高级检索

基于改进A*与DWA算法的井下搬运机器人自主行走路径规划

Autonomous walking path planning of underground handling robot based on improved A* and DWA algorithm

  • 摘要: 针对井下搬运过程用工多且效率低等问题以及转运环节多、转运操作难、人工转运慢等困境,以搬运机器人为研究对象,通过仿真与试验进行井下巷道路径规划方法研究。首先,对井下转运工况进行特定的需求分析,确定井下“多转运点巷道”与“最后1 km直达运输”等典型搬运场景,并分别建立自主行走策略,提升井下搬运机器人自主行走路径规划的适用性与合理性;其次,设计路径规划系统方案,对比分析常规路径规划算法的性能,分别选取A*算法与DWA算法作为井下搬运机器人的全局路径规划与局部路径规划的基本算法;然后,针对常规A*算法搜索效率低与规划路径平滑性不足的问题,依次使用指数优化法与二次B样条曲线法进行改进,同时对比分析改进A*算法的性能,仿真结果发现:改进A*算法在“最后1 km直达运输”与“多转运点自主行走”等典型井下搬运场景的路径规划时间分别缩短了24.62%与22.02%,路径规划效率获得了提升,且生成路径的平滑性较好;接着,针对传统DWA算法存在运行时间与路径长度欠优的问题,统筹井下巷道速度限制、移动底盘驱动电机限制与搬运机器人制动距离限制等约束性要素构建预测路径评价函数,并衡量安全性与高效性引入障碍物评价子函数改进DWA算法,仿真结果发现:基于改进DWA算法的规划路径长度与自主行走时间分别缩短了31.27%与42.33%,有效提升了搬运机器人动态避障能力及路径规划性能;最后,搭建试验巷道场景与搬运机器人模型开展井下搬运机器人路径规划试验,试验结果表明:改进A*-DWA融合算法在多转运点巷道自主行走A→B、A→C以及最后1 km直达运输等试验场景的路径规划效率分别提升了21.90%、18.57%、14.67%,行驶过程实现实时有效动态避障,规划路径安全高效且具有平滑性,能够符合井下“减人、提效、增安”的目标导向以及满足搬运机器人自主行走路径规划需求。

     

    Abstract: In view of the problems of many labors and low efficiency in the underground handling process, as well as the difficulties of many transport links, difficult transport operation and slow manual transport, the handling robot is taken as the research object, and the path planning method of underground roadway is studied through simulation and experiment. Firstly, the specific demand analysis of the underground transport conditions is carried out to determine the typical transportation scenarios such as “multi-transport point roadway” and “last 1km direct transport”, and the autonomous walking strategies are established respectively to improve the applicability and rationality of the autonomous walking path planning of the underground handling robot. Secondly, the path planning system scheme is designed, and the performance of the conventional path planning algorithms is compared and analyzed. The A* algorithm and the DWA algorithm are selected as the basic algorithms for global path planning and local path planning of the underground handling robot. Then, in order to solve the problem of low search efficiency and insufficient smoothness of the planned path of the conventional A* algorithm, the exponential optimization method and the quadratic B-spline curve method are used to improve the algorithm in turn, and the performance of the improved A* algorithm is compared and analyzed. The simulation results show that the improved A* algorithm shortens the path planning time of typical underground handling scenarios such as “the last 1 km direct transport” and “multi-transport point autonomous walking” by 24.62% and 22.02% respectively, and the path planning efficiency is improved, and the smoothness of the generated path is better. Besides, aiming at the problem that the traditional DWA algorithm has poor running time and path length, the predictive path evaluation function is constructed by coordinating the constraints of underground roadway speed limit, mobile chassis drive motor limit and handling robot braking distance limit, and the obstacle evaluation sub-function is introduced to improve the DWA algorithm by measuring safety and efficiency. The simulation results show that the planned path length and autonomous walking time based on the improved DWA algorithm are shortened by 31.27% and 42.33% respectively, which effectively improves the dynamic obstacle avoidance ability and path planning performance of the handling robot. Finally, the experimental roadway scenario and the handling robot model are built to carry out the path planning experiment of the underground handling robot. The experimental results show that the path planning efficiency of the improved A*-DWA fusion algorithm in the multi-transport point roadway autonomous walking A→B, A→C, the last 1 km direct transport and other scenarios is improved by 21.90%, 18.57% and 14.67%, respectively. The driving process realizes real-time and effective dynamic obstacle avoidance, and the planned path is safe, efficient and smooth, which can conform to the goal orientation of “reducing people, improving efficiency and increasing safety” and meet the needs of autonomous walking path planning of handling robots.

     

/

返回文章
返回