An Improved Type REAL Survival Choice Based on Comparison of the Differential Parent and Its Performance Evaluation

Usually, selection operations of individual in the genetic algorithm have the duplicate selection as survival choice, based on these operations of generational change, various generational change models to keep the diversity have been proposed. Especially, by intentionally making difference in evolution speed of individuals, an efficient generational change model REAL has been presented. In this paper, we propose an improved type REAL by add a comparison of differential parent. Some numerical experiments for benchmark function optimization problems, the proposed algorithm show relatively good performance evaluation with comparison of REAL and the various generational change models.