网格下基于并行混合GA的复杂函数优化算法

Complex Data Optimization on Parallel Hybrid GA with Grid

为了解决传统单一GA在解决复杂函数优化时容易陷入局部最优的问题,文中结合模拟退火和网格服务的思想提出了网格下基于并行混合GA的复杂函数优化算法CDOPHGA-Grid。通过比较仿真试验表明:CDOPHGA-Grid算法的收敛速度随着网格节点个数的增加而增加;在相同情况下,CDOPHGA-Grid算法比传统单一的GA的收敛速度提高了约60倍。

To overcome the problem of local minima on standard genetic algorithm （SGA） for complex functions optimization, this paper proposes a novel complex data optimization algorithm on Parallel Hybrid GA with Grid （CDOPHGA -Grid）. A benchmark function is selected as the test functions. The experimental results show that the convergence velocity of CDOPHGA- Grid algorithm increases with Grid＇s node number; the convergence velocity of CDOPHGA-Grid algorithm is about 60 times of SGA＇s under the same conditions.