摘要
该算法充分利用算法前期搜索的有用信息训练出淘汰模式和优秀模式。交叉产生的新群体与上一代群体竞争后进入下一代。有淘汰模式指导变异提高变异能力。根据优秀模式缩短解空间,提高搜索精度和放大适应度值比例进行快速收敛寻优。仿真试验表明,收敛速度和最优解精度都有大幅度提高。
The bad scheme and excellent scheme with the prophase useful information searched by GA are concluded. The new population of crossover competes with the old population. The bad scheme improves the mutating ability. The excellent scheme shortens the solution space, increases the searching precision and zooms out the function fitness value and picks up the convergence and searches the excellent solution. A function optimization problem is presented with the MATLAB language to demonstrate that the convergence velocity and solution precision are improved widely.
出处
《计算机工程与设计》
CSCD
2004年第8期1261-1263,1308,共4页
Computer Engineering and Design
基金
广东省教育厅自然科学基金(200274)
