PARTICLE SWARM SIZING OPTIMIZATION OF PLANAR TRUSS STRUCTURES WITH BUCKLING CONSTRAINTS

Abstract

This research examines the use of PSO sizing optimization for the planar truss mass minimization problem. In order to achieve practically applicable results, the examples used in the paper are also subjected to Euler buckling constraints. All examples use discrete sizing variables to generate models which can be constructed from bar element sizes available in the given material. The complexity added by implementing a dynamic constraint for buckling significantly influences the search space and makes finding global optima more difficult. Test example results are compared to results using GA from literature. As an additional parameter which influences the overall cost of a truss structure, the overall outer areas of all optimal models are compared.