苹果采摘机器人末端执行器恒力柔顺机构研制

    Development of compliant constant-force mechanism for end effector of apple picking robot

    • 摘要: 为了减少采摘机器人末端执行器在夹持过程中对果实造成的损伤,该文通过在末端执行器上设置柔顺机构,并对柔顺机构力学性能进行计算,求解果实无损采摘所需的柔顺恒力特性。首先,基于形状函数建立边界条件约束下的柔顺梁非线性常微分控制方程;然后,利用打靶法将上述边值问题重新描述为初值问题,并结合遗传算法进行初值优化求解,采用序列二次规划法优化梁的形状函数,使其在一定变形范围内实现恒力输出;最后,在给出求解所需参数和柔顺机构初始形状参数基础上,以苹果采摘为例,通过优化计算,使柔顺梁对果实的夹持力维持在7.9 N左右,非线性有限元计算和力-位移特性试验验证了计算结果的准确性,多次苹果夹持试验的抓取完好率为95%,验证了该柔顺机构无损夹持苹果的可行性。研究结果可为不同类型果实的恒力夹持提供参考。

       

      Abstract: Abstract: In order to reduce the damage caused by end effector of picking robot during fruit clamping process, non-destructive fruit picking has gradually become a key technology of fruit harvesters. The application of compliant constant-force mechanism may enable a constant output force within a certain range of operating displacement. The existing compliant constant force mechanism has the defect that the constant force interval is small relative to the size of the mechanism, which makes it difficult to apply in the situations requiring compact structure design such as fruit picking. In this paper, a method to generate compliant constant-force mechanism is introduced. Firstly, based on the shape function, a set of nonlinear ordinary differential equations for buckling deformation of compliant beams under boundary conditions is established. Secondly, the boundary value problem is redescribed as the initial value problem by applying the shooting method combined with the Runge-kutta method, the 3D nonlinear equations are obtained according to three boundary conditions, two unknown variables and one unknown initial value. Thirdly, the nonlinear equations is solved as an optimization problem. To avoid the optimization problem falling into local convergence, the genetic algorithm is employed to globally find the appropriate initial values of the three unknowns in the nonlinear equations, the solution of the differential equations is obtained according to the principle of the shooting method. Finally, the sequential quadratic programming method is implemented to optimize the parameters of the shape function, the optimization objective is to achieve a constant output force within a certain range of displacement input, and the constraints are the design boundary and the maximum permissible stress of the beams. A two-finger apple picking end effector is taken as an example, the initial shape parameters of the compliant mechanism are set and the algorithm is realized in Matlab. After optimization, the constant clamping force of the compliant beam to the fruit is about 7.9 N. In order to verify the correctness of the method, the non-linear finite element simulation is carried out in Abaqus, and the force-displacement experiment is carried out. The simulation results agree well with calculations, compared to the calculations, the average error of force-displacement test is 5.06%, which is acceptable. The main reason for the error may be that the theoretical model only considers the deformation of the neutral axis. The apple picking test is designed to verify the practicality of the method. At constant-force range, the clamping force fluctuates from 7.46-8.42 N, and the average clamping force is about 8.03 N. The clamping force is basically consistent with the calculation results though a certain fluctuation exists. To further verify the flexibility of the clamping mechanism, 20 apples of similar size and intact surface are selected for clamping test. The result shows that the ratio of non-destructive grasping is 95%, no slipping occurred, the compliant mechanism designed can achieve stable clamping of apples. The algorithm can meet the requirements of constant force clamping for different types fruit by properly adjusting the parameters of the compliant beam. The research provides a reference for the effective application of the compliant mechanism in non-destructive fruit-picking.

       

    /

    返回文章
    返回