摘要
遗传算法是一种有效的全局优化算法 ,但存在收敛速度慢和早熟收敛的缺陷。本文提出了具有适应值曲面结构自学习能力的多区域并行局部搜索算子PLS和受控交叉算子GC ,定性地分析了它们的作用机制。引入适应性PLS和GC的改进遗传算法在不增加计算开销的前提下 ,其全局收敛速度和可靠性显著地优于标准遗传算法 。
GAs(Genetic Algorithms)are well known as a class of efficient global optimizing methods,but have slow convergent velocity and are subject to pre-maturing stagnation.This paper proposed a kind of parallel local search operator PLS and a guided crossover operator GC that have self-learning ability of the structure of fitness landscape.The operation mechanisms of the two proposed operators were qualitatively analyzed.Both the convergence velocity and the global convergence reliability of the improved GAs introducing PLS and GC excel greatly that of standard ones,and have good robustness and stability.
