优化搜索空间划分的遗传算法的研究与实现

针对遗传算法中早熟收敛和容易陷入局部收敛的问题,提出优化搜索空间、遗传算法算子的一些改进策略,即利用搜索空间划分实现优良等位基因单元稳定遗传到下一代中,利用禁忌域和有效域快速提高算法的实现性能.改进的算法能有效减少搜索空间、避免算法早熟,使得算法的全局搜索能力和局部搜索能力比其他遗传算法均得到了较大的提高.函数求最优解和服装设计算法的实现,证明了改进算法的平均收敛速度和收敛到最优解的效率都优于其他遗传算法,实验验证了所提出的算法思想的可行性和有效性.

Study and Implementation of Genetic Algorithm Based on Improvment of Search Space Partition

In order to effectively solve the disadvantages that easily run into premature convergence and local extremism in genetic algorithm, the paper puts forward some improved strategies, which consist of genetic operators, the process of searching optima, reserving some elitist genetic sense units that can steadily pass down son generation, and preferably enhancing performance of genetic algorithm by making use of taboo field and satisfaction field. The efficiency of the proposed strategies is analyzed , and the improved algorithm helps better to reduce search space so that the algorithm enhances the globally and locally searching ability and greatly avoids premature convergence, as thus be superior to other genetic algorithms. The experimental validation for the improved algorithm has been passed through being applied to function optimization and fashion design and the experimental results show that the rapidity of convergence and the probability of the improved algorithm can be superior to other genetic algorithms, the simulations test out the feasibility and efficiency of proposed theories by experiment.