Robust parameter design based on Gaussian process with model uncertainty

被引:18
作者
Feng, Zebiao [1 ]
Wang, Jianjun [1 ]
Ma, Yizhong [1 ]
Tu, Yiliu [2 ]
机构
[1] Nanjing Univ Sci & Technol, Dept Management Sci & Engn, Nanjing, Peoples R China
[2] Univ Calgary, Dept Mech & Mfg Engn, Calgary, AB, Canada
基金
中国国家自然科学基金; 加拿大自然科学与工程研究理事会;
关键词
robust parameter design; Gaussian process; model uncertainty; outliers; hypersphere decomposition; MULTIRESPONSE SURFACE OPTIMIZATION; DESIRABILITY FUNCTION-METHOD; JOINT OPTIMIZATION; PROCESS REGRESSION;
D O I
10.1080/00207543.2020.1740344
中图分类号
T [工业技术];
学科分类号
08 ;
摘要
In robust parameter design, it is common to use computer models to simulate the relationships between input variables and output responses. However, for the contaminated experimental data, the model uncertainty between computer models and actual physical systems will seriously impair the robustness of the optimal input settings. In this paper, we propose a new weighted robust design approach concerning the model uncertainty from outliers based on the robust Gaussian process model with a Student-t likelihood (StGP). Firstly, to reduce the impact of outliers on the output means and variances, the StGP modelling technique is adopted to estimate the relationship models for contaminated data. Secondly, the Gibbs sampling technique is employed to estimate model parameters for better mixing and convergence. Finally, an optimisation scheme integrating the quality loss function and confidence interval analysis approach is built to find the feasible optimisation solution. Meanwhile, the hypersphere decomposition method and data-driven method are applied to determine the relative weights of objective functions. Two examples are used to demonstrate the effectiveness of the proposed approach. The comparison results show that the proposed approach can achieve better performance than other approaches by considering the model uncertainty from outliers.
引用
收藏
页码:2772 / 2788
页数:17
相关论文
共 50 条
[21]   PID based nonlinear processes control model uncertainty improvement by using Gaussian process model [J].
Chan, Lester Lik Teck ;
Chen, Tao ;
Chen, Junghui .
JOURNAL OF PROCESS CONTROL, 2016, 42 :77-89
[22]   Consensus cubature filtering based on Gaussian process for distributed sensor network with model uncertainty [J].
Song, Lingyun ;
Jing, Zhongliang ;
Dong, Peng ;
Zou, Ruping ;
Chen, Shichao .
SIGNAL PROCESSING, 2022, 190
[23]   Gaussian process based optimization algorithms with input uncertainty [J].
Wang, Haowei ;
Yuan, Jun ;
Ng, Szu Hui .
IISE TRANSACTIONS, 2020, 52 (04) :377-393
[24]   ROBUST OPTIMIZATION WITH MIXED INTERVAL AND PROBABILISTIC PARAMETER UNCERTAINTIES, MODEL UNCERTAINTY, AND METAMODELING UNCERTAINTY [J].
Zhang, Yanjun ;
Xia, Tingting ;
Li, Mian .
PROCEEDINGS OF THE ASME INTERNATIONAL DESIGN ENGINEERING TECHNICAL CONFERENCES AND COMPUTERS AND INFORMATION IN ENGINEERING CONFERENCE, 2019, VOL 2B, 2020,
[25]   Model Predictive Controller Design Based on Residual Model Trained by Gaussian Process for Robots [J].
Wu, Changjie ;
Tang, Xiaolong ;
Xu, Xiaoyan .
JOURNAL OF MARINE SCIENCE AND ENGINEERING, 2023, 11 (05)
[26]   Robust parameter design based on response surface model considering measurement errors [J].
Wu, F. ;
Wang, J. ;
Ma, Y. .
SCIENTIA IRANICA, 2021, 28 (04) :2323-2332
[27]   Bayesian closed-loop robust process design considering model uncertainty and data quality [J].
Ouyang, Linhan ;
Chen, Jianxiong ;
Ma, Yizhong ;
Park, Chanseok ;
Jin, Jionghua .
IISE TRANSACTIONS, 2020, 52 (03) :288-300
[28]   Robust Gaussian process regression based on iterative trimming [J].
Li, Zhao-Zhou ;
Li, Lu ;
Shao, Zhengyi .
ASTRONOMY AND COMPUTING, 2021, 36
[29]   Robust parameter design based on online adjustment [J].
Wu F. ;
Ma Y. ;
Wang J. .
Jisuanji Jicheng Zhizao Xitong/Computer Integrated Manufacturing Systems, CIMS, 2019, 25 (10) :2487-2492
[30]   Model-Based Reliability Analysis With Both Model Uncertainty and Parameter Uncertainty [J].
Xi, Zhimin .
JOURNAL OF MECHANICAL DESIGN, 2019, 141 (05)