Frequency-Domain Data-Driven Adaptive Iterative Learning Control Approach: With Application to Wafer Stage

Abstract

The feedforward control is becoming increasingly important in ultra-precision stages. However, the conventional model-based methods can not achieve expected performance in new-generation stages since it is hard to obtain the accurate plant model due to the complicated stage dynamical properties. To tackle this problem, this paper develops a model-free data-driven adaptive iterative learning approach that is designed in the frequency-domain. Explicitly, the proposed method utilizes the frequency-response data to learn and update the output of the feedforward controller online, which benefits that both the structure and parameters of the plant model are not required. An unbiased estimation method for the frequency response of the closed-loop system is proposed and proved through the theoretical analysis. Comparative experiments on a linear motor confirm the effectiveness and superiority of the proposed method, and show that it has the ability to avoid the performance deterioration caused by the model mismatch with the increasing iterative trials.Index Terms-adaptive ILC, data-driven, frequency domain, feedforward control, linear motor. I. INTRODUCTIONN OWADAYS, the ultra-precision motion stages have been widely applied in many nanoscale manufacturing industries like IC manufacturing [1]. The precision motion stage has Manuscript

Trajectory Tracking and Vibration Control of Flexible Beams in Multi-Axis System

“…The stability and convergence of the ILC design must satisfy conditions (35) and (36), the frequency domain characteristics Gpc(z) = Mpc(ω)e jθ(ω) , Mpc(ω) is amplitudefrequency characteristics and θ(ω) is phase-frequency characteristics. Similarly, L(z) = ML(ω)e jφ(ω) , L(z) adopts PD learning function, and Q(z) = MQ(ω)e jψ (ω) , Q(z) is a third-order zero-phase Butterworth filter without any lag.…”

“…The amplitude-frequency characteristic and phase-frequency characteristic of E(ω) must satisfy conditions (35) and (36). The parameters of X-axis PD learning algorithm kp = 0.8, kd = 0.014, and the parameter of Y-axis kp = 0.8, kd = 0.02, the the phase margin angle ε=10 °.…”

“…ILC is a suitable algorithm to reduce the contour error of automated equipment to perform periodic, repetitive operation conditions. Using the previous cycle's error information makes the transient response and tracking accuracy of the current cycle system improved [30], [31] The ILC learning function of P (proportional)-type, D (differential)-type, and PD (proportional-derivative)-type is arguably the most widely used learning function that does not depend on the system's accurate model [32]- [35]. In [36], ILC is designed and analyzed in the frequency domain, and various factors are considered, illustrative examples on the designs of the filters and learning functions are also given.…”

An Adaptive Data-Driven Iterative Feedforward Tuning Approach Based on Fast Recursive Algorithm: With Application to a Linear Motor

“…A relatively accuracy model equalling to the inverse of the system plant is required in the model-based feedforward [10], which leads to its high dependence on both the model quality of the approximate model and the accuracy of the modelinversion. In contrary, ILC requires less prior knowledge of the system plant and outperforms the model-based feedforward in applications executing repeated tracking tasks [11]. However, ILC is highly sensitive to the variations of the reference trajectory resulting in limitation of its application.…”

Time‐iteration‐domain integrated learning control for robust trajectory tracking and disturbance rejection: With application to a PMLSM

“…It is essential to address the issues of trajectory tracking and disturbance rejection when the controller of a motion system is designed [1] [2]. For many industrial motion systems, such as pick-and-place robots [3], flatbed inkjet printers [4], positioning stages [5] [6] and wafer scanners [7] [8], a prominent feature is that they typically involve repetitive tasks [9] [10]. Iterative learning control (ILC) methods have been widely used to improve motion performance when reference trajectories and external disturbances are strictly repetitive [11] [12].…”