基于自适应参数趋近律的机械臂滑模控制  

Sliding Mode Control of Manipulator Based on Adaptive Parameter Reaching Law

作　　者：贾立蒙 侯明[1] JIA Limeng;HOU Ming(School of Automation,Beijing Information Science and Technology University,Beijing 100192,China)

机构地区：[1]北京信息科技大学自动化学院,北京100192

出　　处：《机床与液压》2024年第23期51-57,共7页Machine Tool & Hydraulics

基　　金：北京市教委科技计划项目(KM202411232007)。

摘　　要：针对图像采集机械臂在滑模控制过程中存在的精度较低、抖振幅度较大、运动过程易受干扰等问题,提出一种基于自适应参数调节趋近律的机械臂神经网络滑模控制方法。在经典的指数函数趋近律基础上,对等速项系数进行自适应控制。通过引入机械臂连杆长度与转动惯量等参数,构建机械臂名义模型。利用自适应RBF神经网络有利于实时控制和补偿的特点,将其与设计好的线性滑模面相结合。通过合理设计李雅普诺夫函数,对所构建的系统进行稳定性证明。同时,与神经网络自适应指数趋近律及神经网络自适应饱和函数趋近律等方法进行对比。结果表明:在相同条件下,该方法能有效抑制抖振并实现高精度轨迹跟踪。Aiming at the problems such as low precision,large buffeting amplitude and easy interference of image acquisition manipulator in the process of sliding mode control,a neural network sliding mode control method based on adaptive parameter adjustment approach law was proposed.Based on the classical exponential function approach law,the constant velocity coefficients were adaptive controlled.The nominal model of the manipulator was constructed by introducing parameters such as the length of the connecting rod and the moment of inertia.The adaptive RBF neural network was used to facilitate real-time control and adaptive improvement,and was combined with the designed linear sliding mode surface.A suitable Lyapunov function was designed to prove the stability of the designed system.At the same time,it was compared with the adaptive exponential reaching law and adaptive saturation function reaching law of neural network.The results show that the method can effectively suppress chattering and achieve high precision trajectory tracking under the same conditions.

关 键 词：机械臂 自适应RBF神经网络 自适应参数调节 滑模控制 轨迹跟踪 

分 类 号：TP242.2[自动化与计算机技术—检测技术与自动化装置]