Enhanced Nonvector Space Approach for Nanoscale Motion Control

Scanning probe microscopy (SPM) based nanomanipulators always experience probe accuracy issues caused by uncertainties such as scanner hysteresis, drifts, and interaction forces. Although some SPMs contain internal position sensors, they can hardly sense real probe position due to the above uncertainties, especially when implementation scales down below 100 nm. Recently, we proposed a local-image-based nonvector space (NVS) control strategy for regulating nanomanipulator to enhance its accuracy (positioning error is generally maintained below $0.2\%$ of the operation range). Since the previous NVS control strategy employs only local image and pixel-wise information (neglecting global location and local coherence information), it can hardly track trajectories with relatively large “jump” and heavy noise, which might be inefficient for practical implementation. To tackle these problems, this study proposes a novel enhanced NVS (ENVS) control strategy, which performs rough global positioning control through dynamic hysteresis compensation for SPM scanners to solve the “jump” problem; to realize fine positioning control with possible heavily noisy feedback, the ENVS controller utilizes patch-wise-based rather than pixel-wise-based set feedback strategy, and the ENVS is built based on a newly established shape Lyapunov function. Testing results show that the ENVS control scheme can precisely track references with jump changes, such as the step signal. Furthermore, the ENVS controller is able to maintain its accuracy even with extremely noisy feedback.