基于DBS-RRT^(*)算法的机械臂复杂狭窄场景路径规划  

Complex Narrow Scene Path Planning for Robotic Arm Based on DBS-RRT^(*)Algorithm

作　　者：秦鹏飞 王军茹[1] 张菂 孙广彬 QIN Pengfei;WANG Junru;ZHANG Di;SUN Guangbin(School of Automation,Beijing Information Science and Technology University,Beijing 100192,China;One Hundred Meter Robot Technology(Beijing)Co.,Ltd.,Beijing 102200,China)

机构地区：[1]北京信息科技大学自动化学院,北京100192 [2]壹佰米机器人技术(北京)有限公司,北京102200

出　　处：《组合机床与自动化加工技术》2024年第6期62-69,共8页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家自然科学基金项目(61573230)。

摘　　要：针对目前RRT^(*)算法在机械臂复杂狭窄场景路径规划中,存在着规划时间长、路径冗长、狭窄环境规划成功率低的问题,提出一种动态偏置采样DBS-RRT^(*)(dynamic biased sampling RRT^(*))算法。首先,DBS-RRT^(*)算法采用动态偏置率,设计智能椭球子集采样作为偏向采样方法,利用自适应生长策略调整新节点的生长方向与步长,实现动态选择采样方法,提高采样效率,减少无效空间探索,改善搜索导向性的效果;然后,通过设计二维实验验证算法的有效性,实验证明DBS-RRT^(*)算法与RRT^(*)算法相比,规划效率更高,规划路径更短;最后,将DBS-RRT^(*)算法应用于复杂狭窄场景中的机械臂仿真实验。实验数据表明,DBS-RRT^(*)算法与RRT^(*)算法相比,规划路径长度减少了26%,规划时间减少了22.6%,成功率提高了32%。DBS-RRT^(*)算法在复杂狭窄场景中,相比RRT^(*)算法,能够更加有效地实现机械臂避障路径规划。Aiming at the problems of long planning time,lengthy path and low success rate of narrow environment planning in current RRT^(*)algorithm in complex narrow scene path planning of robot arm,a dynamic biased sampling RRT^(*)(DBS-RRT^(*))algorithm is proposed.The DBS-RRT^(*)algorithm adopts a dynamic bias rate,designs intelligent ellipsoid subset sampling as a biased sampling method,and utilizes an adaptive growth strategy to adjust the growth direction and step size of new nodes to achieve the effect of dynamically selecting the sampling method,improving the sampling efficiency,reducing the ineffective spatial exploration,and improving the search orientation.Then,the effectiveness of the algorithm is verified by designing two-dimensional experiments,which prove that the DBS-RRT^(*)algorithm has higher planning efficiency and shorter planning paths compared with the RRT^(*)algorithm;finally,the DBS-RRT^(*)algorithm is applied to a robotic arm simulation experiment in a complex narrow scene.The experimental data show that the DBS-RRT^(*)algorithm reduces the planning path length by 26%,reduces the planning time by 22.6%,and improves the success rate by 32%compared with the RRT^(*)algorithm.The DBS-RRT^(*)algorithm is able to realize the obstacle-avoidance path planning of the robotic arm more efficiently compared with the RRT^(*)algorithm in the complex and narrow scene.

关 键 词：DBS-RRT^(*)算法 动态偏置率 机械臂 路径规划 复杂狭窄场景 

分 类 号：TH112[机械工程—机械设计及理论] TG659[金属学及工艺—金属切削加工及机床]