Abstract: Agricultural picking manipulator is an important means to improve the degree of agricultural automation and production efficiency. Aiming at the problems of single control mode, low interaction with users, and low control efficiency of the traditional agricultural picking manipulator, a multifunctional body sensing manipulator system based on bionics is proposed and designed, which can simulate the picking behavior of upper limbs in realtime. The system uses STM32 microprocessor as the main control chip and six steering engines to construct the sixaxis bionic manipulator structure. MUP6050 accelerometer at the acquisition end obtains the posture information of the human arm and realizes the wireless transmission of the arm posture through the wireless transceiver chip SI4432. The receiving end of the robot arm completes the wireless reception and analysis of the arm posture to control the robot arm to imitate the action of the human upper limb. The experimental results show that the system has an average recognition rate of more than 96% and an average response time of 2-3 s. It can make corresponding simulation actions according to the movement track of the human arm, and follow the human arm to complete all kinds of grasping, flexing, extending, rotating, and other actions to better realize the function of the arm.

Key words: STM32, somatosensory interaction, manipulator, arm attitude, acceleration sensor, picking

摘要： 农业采摘机械臂是提高农业自动化程度和生产效率的重要手段,针对传统农业采摘机械臂控制方式单一、与用户的交互性和控制效率较低的问题,提出并设计一种基于仿生学,能够实时模仿上肢采摘行为的体感交互式仿上肢机械臂系统。系统采用STM32微处理器作为主控芯片,采用6个舵机构建六轴仿生机械臂结构。采集端MUP6050加速度计获取人体手臂姿态信息,通过无线收发芯片SI4432实现手臂姿态的无线传输;机械臂接收端完成手臂姿态的无线接收及解析,从而控制机械臂模仿人体上肢的动作。试验结果表明,系统对仿人体机械臂动作的平均识别率为96%以上,动作跟随的平均响应时间为2～3 s,能够按照人体手臂的运动轨迹做出相应的模拟动作,跟随人体手臂完成各种抓握、屈伸、旋转等动作,较好地实现仿上肢采摘机械臂的功能。

关键词: STM32, 体感交互, 机械臂, 手臂姿态, 加速度传感器, 采摘