Control enhancement of an elliptically bent hard x-ray dynamic mirror bender with machine-learning techniques
The hard X-ray adaptive mirror optics will play an important role at next generation light sources. A dynamic mirror bender with capacitive sensor array as an in-situ mirror profiler is used for initial test for hard x-ray zoom optics has been designed and constructed. Previous work showcases the dynamic control of this elliptically bent hard X-ray mirror through applying a combination of neural networks algorithm and feedback control. In this paper, we present further control enhancement with machine learning techniques through optimization of the number and placement of the capacitive sensors and new sensor calibration with video-based coordinate measuring machine.
Paper
Full text
Control enhancement of an elliptically bent hard x-ray dynamic mirror bender with machine-learning techniques
Semantic Scholar · Engineering · 2021
Abstract
The hard X-ray adaptive mirror optics will play an important role at next generation light sources. A dynamic mirror bender with capacitive sensor array as an in-situ mirror profiler is used for initial test for hard x-ray zoom optics has been designed and constructed. Previous work showcases the dynamic control of this elliptically bent hard X-ray mirror through applying a combination of neural networks algorithm and feedback control. In this paper, we present further control enhancement with machine learning techniques through optimization of the number and placement of the capacitive sensors and new sensor calibration with video-based coordinate measuring machine.