基于自适应RRT与积分滑模的机器人避障跟踪控制

Robot Obstacle Avoidance Tracking Control Based on Adaptive RRT and Integral Sliding Mode

针对机器人在复杂场景下躲避障碍物以及跟踪运动中的目标等问题,提出一种自适应快速扩展随机树(RRT)与积分滑模的机器人避障跟踪控制。首先,在传统的RRT算法中加入自适应采样概率与自适应步长等策略,用于减少采样点,提高路径搜索速度,并通过五次多项式插值将得到的无碰撞路径点拟合成一条光滑的曲线,减少机器人运行中的抖动;其次,采用积分滑模算法设计了基于图像的视觉伺服控制器,保证系统具有快速跟踪性能与良好的鲁棒性;最后,通过仿真和实验证实,自适应RRT算法具有采样节点少,收敛速度快等优势,所提出的视觉伺服控制器也能够在不同的目标移动速度下对其实现精确跟踪。

Aiming at the problems of robot avoiding obstacles and tracking moving targets in complex scenes,a robot tracking control based on adaptive rapidly exploring random tree(RRT)and integral sliding mode is proposed.Firstly,adaptive sampling probability and adaptive step size strategies are added to the traditional RRT algorithm to reduce sampling points and improve the speed of path search.The obtained collision-free path points are fitted into a smooth curve through quintile polynomial interpolation to reduce the jitter problem of robot operation.Secondly,an image-based visual servo controller is designed by using integral sliding mode algorithm to ensure the fast tracking performance and good robustness of the system.Finally,the simulation and experimental results show that the improved RRT algorithm has the advantages of less sampling nodes and faster convergence speed.The proposed visual servo controller is also able to achieve accurate tracking of the target at different moving speeds.