Size Optimization of Space Structures Using Genetic Algorithm and Gradient Methods

Weight optimization is one of important parameters in space structure design. Size optimization is usually performed using gradient or genetic algorithm. Gradient algorithm is based on derivation of objective function and constraints of problem. The performance of gradient method is depended on start point and do not search all design domain. Genetic algorithm searches all design domains, but it cannot get close to the global optimum. In this paper, a new method is presented for size optimization. The algorithm starts with genetic algorithm and result of genetic algorithm is then used as start point for gradient algorithm. The presented method is used for size optimization of two trusses with three and ten elements. It is also applied on for optimization of a lattice structure of parabolic antenna. The results show that the present algorithm can perform better results compared to genetic algorithm alone.