People Education Research Industrial Agenda  
 
Current Research Archive Publications PhD theses Software      

 
Previous      Up      Next

Distributed control for large adaptive optics systems

Project members:  R.M.L. Ellenbroek, prof. M. Verhaegen, R.F.M.M. Hamelinck (TU/e), prof. M. Steinbuch (TU/e), N. Doelman (TNO-TPD)
 
Keywords:  Modeling and analysis, Distributed control, Imaging and adaptive optics, Adaptive and learning control
 
Sponsored by:  TNO-TPD, IOP Precisietechnologie
 
It has been known for a long time that atmospheric turbulence limits the optical telescope image resolution because it distorts the wavefront of the incoming light. Without turbulence, image resolution can be improved by using a larger primary mirror. However, depending on the turbulence strength this trend stops at mirrors with a diameter of approximately half a meter. It is possible to overcome this limitation using an actively controlled deformable mirror or an Adaptive Optics (AO) system.

By now, AO systems are commercially available and AO is considered a proven technology. But there are still many open challenges, amongst which are scalability issues. As telescopes become ever larger, they require AO systems with ever more sensors and actuators. This yields problems not only for the mechanical design, but also for the control design. Distributed control approaches have to be considered, representing a new research field with many challenges.

The goal of this collaboration project between the TU Delft, the TU Eindhoven and TNO, is to design a new AO system consisting of a deformable mirror (DM) and a suitable control system. The distributed control design part is done at the DCSC group of the TU Delft. First steps in controller design consist of system and turbulence modeling. Although models have been created, distributed control requires conversion of those to distributed models.

Two bottlenecks in distributing these models arise from the wavefront sensor and from the coupling between actuators of the deformable mirror. Distributing a required processing step of the sensor signals is not trivial while the same holds for the calculation of actuator signals corresponding to a desired mirror shape.

Previous      Up      Next

Back to top

Last modified: 18 November 2011, 12:25 UTC
Search   Site map