车载激光雷达系统设计与实验研究

发布时间：2018-03-09 19:08

本文选题：车载激光雷达　切入点：扫描系统设计　出处：《南京理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：车载激光雷达是自主驾驶车辆的核心传感器,在世界范围内仍是一项有待进步的新兴技术。在上述背景下,本论文以课题组自行研制的12线测距传感器为基础进行了车载激光雷达的系统设计与扫描实验研究。论文在深入资料调研的基础上,选用米格FHB3910型步进电机作为旋转驱动部件,SRH3899-12S型滑环作为供电和数据传输组件,完成了 12线车载激光雷达原理样机的系统设计,实现了 100米范围内5Hz帧频的360度扫描成像,测距精度达5mm。设计了适用于激光雷达原理样机实验的点云数据处理程序,实现了数据读取、坐标变换、去噪平滑和三角网格重建等功能。在此基础上,利用样机对典型场景进行了扫描成像实验,得到了能反映实际场景的三维点云数据,并进行了分析对比,为后续进行运动状态下三维场景成像奠定了良好基础。此外,论文还提出了一种车载激光雷达扫描系统的升级方案,引入了自适应探测俯仰角调节装置和无线供电及数据传输方案,已申请发明专利。
[Abstract]:Vehicle lidar is the core sensor of autonomous driving vehicle, and is still a new technology in the world. Based on the 12-line ranging sensor developed by our team, the system design and scanning experiment of vehicle lidar are studied in this paper. MIG FHB3910 stepping motor is selected as the rotary driving component, SRH3899-12S sliding ring as power supply and data transmission module, the system design of the principle prototype of 12-line vehicular lidar is completed, and the 360-degree scanning imaging of 5Hz frame rate within 100m range is realized. The precision of ranging is up to 5 mm. the point cloud data processing program suitable for the experiment of the principle prototype of lidar is designed. The functions of data reading, coordinate transformation, noise smoothing and triangular mesh reconstruction are realized. The 3D point cloud data which can reflect the actual scene are obtained by the scanning imaging experiment of the typical scene using the prototype, and analyzed and compared, which lays a good foundation for the subsequent 3D scene imaging under the moving state. This paper also proposes an upgrade scheme of lidar scanning system, which introduces adaptive detecting pitch angle adjusting device, wireless power supply and data transmission scheme, and has applied for invention patent.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN958.98

【参考文献】