论文部分内容阅读
为了有效解决具有不确定性和多极小性的随机优化问题 ,提出了一类基于假设检验的遗传算法 .该方法通过多次评价来进行解性能的合理估计 ,利用遗传操作来进行解空间的有效搜索 ,采用假设检验来增加种群的多样性和算法的探索能力 ,从而避免遗传算法的早熟收敛 .基于典型的随机函数优化和组合优化问题 ,仿真研究了假设检验、性能估计次数、噪声幅度对算法性能的影响 ,验证了所提方法的有效性和鲁棒性
In order to effectively solve the stochastic optimization problem with uncertainties and multipolarity, a kind of genetic algorithm based on hypothesis testing is proposed, which estimates the performance of the solution by multiple evaluations and uses genetic operations to solve the problem of stochastic optimization In order to avoid the premature convergence of genetic algorithm, hypothesis testing is used to increase the diversity of population and the exploration ability of the algorithm.Based on the typical stochastic function optimization and combinatorial optimization problems, the hypothesis testing, the number of performance estimations, the noise amplitude pair The performance of the algorithm is verified, and the validity and robustness of the proposed method are verified