均匀设计理论与应用  

Theory and application of uniform designs

作　　者：贺平[1] 林共进 刘民千 许青松[4] 周永道 Ping He;Dennis K.J.Lin;Min-Qian Liu;Qingsong Xu;Yongdao Zhou

机构地区：[1]北京师范大学-香港浸会大学联合国际学院理工科技学部,珠海519087 [2]Department of Statistics,Pennsylvania State University,University Park,PA 16802,USA [3]南开大学统计与数据科学学院,天津300071 [4]中南大学数学与统计学院,长沙410083

出　　处：《中国科学：数学》2020年第5期561-570,共10页Scientia Sinica：Mathematica

基　　金：国家自然科学基金(批准号:11771220和11871288);国家“万人计划”科技创新领军人才项目(批准号:W03020195);天津市人才发展特殊支持计划“高层次创新创业团队”项目(批准号:TJTZJH-GCCCXCYTD-2-23);天津市“131”创新型人才团队项目;天津市自然科学基金(批准号:19JCZDJC31100)资助项目。

摘　　要：随着科学和技术的发展,试验涉及的因素越来越多,它们之间的关系更加复杂,特别是在高科技的发展中,面临多因素、非线性和模型未知等复杂性,因此,如何科学地组织试验就显得非常重要.常见的试验设计类型有正交设计、均匀设计和最优设计等.均匀设计的主要思想是把设计点均匀地散布在试验区域中,它具有试验点数可灵活选取、对模型稳健、适用于多类试验区域等诸多优点.本文综述均匀设计理论发展过程、最近进展及应用现状.With the development of science and technology, the numbers of factors in experiments become larger and larger, and the relationships among factors become more and more complex, and then it is very important to organize the experiments scientifically. Orthogonal designs, uniform designs and optimal designs are the popular types of experimental designs. The main idea of uniform designs is to scatter the design points on the experimental domain uniformly. Uniform designs have many advantages such as choosing the numbers of runs flexibly, robustness for models, and suitability for different types of experimental domains. This paper reviews the theoretical development, current progress and application of uniform designs.

关 键 词：计算机试验 偏差 优化 均匀设计 

分 类 号：O212.6[理学—概率论与数理统计]