Semidefinite programming and structural optimization

Researchers: Ulf Brännlund, Mathias Stolpe, Krister Svanberg.

Sponsor: The Swedish Research Council for Engineering Sciences (TFR).

Semidefinite programming is a matrix generalization of linear programming which recently has been proven to be valuable in solving many applications of convex optimization.

Structural optimization deals with computer-aided optimal design of load carrying structures. A typical objective is to minimize the structural weight subject to various constraints on structural stiffness and strength.

The subject of this project is to develop mathematical models and numerical methods for optimizing the topology of different types of load-carrying structures. Such problems are in general much harder than for example problems dealing only with element sizes. The emphasis in this project is development of methods for semidefinite programming with the special structure found in truss topology design problems.


Research 1999/2000
Robust control of electrical drives
Some problems in navigation and motion control of nonlinear systems
Last update: February 21, 2001 by Anders Forsgren, anders.forsgren@math.kth.se.