Abstract: The inverse kinematics solution of the banana picking robot with four degrees of freedom should be resolved according to the target position and attitude of the fruit handle clamping actuator at each sampling time. The two degrees of freedom of the base rotation angle and end rotation angle of the banana picking robot, its analytical solution of inverse kinematics is resolved according to the geometric relations. However, the two degrees of freedom of the horizontal displacement and vertical displacement are nonlinear due to their mutual coupling, so it is difficult to resolve their inverse kinematics solutions by analytical solution methods. Therefore, a numerical solution method for inverse kinematics of the banana picking robot based on Newton iteration method is proposed. Firstly, according to the mechanical structure and geometric relations of picking robot, the forward kinematics model solving the position and attitude of the fruit handle clamping actuator by four degrees of freedom is established, then the equations satisfied by the two degrees of freedom of horizontal displacement and vertical displacement are constructed, and the inverse matrix of its Jacobian matrix is obtained. Then the Newton iterative algorithm designed in this paper is used to resolve the inverse kinematics solution of horizontal displacement and vertical displacement according to the target position and attitude of the fruit handle clamping actuator. The experimental results show that the inverse kinematics solution of the base rotation angle and end rotation angle of the picking robot solved by the method proposed in this paper has no error. For the inverse kinematics solution of horizontal displacement and vertical displacement, when the fruit handle clamping actuator takes any target position and attitude, and the initial value of Newton iteration takes a fixed initial position and attitude, the error of inverse kinematics solution with two Newton iterations is less than 01 mm, and has no error with three Newton iterations. When the fruit handle clamping actuator takes the target position and attitude at each sampling time in continuous motion, and the initial value of Newton iteration takes the inverse kinematics solution of previous sampling time, only one iteration is needed to obtain the highprecision inverse solution.

Key words: banana picking robot, Newtons iteration method, solving inverse kinematics, forward kinematics modeling

摘要： 香蕉采摘机器人要在各采样时刻根据果柄夹持机构的目标位置和姿态求得4个自由度的逆运动学解，其底座转角和末端转角可根据几何关系求得解析解，但水平位移和垂直位移因相互耦合呈非线性关系，难以通过解析法求得逆运动学解，故提出一种基于牛顿迭代法的香蕉采摘机器人逆运动学数值求解方法。首先根据采摘机器人的机械结构和几何关系，建立由4个自由度表示果柄夹持机构位置和姿态的正运动学模型，再据此构造水平位移和垂直位移2个自由度满足的方程组，并求出其雅可比矩阵的逆矩阵，然后利用牛顿迭代算法根据果柄夹持机构的目标位置和姿态求得水平位移和垂直位移的逆运动学解。试验结果表明：该方法求得采摘机器人底座转角和末端转角的逆解无误差。而对于水平位移和垂直位移的逆解，当果柄夹持机构取单个任意的目标位置和姿态，且牛顿迭代初值取固定初始位姿时，2次迭代的逆解误差小于01mm，3次迭代的逆解无误差；当果柄夹持机构取连续运动中各采样时刻的目标位置和姿态，且牛顿迭代初值取上一采样时刻的逆解时，只需1次迭代即可得到高精度的逆解。

关键词: 香蕉采摘机器人, 牛顿迭代法, 逆运动学求解, 正运动学建模