Model-Based Control of a High-Precision Measurement Machine for Multiscale Inspection Tasks

Quality inspection of micro systems on wafer scale must cope with opposing demands: nanometre accuracy and high velocity in a comparatively large workspace. An Automated Multiscale Measurement System (AMMS) combines multiple sensors that operate at specific scales by an intelligent measurement strategy in order to balance speed and accuracy. The AMMS demonstrator is based on a modified Mahr MFU100 with a position and tilt measurement resolution of 1 nm. In this paper multiscale models of a horizontal linear axis with an operating range of 200 mm and their application to state linearization and control parameterization of the residual dynamics are developed and discussed. Main modeling issues are sliding and submicron presliding friction for model-based control, reaching from submicron positioning to high-velocity trajectory tracking with desired performance. Therefore, sliding friction parameters and the friction state of an elasto-plastic friction estimator are adapted online. Experimental results show the effectiveness of the approach for fast submicron and large scale positioning.