西瓜采摘末端执行器夹持力精确控制

为实现西瓜的机械化采摘,该文提出了采用液压驱动型末端执行器采摘大型果实的思路。在抓取西瓜时,为避免末端执行器夹持力不稳定引起的果实损伤,需对夹持力进行精确控制,该文建立了末端执行器负载模拟平台和AMESimSimulink联合仿真模型,模拟西瓜采摘夹持力的加载情况。针对执行器夹持力加载过程中位置控制系统对力控制系统产生干扰,影响夹持力精确加载的问题,该文基于速度同步控制原理,设计了简化的加载误差补偿环节,开展了加载误差补偿理论、仿真及试验研究。结果表明,速度同步控制方法能够有效地减小加载误差,提高末端执行器负载模拟精度。该研究可为末端执行器输出力的精确控制和抓取控制策略提供参考。

Precise control of clamping force for watermelon picking end-effector

Abstract: Certain agricultural work requires a very stable robot end-effector to reduce vibration and damage to products. However, the structures of conventional articulated robots tend to be too weak to manipulate heavy objects, such as the watermelon. To achieve mechanized harvesting of the watermelon, we compared the characteristics of different end-effectors with the motor, pneumatic and hydraulic drives. By contrast, the hydraulic drive has many advantages, such as high power-weight ratio, compact structure and convenient layout. Therefore, the end-effector with hydraulic drive is more suitable for heavy fruit harvesting such as the watermelon. The hydraulic drive system based on force perception can output appropriate clamping force according to actual situation. When picking robot crawls and picks, it needs to avoid the damage to fruits and vegetables caused by unstable clamping force and to achieve the precise control of the clamping force for picking robot end-effector. Therefore, how to improve output force's control precision of the valve control cylinder and reduce interference are key to improve the picking performance. Two identical small valve control cylinders were used to establish a load simulation test platform for simulating the loading condition of the clamping force in picking watermelons. There existed the question that position control system interfered the force system on load simulation test platform, which situation would affect the loading accuracy of the end-effector and result in the loading error; and hence in order to reduce the loading error, the control compensation method, which could produce a compensation effect by increasing the compensation algorithm link in the control program, was adopted in this paper. Through this method the loading error was effectively suppressed. Based on the load simulation test platform, the researches of the theory, simulation and experiment on the inhibition of the loading error were carried out. The mathematical relationship between position disturbance and output force was built with mechanism modeling method. Based on the speed synchronization control method, the simplified compensation link of loading error was designed and the co-simulation model of AMESim & Simulink and physical experiment platform were built; in addition, for picking watermelon, the mechanical properties and the situation of actual picking watermelons were taken into consideration in this paper. Researching the compression characteristics of watermelon, it was known that before watermelon reached yield limit, microstructure of watermelon was without damage, so the compression force at this point was considered to be safe. Taking into account the yield limit of loading compression force, the safe loading force in simulation and experiment was set up. In the watermelon's picking process, the position of the disturbance will inevitably produce. Different crawling and moving speeds will produce in various degrees of disturbance. At the same time, picking different qualities of watermelons needs different clamping forces. The accuracies of loading clamping forces at different speeds and different loads during the picking process were studied. By simulation and experiment methods, the loading error can be restrained effectively and the load simulation accuracy of end-effector can be improved greatly, which contributes to the synchronous speed control method. This method provides a powerful guide for the design of end-effector and precise control and real-time tracking of the output force. In this paper, watermelon picking is taken as the example to study the feasibility of control theory, but the study is not limited to the watermelon picking. The results in this paper are generally applicable to the accuracy control of the end-effector driven by hydraulic pressure.