具有运动时间约束的机械手最优平滑轨迹规划

Smoothness-optimal trajectory planning method with constraint on traveling time for manipulators

为了提高机械手在指定时间内完成运动任务的平滑性,提出一种最优平滑轨迹规划方法。采用七次B样条曲线插值关节位置序列,保证关节加加速度连续且关节启停速度、加速度和加加速度可设定。将运动学约束转化为B样条曲线控制顶点约束,以加加速度平方积分的最小值作为平滑性能指标,采用遗传算法对平滑性能指标和相应的时间节点向量全局寻优,进而规划出最优平滑轨迹,解决了现有平滑轨迹规划方法没有考虑加加速度累积效果和运动学、运动时间约束的问题。实验结果表明,该方法规划的轨迹具有加加速度平滑和加加速度累积最小的特点,与现有的三次样条轨迹相比,可有效降低轨迹跟踪误差。

In order to improve running smoothness of joints while manipulators execute a traveling task in definite time, a smoothness-optimal trajectory planning method is proposed. Seven order B-spline was adopted to interpolate joint position sequences, so continuous jerk was ensured and start-stop joint velocity, acceleration and jerk can be configured. Constraints on manipulators＇ kinematics were transferred to constraints on control points of B-splines, and genetic algorithm was exploited to globally minimize the integral of squared jerk and seek the according optimal time nodes, then smoothness-optimal trajectories were gencrated, and problems lied in the existing methods including taking no attention to jerk accumulation and constraints on kinematics or traveling time were solved. Experimental results show that, trajectories generated by the proposed planner have the advantages of continuous and minimum accumulative jerk, and can decrease trajectory tracking errors effectively comparing to the existing cubic spline trajectories.