Impact of population size on bat algorithms performance

In order to optimize the performance of the bat algorithm, and improve the algorithm′s own optimization ability,the effect of performance on population size was analyzed from the bat algorithm convergence speed, solution accuracy, global search ability and convergence time.Eight classical test functions were used to simulate different population sizes.The results shows that when the population size of single-peak and multi-peak problems is 70 and 170,the algorithm has the best performance.