一种新的基于小生境的自适应遗传算法

针对遗传算法在全局优化问题中容易出现早熟和收敛速度慢的问题,根据群体适应值的分布特点,启发性地提出了一种新的基于小生境的自适应遗传算法(ANGA).采用一种新的适应值计算方法,引入了一个自适应的常数Cmin,根据群体中各个个体的适应值分布情况加以启发,通过自适应调整Cmin,以适时改变群体适应值的分布,优化了各个个体被选择的概率.同时采用了小生境技术,并对交叉和变异位置引入了自适应的非均匀选择机制.采用3个典型的全局优化测试函数进行了验证,仿真试验表明该方法能够明显地改善全局寻优能力,并大大加快了收敛速度.

A new self-adaptive genetic algorithm based on niches

To deal with the prematurity and low convergence speed when the genetic algorithm being used for global optimization,a new heuristic self-adaptive genetic algorithm(ANGA) based on niches was presented according to the feature of population fitness distribution.A new method for calculation of the fitness was adopted,and an adaptive constant C_(min) was introduced.By adjusting C_(min) for changing the population fitness distribution in proper time,the selection probability of every individual of the population was optimized.Meanwhile,niches were adopted,and adaptive hyperopic mechanism was introduced to select the position of crossover and mutation.Three classic test functions for global optimization were used to verify the effectiveness.The test results showed that ANGA could evidently improve global optimization ability and convergence speed.