在自主车辆路径规划中逆向D~*算法的应用  

Applied Research of Converse D-star Algorithm in Freedom Vehicle Path Planning

作　　者：雷雨能[1] 赖文娟[1] 曾刊[1] 

机构地区：[1]中国兵器工业第五八研究所军品部,四川绵阳621000

出　　处：《四川兵工学报》2013年第3期96-98,109,共4页Journal of Sichuan Ordnance

摘　　要：针对自主车辆路径规划在大规模环境中存在执行效率低的问题,本文进行了逆向D*路径规划算法的分析和验证。对自主车辆感知的环境信息建立局部栅格地图,在局部栅格地图中建立机器人距离,车辆在行驶过程中采用A*启发方法搜索中间目标节点,以滚动方式规划出中间目标节点到达当前位置的局部路径。自主车辆沿着局部路径行驶到这个中间目标节点后,再搜索下1个中间目标节点,滚动进行规划。依此循环,直到车辆达到最终目标节点。Matlab仿真实验表明,该方法具有可行性、有效性和实时性。Facing to the problem of low efficiency of autonomous vehicle path planning in big scale environment, an converse D-star algorithm is analyzed and validated. Establishing the part grid map for per- ceptive environment information, in which robot-distance is established, and then using A-star algorithm to search for middle goal node in process of steering, in order to plan the part path from middle goal node to current position node in roll mode. Steering to the middle goal node, autonomous vehicle searches for the next middle goal node and go on to plan in roll mode. The cycle goes on like this until the vehicle steers to the goal node. Experimental results using matlab tool indicate that the method is feasible, effective and real-time.

关 键 词：自主车辆 路径规划 逆向 D＊算法 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]