面向自动驾驶碾压作业车的障碍物检测研究

Obstacle Detection for Automatic Driving Roller Compacting Vehicle

为解决自动驾驶碾压作业车在非结构化环境下的障碍物检测问题,研究了一种基于D-S证据理论的障碍物检测方法。对某型号碾压作业车,设计对应的三维激光雷达支架,安装及标定激光雷达,构建碾压作业车障碍物检测平台。针对碾压作业面非结构化道路环境,分析作业面中障碍物的实际特点,采用D-S证据理论融合相对高度差、单线径向距离跳变、局部区域双梯度阈值等方法,实现对正障碍物、负障碍物和坡道的检测,获得全面的障碍物栅格地图。试验结果表明,该融合检测方法有效实现碾压作业面的障碍物检测,满足实际工程应用需求。

To solve the problem of obstacle detection in the unstructured environment of automatic driving rolling vehicles, a method of obstacle detection based on D-S evidence theory was studied.For a certain type of rolling compacting vehicle, the installation method of a 3D LiDAR bracket is designed, the 3D LiDAR is installed and calibrated. The obstacle detection platform of a rolling compacting vehicle is constructed.Aiming at the unstructured road environment of the rolling surface, the actual characteristics of obstacles in the working face are analyzed. D-S evidence theory is used to fuse the relative height difference, single-line radial distance jump, local area double gradient threshold, and other methods to realize the detection of positive obstacles, negative obstacles and ramps, to obtain a comprehensive obstacle grid map. The experimental results show that the fusion detection method can effectively detect the obstacles in the rolling operation surface and meets the actual needs of engineering applications.