摘要
采摘机械臂在夹住柔性果茎后运输果实时,执行器末端的加减速运动使得果实在移动过程中产生摆动,易引发掉落,进而导致采摘失败.本文以单个西红柿作为负载,将果茎近似为柔性连杆.由于每一个果实的质量是不同的,因此,针对机械臂抓取可变柔性负载移动过程中的振动抑制问题,提出了自适应输入整形控制方法.当系统模型由于负载的不确定性发生变化后,传统的输入整形算法无法抑制柔性连杆移动过程中产生的振动.因此采用自适应输入整形算法,实时计算脉冲的幅值和时间.构造二次性能指标函数,通过对机械臂移动的加速度和负载的摆角实时数据进行迭代运算,达到零残余振动的目的.仿真实验结果表明,在变负载情况下,自适应输入整形算法有良好的末端振动抑制能力,获得满意的控制效果.
The acceleration and deceleration motion of the end of the actuator causes the fruit swing during the movement after vegetable harvesting robot arm clamp the stem,which is easy to lead to fruit falling and failure of picking.In this paper,a single tomato is treated as a payload and stem as flexible link.An adaptive input shaping algorithm for vibration control of variable flexible load of the fruits and vegetable harvesting robot arm was proposed as the mass of each fruit is different.Traditional input shaping algorithm could not restrain vibration of flexible link in the process of movement while the system model changed during load varied.Then an adaptive input shaping algorithm where amplitude and timing of impulses are tuned during operation to match the system under control.Solutions giving zero residual vibration are formulated in terms of a quadratic cost function and constructed by iterative operations on measured sets of robot arm and slant angle of load data.The adaptive input shaping algorithm is tested on a one-dimensional numerical control slide rail,and the results show the adaptive input shaping algorithm can solve load-varying vibration control problem.Simulations are carried out and tail vibration suppressing ability is qualified.
作者
刘德馨
张建成
李媛
方建军
LIU De-xin;ZHANG Jian-cheng;LI Yuan;FANG Jian-jun(College of Robotics,Beijing Union University,Beijing 100101,China;College of Urban Rail Transit and Logistics,Beijing Union University,Beijing 100101,China)
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
2022年第6期1043-1050,共8页
Control Theory & Applications
基金
北京市教育委员会科技计划一般项目(KM202011417003)
北京联合大学科研项目(ZK30202102)资助。
关键词
采摘机械臂
可变柔性负载
振动控制
自适应输入整形
harvesting robot arm
variable flexible load
vibration control
adaptive input shaping