Smart Sampling Algorithm for Surrogate Model Development

详细信息    查看全文

• 作者：Sushant Suhas Garud^a ; ^b ; I.A. Karimi^a ; ^b ; ^{cheiak@nus.edu.sg} ; Markus Kraft^a ; ^c ; ^d
• 关键词：Smart sampling ; Adaptive surrogate modelling ; Point placement
• 刊名：Computers & Chemical Engineering
• 出版年：2017
• 出版时间：4 January 2017
• 年：2017
• 卷：96
• 期：Complete
• 页码：103-114
• 全文大小：1487 K
• 卷排序：96

文摘

Novel sampling technique for surrogate modelling of arbitrary functions. Sample points are placed adaptively based on spatial and quality considerations. Optimal sample placement is formulated as an NLP using surrogate models. Our technique outperforms uniform, random, and Sobol sampling on 1-variable test problems.