采摘机械手结构和控制系统优化—基于神经网络武术套路训练

摘要

Picking robot has broad application prospects in agricultural production,and its picking efficiency and picking quality are mainly affected by the structure of robot hand.In order to improve the operating performance of the picking ro-bot,Wushu training hook action into the manipulator structure design of the picking robot,according to the shape of the boxer hook picking robot movement by optimization, and combined with the neural network training method of Wushu movement error,the robot control system is improved, thus effectively improving the manipulator flexibility.In order to verify the feasibility and reliability of the scheme, the picking manipulator is verified by experiments, the experimental results show that the optimized manipulator structure and control system, the picking range is larger, can satisfy the operation requirement of three-dimensional space, picking manipulator moves faster in response,less error, high preci-sion and fast picking meet the demand and provide technical reference for the innovative design of picking manipulator.%采摘机器手在农业生产中具有广阔的应用前景,其采摘效率和采摘质量主要受机械手结构的影响.为了提高采摘机器人的作业性能,将武术套路训练的勾拳动作引入到了采摘机器人的机械手结构设计中,根据勾拳手形对采摘机器人的动作进行了优化,并结合神经网络武术动作误差训练方法,对机械手控制系统进行了改进,从而有效地提高了机械手作业的灵活性.为了验证该方案的可行性和可靠性,对采摘机械手进行了实验验证.结果表明:对机械手进行结构和控制系统进行优化后,其采摘范围较大,可以满足三维空间作业需求,采摘机械手动作响应较快,误差较小,满足高精度快速采摘的需求.