自适应局部微调遗传算法

A genetic algorithm with adaptive local adjustment

针对遗传算法在有限时间内难于给出高精确度解的问题,在传统遗传操作的前期与后期分别采用自适应与局部微调相结合的方法,给出了一种自适应局部微调的遗传算法,即将遗传代数划分为自适应概率搜索阶段和局部微调阶段,在交叉操作中分别采用自适应概率算术交叉和部分确定性诱导交叉;在变异操作中分别采用自适应随机扰动变异操作和最优个体诱导变异操作。应用该算法对全局最优解领域进行搜索,能在较短的时间内找到高精确度的数值解。对6个典型测试函数的优化问题实验表明,该方法具有快速、稳定和易于实现的优点。

It is difficult for genetic algorithm to give a precise solution in limited time. In order to solve this problem, an adaptive and local adjustment algorithm is adopted in the prophase and anaphase of traditional genetic algorithm, and a genetic algorithm with adaptive local adjustment is presented. In this algorithm, the numbers of genetic generations are divided into two steps, namely adaptive probability search and local adjustment. Adaptive probability arithmetic cross and portion certain abduction cross are adopted in the cross operation, and adaptive stochastic perturbation mutation operation and optimal individual abduction mutation operation are adopted in mutation operation. The high precision numerical solution can be found in a short time, when using this algorithm to search the domain of global optimal solution. The experiments of optimal problem of 6 typical testing functions are given. The experiments indicate that this method is fast, stable and easy to implement.