A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Robust real-time control of an adaptive optics system
This research contributes to the understanding of the limitations when designing a robust control real-time system for Adaptive Optics (AO). One part of the research is a new method regarding the evaluation of a Shack-Hartmann wavefront sensor (SHWFS) to enhance the overall performance. The method is presented based on the application of a Field Programmable Gate Array (FPGA) using Connected Component Labeling (CCL) for blob detection. The FPGA has been utilized since the resulting delay is crucial for the general AO performance. In this regard, the FPGA may accelerate the evaluation largely by its parallelism. The developed algorithm does not rely on a fixed assignment of the camera sensor area to the lenslet array to maximize the dynamic range. In extension to the SHWFS evaluation, a new rapid control prototyping (RCP) system based on hard real-time RTAI-patched Linux kernel has been developed. This system includes the required hardware e.g.~the analog output cards and FPGA based frame-grabber. Based upon a Simulink model, accelerated C/C++ code is automatically generated which uses the available parallel features of the processor. A continuative contribution is the application of a robust control scheme using H-infinity methods for designing a controller while considering uncertainty of the identified model. For synthesizing the controller, a special optimization technique called non-smooth mu-synthesis is utilized which minimizes the H-infinity norm while coping with pre-specified controller schemes. Depending on the pre-specified controller scheme, the resulting controller can be computationally costly but the RCP approach is designed to cope with the problem. Based on simulations and according to experiments, the validity of the identified models of the AO setup is assured. At the same time, the enhanced performance of the new RCP setup is demonstrated. ; Die wissenschaftliche Arbeit trägt maßgeblich zum Verständnis der gängigen Limitierungen bei robusten echtzeit-fähigen Regelungssystemen für Adaptiv ...
Robust real-time control of an adaptive optics system
This research contributes to the understanding of the limitations when designing a robust control real-time system for Adaptive Optics (AO). One part of the research is a new method regarding the evaluation of a Shack-Hartmann wavefront sensor (SHWFS) to enhance the overall performance. The method is presented based on the application of a Field Programmable Gate Array (FPGA) using Connected Component Labeling (CCL) for blob detection. The FPGA has been utilized since the resulting delay is crucial for the general AO performance. In this regard, the FPGA may accelerate the evaluation largely by its parallelism. The developed algorithm does not rely on a fixed assignment of the camera sensor area to the lenslet array to maximize the dynamic range. In extension to the SHWFS evaluation, a new rapid control prototyping (RCP) system based on hard real-time RTAI-patched Linux kernel has been developed. This system includes the required hardware e.g.~the analog output cards and FPGA based frame-grabber. Based upon a Simulink model, accelerated C/C++ code is automatically generated which uses the available parallel features of the processor. A continuative contribution is the application of a robust control scheme using H-infinity methods for designing a controller while considering uncertainty of the identified model. For synthesizing the controller, a special optimization technique called non-smooth mu-synthesis is utilized which minimizes the H-infinity norm while coping with pre-specified controller schemes. Depending on the pre-specified controller scheme, the resulting controller can be computationally costly but the RCP approach is designed to cope with the problem. Based on simulations and according to experiments, the validity of the identified models of the AO setup is assured. At the same time, the enhanced performance of the new RCP setup is demonstrated. ; Die wissenschaftliche Arbeit trägt maßgeblich zum Verständnis der gängigen Limitierungen bei robusten echtzeit-fähigen Regelungssystemen für Adaptiv ...
Robust real-time control of an adaptive optics system
Mauch, Steffen (author)
2016-09-29
Theses
Electronic Resource
English
Doktorarbeit , control closed-loop , thesis , SHWFS , ddc:621.3 , adaptive optics , FPGA , H infinity
Robust adaptive model‐based compensator for the real‐time hybrid simulation benchmark
| Wiley | 2022
Robust Adaptive Structural Control
| British Library Conference Proceedings | 1995
Robust decentralized adaptive compensation for the multi-axial real-time hybrid simulation benchmark
| DOAJ | 2024
Integration of Dynamic Traffic Assignment with Real-Time Traffic Adaptive Control System
| British Library Online Contents | 1998