Abstract: In order to resolve the problem that the harvesting robot vision system is not able to acurately provide the growth posture of citrus fruit under complex natural environment, thus the successful rate of harvesting is reduced, this paper proposed an optimum harvesting posture determination method based on citrus harvesting robot occlusal end-effector. This method based on end-effector configuration parameters, established the performance evaluation function of the harvesting posture to the center position of fruit, and gained the optimal harvesting posture of actuator by using calculation of this function. The calculated optimal harvesting posture was verified by constructing harvest experiment platform and designing harvest experiment. The results showed that citrus harvesting successful rate was increased by 26.32% using the harvesting posture optimized by the optimal harvesting posture evaluation method, compared with the general horizontal harvesting posture.

Key words: citrus harvesting robot, end-effector, configuration, harvesting posture

摘要： 针对采摘机器人视觉系统在复杂自然环境中无法准确提供柑橘果实生长姿态，进而导致采摘成功率下降的问题,基于柑橘采摘机器人咬合型末端执行器提出了一种最佳采摘姿态确定方法。该方法依据末端执行器构型参数，建立其采摘姿态对果实中心位置影响的性能评价函数，并使用该函数计算得到执行器最佳采摘姿态。通过搭建采摘实验平台和设计采摘实验，对计算出的最佳采摘姿态进行验证。实验结果表明，与一般的水平采摘姿态相比，采用最佳采摘姿态评价方法优化后的采摘姿态，在进行柑橘采摘时采摘成功率提高26.32%。

关键词: 柑橘采摘机器人, 末端执行器, 构型, 采摘姿态