基于改进RRT的核退役机器人避障方法

核退役机器人工作过程中,传统快速扩展随机树(Rapidly-exploring Random Tree, RRT)路径规划算法缺乏导向性,路径规划效率低,避障能力弱;为此,提出改进RRT路径规划算法,以提高作业效率和准确率。首先,引入目标偏置函数,并提出自适应步长,使RRT路径规划具有导向性,避免陷入局部最优;其次,采用启发式搜索思想,保留优于其父节点的随机搜索点为新节点;最后,修剪路径中的冗余节点,并采用贝塞尔曲线对路径进行平滑处理。在MATLAB平台上进行仿真,结果表明,改进RRT路径规划算法较传统RRT路径规划算法、RRT-connect路径规划算法效率更高,收敛性更强,可以很好地提高核退役机器人的避障能力。

Obstacle avoidance method for nuclear decommissioning robot based on improved RRT

In the working process of nuclear decommissioning robot,the traditional RRT path planning algorithm is lack of guidance,path planning efficiency is low,and obstacle avoidance ability is weak.Therefore,an improved RRT path planning algorithm was proposed to improve the operation efficiency and accuracy.Firstly,the target bias function was introduced and the adaptive step size was proposed to make the RRT path planning oriented and avoid falling into the local optimum.Secondly,the heuristic search idea was used to retain the random search points which were better than their parent nodes as new nodes.Finally,the redundant nodes in the path were pruned,and Bessel curve was used to smooth the line.The simulation results on the MATLAB platform show that the improved RRT path planning algorithm has higher efficiency and stronger convergence than the traditional RRT path planning algorithm and RRT-connect path planning algorithm,and can improve the obstacle avoidance ability of the nuclear decommissioning robot.