伺服驱动系统无模型自适应控制

Research on Model-free Adaptive Control for Servo Driver System

高速高精的应用场合要求伺服驱动系统需要具备很好的动态响应性能和很强的鲁棒性来应对外界的干扰.但系统参数时变特性、不确定性以及非建模动态等因素导致伺服驱动系统的精确模型无法辨识得到.提出一种基于虚拟参考反馈校正的伺服驱动系统无模型自适应控制方法.该方法实时采集过程输入和输出数据,以当前系统运行状态的最新数据序列更新PI控制器参数,达到自适应控制的目的,保证系统的跟随性能.并且该方法结合稳定性约束条件以确保整定出来的伺服参数位于系统的稳定域内.仿真和实验结果表明,相比传统PI控制,提出的无模型自适应控制方法具有更好的动态响应性能、稳定性和鲁棒性.

High precision and high speed applications demand satisfied dynamic performance and strong robustness to against external disturbances for AC driver system. However,it was difficult to establish accurate mathematical models for the controlled system considering of time-varying parameters,uncertainness and unmodeled dynamics. A model-free adaptive control for AC servo driver system,based on enhanced virtual reference feedback control( VRFT),was proposed. The improved adaptive VRFT method collected the current process data at each sampling instant to update the controller parameters so that the good tracking performances were ensured. Stability constraints were incorporated in enhanced adaptive method to guarantee the stability the closed-loop system. Simulation and experimental results indicate that the proposed adaptive method processes better dynamic performances,stability and stronger robustness compared to traditional PI controller.