遗传参数协同进化的自适应遗传算法

An Adaptive Genetic Algorithm Based on Parameters Cooperated with Evolution

在智能计算技术中,为了提高收敛速度和缩短运行时间,提出了一种遗传参数协同进化的自适应遗传算法,解决遗传算法优化复杂的多峰值函数时出现早熟的问题。在其中的选择运算中引入一种抗早熟机制,在交叉运算中动态调整交叉算子,并在交叉率的调整中考虑海明距离的因素。在仿真实验中,用新算法对几个典型的多峰值函数进行优化,实验结果表明算法可以有效的避免早熟问题,收敛性能明显优于基本遗传算法和自适应遗传算法,适用于信息处理和模式识别领域。

An Improved Adaptive Genetic Algorithm based on Parameters Cooperated with Evolution（PCEAGA） is introduced for solving the problem of premature of genetic algorithm.In the algorithm,a resisting premature mechanism is introduced into selection operation,crossover operator is dynamically adjusted in the process of crossover operation,and the hamming distance is considered as an important factor of self-adaptation of crossover ratio.In the simulation experiment,several multiple peaks functions are optimized by the new algorithm.The results prove that the algorithm avoids the convergence of local optimal solution and shows great advantages of simple genetic algorithm and adaptive genetic algorithm in convergence.