. The influence of the parameters on the measurement accuracy is analyzed, and the appropriate parameters are selected for the practical measurement. The splicing method of three-dimensional point cloud measured at different times is preliminarily studied. Finally, the accuracy and efficiency of the measurement system are verified by simulation experiments.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U456.3;TP274

【参考文献】