摘要
提出了一种基于PRM(probabilistic roadmaps)算法思想的新型工业机器人运动规划算法.将PRM规划算法由全部C空间计算优化为大部分在欧氏空间、小部分在C空间的计算模式,大幅减少了计算量;使用优化空间分割方法提高采样效率;采用凸优化方法解决了机器人静止姿态碰撞检测问题,并结合自适应动态碰撞检测算法,实现在不降低计算精度的前提下,提高运算速度,使规划结果具有完备性.
A PRM （probabilistic roadmaps） based novel motion planning algorithm was proposed for industrial robots. It could significantly reduce computational efforts by carrying out robot path calculation in both Cartesian space and configuration space. This algorithm was embedded with a fast space partition method to improve sampling efficiency. Since both the robot and obstacles were modeled as cuboids, the static collision detection was formulated as a convex optimization problem. By applying the adaptive dynamic collision detection method, the motion planning problem was solved efficiently and completely without loss of computation precision.
基金
国家自然科学基金(51075085)
国家高技术研究发展专项经费(2011AA04A103)资助
关键词
工业机器人
运动规划
空间分割
碰撞检测
industrial robots
motion planning
space partition
collision detection
