Convergence enhanced genetic algorithm with successive zooming method for solving continuous optimization problems

A dynamic parameter encoding method was previously presented by Schraudolph and Belew [J Mach Learn 9 (1992) 9] for solving optimizing problems using discrete zooming factors. In contrast, the current paper proposes a successive zooming genetic algorithm (SZGA) for identifying global solutions using continuous zooming factors. To improve the local fine-tuning capability of a genetic algorithm (GA), a new method is introduced whereby the search space is zoomed around the design point with the best fitness per 100 generations. Furthermore, the reliability of the optimized solution is determined based on a theory of probability. To demonstrate the superiority of the proposed algorithm, a simple genetic algorithm, micro-genetic algorithm, and the proposed algorithm were compared as regards their ability to minimize multi-modal continuous functions and simple continuous functions. The results confirmed that the proposed SZGA significantly improved the ability of a GA to identify a precise global minimum. As an example of structural optimization, SZGA was applied to the optimal location of support points for weight minimization in the radial gate of a dam structure. The proposed algorithm identified a more exact optimum value than the conventional GAs.