基于改进APF-Informed-RRT^(*)的机械臂避障路径规划

Obstacle Avoidance Path Planning of Manipulator Based on Improved APF-Informed-RRT^(*)

针对Informed-RRT^(*)算法在避障路径规划中缺乏目的性和方向性,存在规划时间长、迭代效率低等问题,提出了结合人工势场法和Informed-RRT^(*)算法的避障规划算法。首先,针对传统人工势场法存在目标点不可达、易与障碍物碰撞的问题,提出了改进后的人工势场法,并将其融入Informed-RRT^(*)算法中,使随机树沿势场下降的方向生长,增强其方向性;其次,依据随机树与障碍物间的距离,提出了一种自适应生长步长策略,提高了对空间的探索能力;最后,引入贪心算法的思想,在生长时直接判断随机树能否直达目标点,提高了路径规划效率。在二维和三维环境下对改进后的算法与传统算法及其衍生算法进行对比实验,仿真结果表明改进后的Informed-RRT^(*)算法相较于原始算法规划的路径长度和规划耗时分别减少了17.42%和36.21%。

Aiming at the problem of informed-RRT^(*)algorithm lacks purpose and direction,has long planning time and low iteration efficiency in the obstacle avoidance path planning.An obstacle avoidance planning algorithm combining artificial potential field method and Informed-RRT^(*)algorithm was proposed.Firstly,aiming at the problems of unreachable target points and collision with obstacles in original artificial potential field method,an improved artificial potential field method is proposed and integrated into the Informed-RRT^(*)algorithm to make the random tree grow along the direction of potential field decline and enhance its directivity.Secondly,according to the distance between the random tree and the obstacle,propose an adaptive growth step strategy to improve the ability of space exploration.Finally,the greedy algorithm is introduced to directly judge whether the random tree can reach the target point during growth,which improves the efficiency of path planning.The improved algorithm is compared with the traditional algorithm and its derived algorithm in two and three dimensional environments.Simulation results show that compared with original algorithm,the improved Informed-RRT^(*)algorithm reduces the path length and plan time by 17.42%and 36.21%.