摘要
设计了一种基于支配关系下的局部搜索方法,将此局部搜索方法嵌入到多目标遗传算法中,从而提出一种有效的求解多目标优化问题的混合遗传算法。为加速遗传算法在全局优化问题上的收敛性,发挥传统数值优化算法在计算速度与计算精度上的优势,在遗传算法中镶嵌一个多目标线搜索算子。线搜索算子与遗传算法中的选择算子、交叉算子和变异算子共同作用,使全局搜索和局部搜索都能够很好的实现。数值实验表明,该混合遗传算法能求得问题的数量更多、分布更广的Pareto最优解。
A new local searching method based on dominance is presented , by joining the local searching method into the multi-objective optimization genetic algorithm , so a hybridized algorithm is proposed for solving multi-objective optimization problem. To accelerate astringency of genetic algorithm in global aid fatigue optimization and bring tradition numerical optimization in calculation speed and precision into play, a multi-objective line search operator is been set in genetic algorithm. The multi-objective line search operator interacts with selected operator, crossover operator and mutation operator, making global searching and local searching be well actualized. The numerical experiments show that this algorithm can find more and wider Pareto-optimal solutions than the original one.
出处
《天津工程师范学院学报》
2006年第3期24-26,35,共4页
Journal of Tianji University of Technology and Education
关键词
多目标优化
遗传算法
PARETO最优解
multi-objective optimization
genetic algorithm
Pareto-optimal solution
