针对液压重载机械臂的动态倾覆稳定性问题,提出了一种基于改进快速扩展随机树(Rapidly-exploring Random Tree,RRT)算法的路径规划方法。与只对危险工况的静态稳定性校核不同,该算法以机械臂运动过程中的动态倾覆稳定性最优为目标,在机...针对液压重载机械臂的动态倾覆稳定性问题,提出了一种基于改进快速扩展随机树(Rapidly-exploring Random Tree,RRT)算法的路径规划方法。与只对危险工况的静态稳定性校核不同,该算法以机械臂运动过程中的动态倾覆稳定性最优为目标,在机械臂的关节空间内进行路径规划。以7个关节变量组成的七维数组作为采样点,结合正运动学与力矩法建立机械臂的动态倾覆稳定性计算模型,利用双采样点择优原则,选择其在对应位姿下抗倾覆稳定力矩最优的随机点作为采样点,以增强算法的启发性。在Matlab平台进行的仿真实验表明,改进RRT算法规划路径的倾覆裕度在3种典型工况下分别提升了37%、28%和38%,有效地改善了液压重载机械臂作业平台的抗倾覆稳定性。展开更多
针对人工势场(Artificial Potential Field,APF)法对机械手进行路径规划时存在的问题,提出了关节空间APF自适应变步长和目标偏置的快速扩展随机树(Rapidly-exploring Random Tree,RRT)相结合的方法。在关节空间中进行APF法路径规划,减...针对人工势场(Artificial Potential Field,APF)法对机械手进行路径规划时存在的问题,提出了关节空间APF自适应变步长和目标偏置的快速扩展随机树(Rapidly-exploring Random Tree,RRT)相结合的方法。在关节空间中进行APF法路径规划,减少逆向运动学次数和关节角突变;通过改进斥力和引力势场函数,解决APF法中碰撞和目标不可达问题;采用柯西概率分布,根据末端点与障碍物的距离来改变关节角步长;通过调节RRT算法的目标偏置值,产生合适临时目标点,从而解决APF法局部极小值问题。在APF法存在局部极小值情况下进行机械臂避障仿真,结果表明,自适应变步长路径规划能够生成平滑轨迹并能提高搜索效率,目标偏置RRT选取临时目标点后整体路径长度变短。捡拾机械手在该改进算法下能够有效实现避障拾取任务需求。展开更多
The Chang＇e-3（CE-3） spacecraft successfully landed on one of the youngest mare surfaces on the Moon in December 2013. The Yutu rover carried by CE-3 was equipped with a radar system that could reveal subsurface str...The Chang＇e-3（CE-3） spacecraft successfully landed on one of the youngest mare surfaces on the Moon in December 2013. The Yutu rover carried by CE-3 was equipped with a radar system that could reveal subsurface structures in unprecedented details, which would facilitate understanding regional and global evolutionary history of the Moon. Based on regional geology, cratering scaling, and morphological study, here we quantify the subsurface structures of the landing site using high-resolution orbital and in-situ imagery data. Three layers of lunar regolith, two layers of basalt units, and one layer of ejecta deposits are recognized at the subsurface of the landing site, and their thicknesses are deduced based on the imagery data. These results could serve as essential references for the on-going interpretation of the CE-3 radar data. The ability to validate our theoretical subsurface structure using CE-3 in-situ radar observations will improve the methods for quantifying lunar subsurface structure using crater morphologies and scaling.展开更多
文摘针对液压重载机械臂的动态倾覆稳定性问题,提出了一种基于改进快速扩展随机树(Rapidly-exploring Random Tree,RRT)算法的路径规划方法。与只对危险工况的静态稳定性校核不同,该算法以机械臂运动过程中的动态倾覆稳定性最优为目标,在机械臂的关节空间内进行路径规划。以7个关节变量组成的七维数组作为采样点,结合正运动学与力矩法建立机械臂的动态倾覆稳定性计算模型,利用双采样点择优原则,选择其在对应位姿下抗倾覆稳定力矩最优的随机点作为采样点,以增强算法的启发性。在Matlab平台进行的仿真实验表明,改进RRT算法规划路径的倾覆裕度在3种典型工况下分别提升了37%、28%和38%,有效地改善了液压重载机械臂作业平台的抗倾覆稳定性。
文摘针对人工势场(Artificial Potential Field,APF)法对机械手进行路径规划时存在的问题,提出了关节空间APF自适应变步长和目标偏置的快速扩展随机树(Rapidly-exploring Random Tree,RRT)相结合的方法。在关节空间中进行APF法路径规划,减少逆向运动学次数和关节角突变;通过改进斥力和引力势场函数,解决APF法中碰撞和目标不可达问题;采用柯西概率分布,根据末端点与障碍物的距离来改变关节角步长;通过调节RRT算法的目标偏置值,产生合适临时目标点,从而解决APF法局部极小值问题。在APF法存在局部极小值情况下进行机械臂避障仿真,结果表明,自适应变步长路径规划能够生成平滑轨迹并能提高搜索效率,目标偏置RRT选取临时目标点后整体路径长度变短。捡拾机械手在该改进算法下能够有效实现避障拾取任务需求。
基金supported by the Key Research Program of the Chinese Academy of Sciences (No. KGZD-EW-603)the National Natural Science Foundation of China (Nos. 41373066, 41403053)the State Scholarship Fund of China (No. 201406410040)
文摘The Chang＇e-3（CE-3） spacecraft successfully landed on one of the youngest mare surfaces on the Moon in December 2013. The Yutu rover carried by CE-3 was equipped with a radar system that could reveal subsurface structures in unprecedented details, which would facilitate understanding regional and global evolutionary history of the Moon. Based on regional geology, cratering scaling, and morphological study, here we quantify the subsurface structures of the landing site using high-resolution orbital and in-situ imagery data. Three layers of lunar regolith, two layers of basalt units, and one layer of ejecta deposits are recognized at the subsurface of the landing site, and their thicknesses are deduced based on the imagery data. These results could serve as essential references for the on-going interpretation of the CE-3 radar data. The ability to validate our theoretical subsurface structure using CE-3 in-situ radar observations will improve the methods for quantifying lunar subsurface structure using crater morphologies and scaling.
