一种相对快速收敛的遗传算法的探讨

GA作为一种新的全局优化搜索技术比起其他搜索算法,优点明显,其不足之处是当搜索具有复杂染色体结构的求解空间时收敛速度慢。针对这问题提出了一种改进的相对快速收敛的GA算法的思路:增加对染色体的分割与重组操作,依据于各段的结构和段长,组成段群体,对其实施遗传操作以寻找优化段,重新组合成完整的染色体来搜索优化解。

Discussion on a Genetic Algorithm of Relatively Fast Convergence

As a new research technique of global optimization,the Genetic Algorithm has obvious advantages over other research algorithms.Its weakness lies in its low convergence rate while searching for the solution space having complicated chromosome structure.In order to speed up convergence rate,a new kind proposal about it is summed up in this thesis:increasing the operations(dividing and recombination)of the chromosome,composing segmented population based on structure and length of each segment,looking for optimal segment by carrying our genetic operation,and composing complete chromosome to search for optimal solution.