Abstract: In order to achieve the high-speed smooth movement to play the badminton continuously for the manipulator of badminton robot, this paper proposes a new multi-objective trajectory optimization model of manipulator. To begin with, the optimization establishes the pose expression of manipulator through successive coordinate transformation, according to D-H model. After that, it uses the Newton descent method to solve the disaggregation of inverse kinematics for the given trajectory key points, then solves the optimal solution from the disaggregation of inverse kinematics by the shortest path algorithm. Finally, to achieve the continuous smooth trajectory of manipulator, it uses the cubic spline interpolation to establish the function of motors’ rotating angle, evolving in time, in accordance with the optimal solution. The results of experiments show that the new trajectory optimization model can effectively reduce?energy consumption of motor, increase the turning efficiency, and ensure the response?rate of manipulator sequentially.

Key words: trajectory planning, D-H model, Newton descent method, shortest path algorithm, cubic spline interpolation

摘要： 为了实现羽毛球机器人机械臂高速连续平滑地击打羽毛球动作，提出了一种新的多目标机械臂运动轨迹优化模型。首先，该轨迹优化模型根据D-H运动学模型，通过坐标变换建立机械臂的位姿表达式。然后，采用牛顿下山法求出给定路径关键点的运动学逆解集，并基于最短路径算法从逆解集中求出最优解。最后，根据所求出最优解，采用三次样条插值建立电机转角函数，以实现机械臂的连续平滑运动。实验结果表明：新的轨迹优化模型能够有效地降低电机能耗和提高转动效率，从而保证了机械臂响应速度。

关键词: 轨迹规划, D-H模型, 牛顿下山法, 最短路径算法, 三次样条插值