改善低频线振动台加速度失真度的自适应重复学习控制

Adaptive repetitive learning control for improvement of acceleration distortion of low-frequency linear vibration table systems

为了抑制低频线振动台系统中存在的摩擦力、推力波动及死区特性,提出了一种新的自适应重复学习控制方法。利用函数逼近技术将死区不确定性表示成正交基函数的线性组合。所设计的控制律由自适应构件、等效PID构件、重复学习构件组成。自适应构件用于在线估计线性组合中的未知参数及摩擦力数学模型的参数;PID构件用来镇定低频线振动台系统,抑制非周期性扰动;重复学习构件在未知对象模型的情况下抑制周期性推力扰动,并提高对周期性输入信号的跟踪能力。采用Lyapunov理论设计的自适应重复学习控制律保证了闭环系统的渐近稳定性和跟踪性能。仿真结果表明该控制方法提高了低频线振动台的跟踪性能,并改善了加速度失真度。

An adaptive repetitive learning control （ARLC） scheme is presented to restrain the friction force, the ripple force and the dead zone characteristics in low-frequency linear vibration table systems. The function approximation technique is employed to transform the dead zone uncertainties into finite combinations of orthonormal basis functions. The control algorithm consists of an adaptive component, a PID component and a repetitive learning component. The adaptive component is used to estimate unknown parameters of finite combinations and those of the friction force model on line. The PID component can stabilize the low-frequency linear vibration table system and suppress aperiodic disturbances. The repetitive learning component is used to restrain the periodic ripple force without knowledge of the plant model and improve the performence of tracking periodic input signals. The ARLC law designed by using the Lyapunov theory guarantees the system stability and the tracking performance. The simulation results demonstrate the proposed scheme can improve the tracking performance and the acceleration distortion for low-frequency linear vibration table systems.