Optimization of tensegrity structures

Abstract

This thesis presents a new approach to solve the optimization of articulated structures and especially looks into the performance of tensegrity systems compared to regular trusses. Volume is the objective to minimize and a wide range of constraints can be considered from the required mechanical constraints to more architectural ones. The resulting nonlinear non-convex mixed-integer optimization problem is approached with a twofold algorithm. A global genetic algorithm controls the connectivity and geometric variables while the cross-sectional area and prestress forces variables are dealt with an internal sequential quadratic programming algorithm. The performance of this approach is evaluated on a typical cantilever configuration.