Retaining Population Diversity in Genetic Algorithms By Introducing New Populations.
Susumu Sakano, Hirotsugu Nagami · TRANSACTIONS OF THE JAPAN SOCIETY OF MECHANICAL ENGINEERS Series C · 1998
The genetic algorithm (GA) simulates the process of natural evolutions and the usual GA is based on the mechanisms of natural selection in natural genetics : selection, crossover and mutation. The GA has problems of premature local convergences and long repitional computation which arise from a loss of population diversity. This paper proposes a new genetic algorithm technique by introducing new populations into each generations of the computation. To retain population diversity in GA, we propose the genetic algorithm by introducing activation method. This method is applied to the inverse kinematics of the five degrees robot and shows the effectiveness of this method better than the simple GA (SGA). The efficient difference between the new genetic algorithm and the commonly used genetic algorithm is shown as the comparison of the required genetic generations.