全局收敛的带有输出约束柔性关节机械臂分段控制

针对带有输出约束和模型不确定的柔性关节机械臂系统,运用奇异摄动法将系统解耦成慢子与快子系统且分别进行控制器设计,从而实现与刚性控制方法的联系且能减少计算量.针对快子系统,采用速度差值反馈来抑制关节柔性引起的系统弹性振动.针对慢子系统提出了一种全局收敛的分段控制策略,将收敛域拓展到全局,克服了基于tan-障碍Lyapuov函数(BLF)反演控制需要系统初始误差在收敛域内的缺陷,且应用径向基(RBF)神经网络消除未知干扰和模型不确定性引起的误差,至此保证了系统的轨迹跟踪和输出约束要求.仿真对比表明,所提方法能使柔性关节机械臂在任意初始位置均能保持良好的跟踪性能,体现了控制器的有效性和优越性.

Globally convergent segmented control of flexible joint manipulator with output constraints

For flexible joint manipulator systems with output constraints and model uncertainties, this paper used the singular perturbation method to decouple the system into slow and fast subsystems and designed the subcontrollers separately, so as to realize the connection with the rigid control method and reduce the amount calculation. The fast subsystem used the speed difference feedback to suppress the elastic vibration of the system caused by the joint flexibility. For the slow subsystem, this paper proposed a globally convergent segmented control strategy, which extended the convergence range to the global level, and overcomed the defect that the initial error of the system needed to be within the convergence range for the backstepping control based on tan-barrier Lyapuov function(BLF). Then the radial basis(RBF) neural network eliminated errors caused by unknown disturbances and model uncertainties, thus ensuring the system''s trajectory tracking and output constraints of the system. The simulation shows that the proposed method can make the flexible joint manipulator maintain good tracking performance at any initial position, which reflects the effectiveness and superiority.