论文部分内容阅读
针对未知随机变量分布环境下的非线性概率优化模型,探讨微种群免疫优化算法。算法设计中,基于危险理论的应答模式,设计隐并行优化结构;经由自适应采样方法辨析优质和劣质个体;通过动态调整个体的危险半径确定危险区域和不同类型子群;利用多种变异策略指导个体展开多方位局部和全局搜索。该算法的计算复杂度依赖于迭代数、变量维数和群体规模,其具有进化种群规模小、可调参数少和结构简单等优点。借助理论测试例子和公交车调度问题,比较性的数值实验显示,此算法在寻优效率、搜索效果等方面均有一定的优势,对复杂概率优化模型有较好潜力。
In view of the nonlinear probability optimization model under unknown random variable distribution environment, the micro-population immune optimization algorithm is discussed. In the algorithm design, based on the hazard theory, the response model is designed to conceal the parallel optimization structure; the adaptive sampling method is used to discriminate between high quality and low quality individuals; the dangerous radius and the different types of subgroups are determined by dynamically adjusting individual risk radius; Individuals start multi-faceted local and global searches. The computational complexity of this algorithm depends on the number of iterations, the number of variables and the population size. It has the advantages of small evolutionary population, fewer adjustable parameters and simple structure. With the example of theoretical test and bus scheduling problem, comparative numerical experiments show that this algorithm has some advantages in terms of optimization efficiency, search effect and so on, and has better potential for the complex probability optimization model.