Public defense PhD thesis based on research in Condor project
On June 21, 2012, David Rijlaarsdam will defend his PhD thesis entitled 'Frequency domain based performance optimization of systems with static nonlinearities'. The defense is open to all interested and will take place at the TU/e in the Auditorium starting at 16:00.
The research presented in David’s thesis is part of the Condor project. This project investigated how to adapt a highly specialised, complex system to meet new market requirements, taking a FEI transmission electron microscope (TEM) as a prime example of a complex system.
The work presented in David’s thesis focuses on the analysis and optimal control of the mechanics of the motion stage in the electron microscope.
The results contribute to the project by supplying a new, improved methodology to assess and optimize the performance of the system by minimizing performance degrading nonlinear effects such as friction.
FEI electron beam microscopes make the Ångström world (a tenth of a million part of a millimetre) visible. At these levels of magnification, environmental influences, alignment issues, as well as instrument characteristics have significant impact on image quality. Operators, performing research by electron microscopes, are forced to continuously correct the microscope for these types of imperfections. Hence, because of the large amount of variables to correct, the operator needs to combine its skills as experienced researcher and experienced microscope operator.
New market segments with less skilled operators, such as quality control of mining processes, chemical bulk processes, and life science research, need to have electron beam microscopes that work without the necessity of manual correction of instrument imperfections. They are asking for robust, stable microscopes as industrial tools. The Condor project has adopted this challenge for the research areas of its twelve academic researchers.
The project resulted in techniques to extract data from images and to use that data for automated control of a wide variety of variables (parameters), without explicitly knowing the influence of each parameter on the images. For some disturbances it proved possible to reconstruct the images in real-time, without the need of correcting the instruments imperfections.
Much of the results can be used in other science fields and applications, such as image reconstruction in medical applications, real-time data extraction for observation, surveillance or transport related applications, as well as modelling techniques for analysis of drive components to overcome friction-induced imperfections.
Next to FEI Company as carrying industrial partner, a second industrial partner was Technolution, a company developing electronics and software solutions for technical information systems and embedded systems. They have know-how in the field of image processing techniques, and work as development partner for FEI.
ESI coordinated the project and was responsible for the overall project management and knowledge dissemination. ESI also coached the research activities by assigning research fellows to the project.
Results of the project are published in several (scientific) publications.