Research on optimization efficiency of Genetic Algorithms

In order to evaluate the optimization efficiency of Genetic Algorithms (GA), this paper presents an efficiency evaluation criterion based on average optimization generation and time efficiency of GA, which not only can avoid infection evaluating the efficiency of GA on random factors commendably, but also consider the time firstly. So that it provides gist of evaluation criterion and theory for selecting the efficient GA parameters. According to this criterion, we have made an evaluation and analysis for GApsilas efficiency influence about the population size, crossover probability and mutation probability. Based on the statistical of function F2, simulation result shows the highest efficiency when GApsilas population size, crossover probability, mutation probability are 30, 0.7~0.8, 0.001~0.05 respectively.