一种基于增量元模型方法的全局优化算法

摘要

A new global optimization method based on incremental metamodel was proposed for the complex simulation model problem. Firstly, to overcome the defects of existing incremental Latin hyper-cube design, which was hard to control the number of sampling points and limited to multiples, we proposed an improved incremental Latin hyper-cube sampling method based on subtraction rule idea. Secondly, combined incremental Latin hyper - cube design and the method of incremental update radial basis functions, we proposed a new efficient global optimization algorithm. Finally, the method was applied to a pressure vessel design problem and the results of the example demonstrate the efficiency and engineering practicability of the presented method.%针对复杂仿真模型的全局优化问题,提出一种基于增量元模型方法的全局优化算法.首先,分析了现有增量拉丁超立方采样方法新增点数量难以控制以及其数值必须为原采样点数量的整数倍的缺陷,在此基础上利用减法规则思想改进了增量拉丁超立方采样；其次,将改进后的增量拉丁超立方采样与径向基函数的增量更新方法相结合,提出了一种全新的高效全局优化算法；最后,将该算法应用于压力容器的优化设计,计算结果证明了该方法的实用性与工程有效性.