基于三维线激光技术的路面车辙检测技术研究

发布时间：2019-01-22 10:33

【摘要】：我国已从道路建设的高峰期快速进入以养护和维修为主的管养期,道路管理部门将长期面临艰巨的养护管理任务。全面准确、快速可靠地掌握路面病害数据是科学养护管理决策的关键。车辙作为沥青路面主要病害形式之一,不仅直接影响行车安全性和舒适性,还严重危害道路结构整体性和稳定性。因此获取准确、可靠的车辙检测数据是路面状况评价与养护决策的重要任务。目前国内外广泛采用的路面车辙人工直尺测量方法因耗时、费力且危险性高已逐渐被多点激光检测系统替代。然而,多点激光技术受激光点数量制约和检测车辆横向偏移的影响,常产生车辙深度检测结果的低估,进而导致车辙严重程度的误判。同时,多点激光检测系统难以获取车辙三维细观形貌,难以准确计算车辙多维度特征,分析现有的车辙一维深度检测数据无法为管理者提供全面、真实、准确的车辙信息,制约了公路养护管理的科学化、精细化发展。本文首先通过收集大量相关资料和文献,基于先进的商用三维线激光设备,分析了三维线激光检测原理与特性,进一步地,通过室内实验分析了检测系统的扫描线间隔与架设高度等参数对三维成像特性的影响;然后,分别以大理石、水泥板、沥青板和标定板为检测对象,研究不同特性的材料对激光点云数据的准确性与重复性影响;针对黑、黄、白三种路面常见颜色和不同路面构造深度,研究了被测物体表面色彩和构造深度对激光点云数据特性的影响;最后,分析了检测过程中线激光数据异常值特性,通过室内外实验研究了缺失点与突变点的可能成因;借助MATLAB平台编制了路面车辙多维度特征参数提取软件,实现了车辙多维度特征参数提取,并通过室内试验验证了车辙多维度指标提取算法的准确性。通过对新兴三维线激光点云数据特性分析,并对其在路面车辙多维度指标检测进行研究,是对三维线激光技术在路面管理领域适用性的初步探究,是对我国现有车辙分析技术的完善,对推进公路养护管理的科学化、精细化建设有益补充。
[Abstract]:Our country has entered the maintenance and maintenance period from the peak period of road construction, and the road management department will face the arduous task of maintenance and management for a long time. It is the key of scientific maintenance management to grasp the pavement disease data accurately and quickly. Rutting, as one of the main diseases of asphalt pavement, not only directly affects the driving safety and comfort, but also seriously endangers the integrity and stability of road structure. Therefore, obtaining accurate and reliable rut detection data is an important task in pavement condition evaluation and maintenance decision. At present, the manual measuring method of pavement rutting has been gradually replaced by multipoint laser detection system because of its time-consuming, laborious and dangerous measurement method, which is widely used at home and abroad. However, the multi-point laser technology is influenced by the number of laser points and the lateral migration of vehicles, which often leads to the underestimation of the detection results of rutting depth, which leads to the misjudgment of the severity of rutting. At the same time, the multi-point laser detection system is difficult to obtain 3D micro-morphology of ruts and calculate the multi-dimensional features of ruts accurately. Analyzing the existing data of one-dimensional rutting depth detection can not provide the administrator with comprehensive, true and accurate rutting information. It restricts the scientific development of highway maintenance management. Based on the advanced commercial 3D line laser equipment, the principle and characteristics of 3D line laser detection are analyzed in this paper. The effects of scanning line spacing and erection height on the 3D imaging characteristics of the detection system are analyzed through laboratory experiments. Then, the accuracy and repeatability of laser point cloud data were studied by using marble, cement slab, asphalt plate and calibration board respectively. Aiming at the common colors of black, yellow and white pavement and different road structure depth, the influence of surface color and structure depth on the characteristics of laser point cloud data is studied. Finally, the characteristics of abnormal values of line laser data in the detection process are analyzed, and the possible causes of missing points and abrupt points are studied through indoor and outdoor experiments. With the help of MATLAB platform, the software for extracting multi-dimensional rutting parameters is developed, and the accuracy of the algorithm is verified by indoor experiments. By analyzing the characteristics of the new 3D line laser point cloud data, and studying the multi-dimensional index detection of rutting on the road surface, it is a preliminary exploration of the applicability of 3D line laser technology in the field of road surface management. It is a useful supplement to the improvement of rutting analysis technology in China, to the scientific management of highway maintenance and to the fine construction of highway maintenance management.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：U418.6

【相似文献】