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

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