Funding: SenterNovem
Running Period: 2009-2013
Staff: George Still
Ph.D. student: Matthijs Bomhoff
Project Website
ABSTRACT
The goal of the project is to develop a next generation of CAD tools. This new type of synthesis based computer tools help mechanical engineers to design solutions of higher quality in a significantly shorter time than is currently done. Close cooperation with industrial partners insures practical applicability.
Subproject: Large solution spaces
The synthesis tool will in general allow choosing the design from a large number of feasible constructions in a high dimensional space. The goal is then to find ‘the best design’.
Only for ‘simple designs’ this optimization problem can be expected to be solved exactly in an automatic process. In most practical cases the optimal choice has to be found interactively by ‘navigating’ in an efficient way through the space of feasible designs. Main aspects of the research are:
- techniques from Operations Research to find a ‘representative subset’ of the feasible set. (Multi-objective programming, sensitivity analysis.),
- techniques from mathematics and computer science which allow ‘rough visualization’ of a high-dimensional solution space and an efficient interactive selection of an optimal solution.