基于线杂交和面变异的遗传算法DGA

LINE CROSSOVER AND PLANE MUTATION BASED GENECIT ALGORITHM: DGA

遗传算法由于其并行性和对全局信息的有效利用能力在化学和化工界得到越来越广泛的应用。但经典的遗传算法存在着一些缺点， 如优化速度慢、空间搜索不均匀、搜索比较盲目等［１］ 。针对这些缺点， 我们提出了结合均匀设计、有方向的搜索和遗传算法的确定性遗传算法ＤＧＡ， 并用１８ 个经典测试函数和３ 个非线性规划问题对ＤＧＡ进行了测试。

Genetic algorithms (GAs) have been applied widely in chemistry and chemical field due to their parallelism and effective utilization of global information. However, classical GAs have some flaws, such as low optimization speed, heterogeneous space search, blind search directions etc. Aimed at these shortages, a deterministic GA(DGA) combined with homogeneous design, directed search and genetic algorithms is proposed, and a test for DAG by 18 classical test functions and 3 nonlinear programming problems is performed.