Fusing Online Gaussian Process-Based Learning and Control for Scanning Quantum Dot Microscopy
| Autor: | Rolf Findeisen, Michael Maiworm, Christian Wagner, Maik Pfefferkorn, F. Stefan Tautz |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Physics - Instrumentation and Detectors
Basis (linear algebra) Computer science Measure (physics) Process (computing) FOS: Physical sciences 02 engineering and technology Sparse approximation Systems and Control (eess.SY) Instrumentation and Detectors (physics.ins-det) 021001 nanoscience & nanotechnology Electrical Engineering and Systems Science - Systems and Control 01 natural sciences 3. Good health symbols.namesake Quantum dot 0103 physical sciences Hyperparameter optimization Microscopy symbols FOS: Electrical engineering electronic engineering information engineering 010306 general physics 0210 nano-technology Gaussian process Algorithm |
| Zdroj: | 2020 59th IEEE Conference on Decision and Control (CDC) CDC |
| DOI: | 10.48550/arxiv.2004.02488 |
| Popis: | Elucidating electrostatic surface potentials contributes to a deeper understanding of the nature of matter and its physicochemical properties, which is the basis for a wide field of applications. Scanning quantum dot microscopy, a recently developed technique allows to measure such potentials with atomic resolution. For an efficient deployment in scientific practice, however, it is essential to speed up the scanning process. To this end we employ a two-degree-of-freedom control paradigm, in which a Gaussian process is used as the feedforward part. We present a tailored online learning scheme of the Gaussian process, adapted to scanning quantum dot microscopy, that includes hyperparameter optimization during operation to enable fast and precise scanning of arbitrary surface structures. For the potential application in practice, the accompanying computational cost is reduced evaluating different sparse approximation approaches. The fully independent training conditional approximation, used on a reduced set of active training data, is found to be the most promising approach. Comment: This paper is currently under review for CDC 2020 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|