Heuristic Methods in Space Frame Optimization

Abstract

*† In this study four heuristic optimization algorithms are used as a solution method for a discrete space frame sizing optimization problem. Suitable profiles for each beam have to be selected from a given standard selection and the mass of frame is minimized regarding to stress, displacement, buckling and frequency constraints. Selected heuristic algorithms are simulated annealing, tabu search, genetic algorithm and particle swarm optimization. The main idea is to compare the efficiency of these algorithms by using example problems. The criteria for the efficiency is considered to be the improvement of the object function as the function of needed FEM-analysis. Numerical calculations show that population based methods (genetic algorithm and particle swarm optimization) work better than local search algorithms (simulated annealing and tabu search).