基于代沟信息的可变种群规模遗传算法

Genetic Algorithm with Varying Population Size Based on Generational Gap Information

针对遗传算法种群规模难以估计的问题,提出了一种基于代沟信息的可变种群规模遗传算法。利用相邻几代群体间的极优解差异信息,在遗传算法发生早熟现象时根据逻辑斯蒂模型来改变种群规模,能以较小的计算代价获得与其它遗传算法性能相近的解。实验结果证明了算法的有效性。

Focusing on the problems existing in some genetic algorithms that population size is difficult to estimate or compute, a genetic algorithm with varying population size based on generational gap information （SAVPGA） is proposed. By using the information of local optimal solutions difference among some neighbor generation, SAVPGA can obtain approximate solutions with less computational cost by varying population size based on logistic model when premature convergence occurs. The computing of some examples is made to show that the algorithm is useful.