摘要
针对柔性机械臂在运动过程中受到柔性因素的影响会出现剧烈振动的问题,提出一种采用径向基(RBF)神经网络辨识的柔性机械臂抑振控制策略,通过减弱机械臂转角波动的方式间接抑制振动。首先,根据拉格朗日原理和假设模态法建立柔性机械臂的动力学模型,其中的不确定项由模态坐标和转角耦合的非线性项构成;其次,在控制律的设计中采用RBF神经网络对动力学模型的不确定项进行辨识补偿,从而提高驱动力矩的精度;最后,通过调整神经网络权重自适应律的系数,使包含辨识结果的控制律满足李亚普诺夫稳定性定理,从而保证动力学系统的稳定性,其中权重自适应律由高斯函数和误差向量组成。采用柔性机械臂实物控制平台的对比实验结果表明:所提出的控制策略能够有效减小柔性机械臂的转角误差和振动幅值;当柔性机械臂长度为1.5 m时,相比常规比例微分控制策略,采用RBF神经网络辨识的控制策略使机械臂末端振动敏感方向的加速度的方差下降了5.8%。该控制策略为柔性机械臂的振动抑制提供了新思路。
To solve the problem that flexible manipulators will vibrate violently under the influence of the flexibility factors during the moving process,a vibration suppression control strategy for flexible manipulators using the radial basis function(RBF)neural network identification is proposed.Vibration is suppressed indirectly by weakening the rotation angle fluctuation of flexible manipulators.First,the dynamic model of flexible manipulators is established according to the Lagrangian principle and the assumed modal method.The uncertain term is composed of the nonlinear terms coupled with the modal coordinates and the rotation angle.Secondly,the RBF neural network is used in the design of the control law to identify and compensate the uncertain terms of the dynamic model,so as to improve the precision of the driving torque.Finally,by adjusting the coefficients of the neural network weight adaptive law,the control law containing the identification results satisfies the Lyapunov stability theorem,so as to ensure the stability of the dynamic system.The weight adaptive law consists of the Gaussian function and the error vector.The comparative experimental results using the physical control platform of the flexible manipulator show that the proposed control strategy can effectively reduce the rotation angle error and the vibration amplitude of the flexible manipulator.When the flexible manipulator length is 1.5 m,compared with the conventional proportional differential control strategy,the RBF neural network identification control strategy reduces the variance of the acceleration in the vibration-sensitive direction at the end of the manipulator by 5.8%.This control strategy provides a new idea for vibration.
作者
尚东阳
李小彭
尹猛
李凡杰
杨贺绪
SHANG Dongyang;LI Xiaopeng;YIN Meng;LI Fanjie;YANG Hexu(School of Mechanical Engineering&Automation,Northeastern University,Shenyang 110819,China;Shenzhen Institute of Advanced Technology,Chinese Academy of Sciences,Shenzhen,Guangdong 518055,China)
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2022年第6期76-84,共9页
Journal of Xi'an Jiaotong University
基金
国家自然科学基金资助项目(51875092)
国家重点研发计划重点专项资助项目(2020YFB2007802)
宁夏回族自治区自然科学基金资助项目(2020AAC03279)。
关键词
柔性机械臂
RBF神经网络
不确定项
振动抑制
非线性项
flexible manipulator
RBF neural network
uncertain term
vibration suppression
nonlinear term
