基于RBF神经网络补偿的自适应滑模机械手控制  

Adaptive Sliding Mode Manipulator Control Based on RBF Neural Network Compensation

作　　者：刘洋洋[1] 院老虎[1] 腾英元 吴彪 张舒然 LIU Yangyang;YUAN Laohu;TENG Yingyuan;WU Biao;ZHANG Shuran(College of Aeronautics and Astronautics,Shenyang Aerospace University,Shenyang 110136,China)

机构地区：[1]沈阳航空航天大学航空宇航学院,沈阳110136

出　　处：《组合机床与自动化加工技术》2023年第6期119-122,127,共5页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家自然科学基金项目(11302134)。

摘　　要：针对机械臂在实际工作中无法完全建立精确的数学模型以及系统具有较大的不确定性,提出一种基于RBF神经网络补偿的自适应滑模鲁棒控制器。首先,利用计算力矩控制方法,结合基于RBF神经网络实现的补偿控制器使系统获得较好的跟踪效果;其次,引入滑模鲁棒控制,将建模误差以及外部干扰通过在线训练和调整参数对机械臂的未知不确定量部分进行实时估计和补偿,提高了跟踪精度以及提升了逼近的速度,同时,引入饱和函数降低了滑模控制中的抖振现象。仿真结果表明,改进后的控制器,对比文中传统控制器2和3在跟踪机械臂关节位置的响应速度上分别有约4 s和2 s的提升,同时,改进后的控制器在对于系统不确定性的补偿速度上有提升,约在1 s内做出响应。Aiming at the fact that the manipulator cannot completely establish an accurate mathematical model in practical work and the system has great uncertainty,an adaptive sliding mode robust controller based on RBF neural network compensation is proposed in this paper.First,using the computational torque control method,combined with the compensation controller based on the RBF neural network,the system can obtain a better tracking effect.Second,the sliding mode robust control is introduced,and the modeling error and external disturbance are adjusted through online training and parameter adjustment.The unknown uncertainty part of the manipulator is estimated and compensated in real time,which improves the tracking accuracy and the approximation speed.At the same time,the introduction of the saturation function reduces the chattering phenomenon in the sliding mode control.The simulation results show that the improved controller enhances the response speed of tracking manipulator joint position by about 4 s and 2 s respectively compared with the traditional controllers 2 and 3.Simultaneously,the improved controller accelerates the compensation speed of the system uncertainty and responds within about 1 s.

关 键 词：机械臂 RBF神经网络 自适应 滑模控制 

分 类 号：TH165[机械工程—机械制造及自动化] TG659[金属学及工艺—金属切削加工及机床]