Extreme Precise Motion Tracking of Piezoelectric Positioning Stage Using Sampled-Data Iterative Learning Control

Positioning stages using piezoelectric stack actuators have been widely used in industrial applications. In this brief, we explore practical control algorithms that can achieve extreme precision motion tracking. Extreme precision is defined as the acquisition of tracking accuracy up to the hardware limit of a control system, for instance, the sensor resolution, the actuator resolution, the quantization limit of analog-to-digital converters (ADC), the limit of sampling interval, and the system repeatability. Sampled-data feedback control algorithms are unable to achieve such extreme precision tracking because of actuator saturation and stability margin. In this brief, we apply an iterative learning control (ILC) approach that can achieve the extreme precision for motion tracking tasks that repeat. ILC is essentially a feedforward control approach that fully utilizes the past control information, and hence is able to overcome the limit of feedback algorithms. The ILC algorithm used is simple in design and implementation, and fast in learning convergence. The sampled-data ILC is implemented on a piezoelectric positioning stage, in which the ADC device has a limited quantization of 49 nm. With only a few iterations of learning, the extreme precision motion tracking is achieved monotonically. Compared to well-tuned open-loop and proportional integral control algorithms, ILC can further reduce the tracking error by at least fourfold.