An Optimal Design of Genetic Operation

Simple genetic algorithm is prone to premature and has a slow convergence. In view of the above inadequacies, we carried on an optimal design of selection operation and proposed a model of selection operation creatively in this paper. That is the strategy of ultra expectations - compared selection. Besides, we proved its advantages in global optimal solution and convergence speed. It can avoid effectively the premature maturation of genetic algorithm, maintain the diversity of populations and accelerate the solving speed.