摘要
针对具有参数不确定刚性机器人系统的跟踪控制问题,提出了一种基于视觉反馈和全调节RBF神经网络的自适应反演控制器设计方法。根据安装在末端执行器的CCD摄像机提取的特征点确定期望位置,利用与一般设计不同的全调节RBF神经网络逼近系统的不确定项及外界干扰。在调节RBF神经网络权值的同时调节中心点值和影响范围,使得全调节RBF神经网络具有了更强的在线逼近能力。应用Lyapunov稳定性理论,证明了系统的所有信号均有界,控制器可以保证机械臂的运动按指数收敛到期望位置。仿真结果验证了所提控制器的有效性。
An adaptive backstepping design scheme for rigid robot manipulators with parametric uncertainties is proposed based on the visual feedback and fully tuned radial basis function（RBF） neural networks.The feature extraction by the CCD camera mounted on the end effector makes the desired position.Fully tuned RBF neural networks which are different from the general RBF neural networks are used to approximate parametric uncertainties and disturbances of the system.The center value and incidence as well as the weight of the fully tuned RBF neural networks are adjusted.Therefore,the approximate ability of the networks is improved on line.It is proved by Lyapunov stability theory that all signals of the system are bounded and the controller guarantees exponential convergence of the motion of robot manipulators to the desired position.Finally,simulation results show the effectiveness of the proposed controller.
出处
《机械设计》
CSCD
北大核心
2012年第12期28-33,共6页
Journal of Machine Design
基金
河北省自然科学基金资助项目(F2012203088)
关键词
机器人
视觉反馈
全调节RBF神经网络
自适应控制
反演
robot manipulator
visual feedback
fully tuned RBF neural networks
adaptive control
backstepping
