基于生物入侵思想的自适应遗传算法优化

Exploration of improved adaptive genetic algorithm based on the idea of biological invasion

为了改进遗传算法的性能,针对标准遗传算法存在的早收敛及稳定性差的缺点,分析这一问题的原因及应对措施。选择使用基于生物入侵思想的自适应遗传算法进行优化,根据种群的实际情况对交叉概率、变异概率及入侵概率进行动态调整,使算法具有更好的鲁棒性。通过对标准遗传算法、自适应遗传算法、改进的自适应遗传算法和基于生物入侵思想的自适应遗传算法进行仿真验证得知,优化后的算法具有更好的适应度曲线,表明其能够有效的克服不成熟收敛,具有更快的收敛速度及更好的的稳定性。

In order to improve the performance of the genetic algorithm,this essay first aims at the shortcomings of premature convergence and poor stability in standard genetic algorithm. And it analyzes the causes of this problem and its countermeasures. It then chooses the adaptive genetic algorithm that is based on biological invasion（ IIAGA）to improve these shortcomings. In order to achieve better robustness,the study conducts a dynamic adjustment of the crossover probability,mutation probability and invasion probability ac-cording to the actual situation of the population. Finally,by way of the verified simulation of SAG,AGA,IA-GA and IIAGA algorithm,we know that the optimized algorithm has better fitness curves. It proves that it can effectively overcome the difficulty of the premature convergence,and it has faster convergence speed and better stability.