基于自抗扰的快速刀具伺服系统复合控制

Composite control based on active disturbance rejection for a fast tool servo system

快速刀具伺服(fast tool servo,FTS)系统是实现微机械零件加工的关键部件。以压电陶瓷型FTS系统为研究对象,将迟滞状态时滞模型、时变时滞模型以及未建模动态非线性模型引入到FTS系统模型设计中以描述颤振现象,并基于神经网络控制提出具有迟滞时滞补偿功能的复合自抗扰控制方案,以实现FTS系统的颤振控制。其中,线性自抗扰控制将内部不确定性及迟滞时滞非线性和其他干扰视为总扰动并实时估计补偿,自适应BP(back propagation)神经网络用来对扰动估计误差进行逼近。与传统模型求逆方法相比,复合控制方案无需精确的数学模型易于初始设计。与线性自抗扰相比,复合控制方案减少了需要整定的参数数目,在相同带宽下具有更高的跟踪精度。仿真结果表明,所设计的复合控制具有更好的鲁棒性,能够实现压电陶瓷型FTS系统的快速精密跟踪控制。

A fast tool servo(FTS)system is a key component to realize the machining of micromachined parts.Taking a piezoelectric ceramic type FTS system as the research object,the hysteresis state delay model,the time varying delay model,and the unmolded dynamic nonlinear model were introduced into the FTS system model design to describe the chattering phenomena.Based on neural network control,an active disturbance rejection composite control scheme with hysteresis delay compensation function was proposed to realize the chattering control of the FTS system.The linear active disturbance rejection control(LADRC)considers the internal uncertainty,delay nonlinearity,and other disturbances as total disturbance and estimates compensation in real time,and an adaptive BP neural network was used to approximate the disturbance estimation error among them.Compared with the traditional model inversion method,the composite control scheme is easy to be designed initially without a precise mathematical model.Compared with the LADRC,the composite control scheme reduces the number of parameters that need to be adjusted,has higher tracking accuracy under the same bandwidth.Simulation results show that the designed composite control has better robustness,can effectively realize fast and precise tracking control of piezoelectric ceramic type FTS systems.