论文部分内容阅读
基于无急回摆动机构,实现了单线激光雷达对周围区域环境的实时感知与建图。为了准确地检测障碍物,在讨论了摆动机构的安装标定及扫描帧与角度编码器对准方法的基础上,采用了一种生成可通行性代价地图(costmap)的算法。该算法以2-D栅格模型方式,通过综合高程信息、测量值个数、扫描帧个数和测量时间等信息,得到各栅格单元的代价值,生成一张关于环境可通行性代价地图。实车验证结果表明:基于该算法得到的地图能够较好地描述环境的可通行性,实现了对环境的实时感知。
Based on the non-hitch swinging mechanism, the real-time perception and the construction of the surroundings of the single-line laser radar are realized. In order to accurately detect the obstacle, based on the discussion of the installation of the swinging mechanism and the alignment of the scanning frame and the angle encoder, an algorithm to generate costmap is adopted. The algorithm takes the 2-D raster model as a way to obtain the cost value of each grid cell by synthesizing the information of elevation, number of measurements, number of scanning frames and measurement time, and generates a map about the cost of environmental viability . The real car verification results show that the map based on this algorithm can describe the environment’s viability and realize the real-time perception of the environment.