A multi-fidelity surrogate modeling approach for incorporating multiple non-hierarchical low-fidelity data

Zhang, Lili; Wu, Yuda; Jiang, Ping; Choi, Seung-Kyum; Zhou, Qi

doi:10.1016/j.aei.2021.101430

Cited by 25 publications

(4 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Prior publications have proposed multi-level MFSM frameworks for incorporating non-level data based on the weighted average of statistical surrogate models [34][35][36][37][38]. Several LF models were created in this scenario by utilizing various methods to simplify the HF model, leading to non-level LF models with varying degrees of correlation to the HF model in the subregion of the design space.…”

Section: Related Workmentioning

confidence: 99%

“…M. Xiang et al [36] proposed extended co-Kriging (ECK) to incorporate non-level LF models to improve the HF model's accuracy. Zhang et al [38] proposed an improved version of ECK, called NHLF-co-Kriging, featuring a different strategy to obtain optimal scaling factors of the LF models.…”

Section: Related Workmentioning

confidence: 99%

“…Co-Kriging, 2000 [22,24,30] √ Bayesian HK, 2012 [25] √ Bayesian MCK, 2012 [24] √ Bayesian IHK, 2018 [12] √ Bayesian MVCM, 2015 [19] √ BF GVFM, 2015 [20] √ BF co-BRF, 2017 [40] √ Bayesian MFGP, 2018 [41] √ Bayesian POD-co-Kriging, 2018 [42] √ Bayesian SM-VFM, 2018 [26] √ SM, BF MHK, 2020 [31,32] √ Bayesian MDNN, 2020 [13] √ BF GCK, 2020 [43,44] √ Bayesian TL-VFSM, 2021 [27] √ BF GAN-MDF, 2022 [29] √ BF MMGP, 2021 [45] √ Bayesian WS, PC-DIT, PC-CSC, 2016 [39] √ Bayesian ECK, 2018 [36] √ Bayesian LRMFS, 2018 [37] √ BF VWS-MFS, 2021 [46] √ Bayesian NHLF-co-Kriging, 2022 [38] √ Bayesian Our work (EHK)…”

Section: Granularity Of Lf Datasets Mfsm Framework Type Multi-level N...mentioning

confidence: 99%

See 2 more Smart Citations

Extended Hierarchical Kriging Method for Aerodynamic Model Generation Incorporating Multiple Low-Fidelity Datasets

Pham,

Tyan,

Nguyen

et al. 2023

Aerospace

View full text Add to dashboard Cite

Multi-fidelity surrogate modeling (MFSM) methods are gaining recognition for their effectiveness in addressing simulation-based design challenges. Prior approaches have typically relied on recursive techniques, combining a limited number of high-fidelity (HF) samples with multiple low-fidelity (LF) datasets structured in hierarchical levels to generate a precise HF approximation model. However, challenges arise when dealing with non-level LF datasets, where the fidelity levels of LF models are indistinguishable across the design space. In such cases, conventional methods employing recursive frameworks may lead to inefficient LF dataset utilization and substantial computational costs. To address these challenges, this work proposes the extended hierarchical Kriging (EHK) method, designed to simultaneously incorporate multiple non-level LF datasets for improved HF model construction, regardless of minor differences in fidelity levels. This method leverages a unique Bayesian-based MFSM framework, simultaneously combining non-level LF models using scaling factors to construct a global trend model. During model processing, unknown scaling factors are implicitly estimated through hyperparameter optimization, resulting in minimal computational costs during model processing, regardless of the number of LF datasets integrated, while maintaining the necessary accuracy in the resulting HF model. The advantages of the proposed EHK method are validated against state-of-the-art MFSM methods through various analytical examples and an engineering case study involving the construction of an aerodynamic database for the KP-2 eVTOL aircraft under various flying conditions. The results demonstrated the superiority of the proposed method in terms of computational cost and accuracy when generating aerodynamic models from the given multi-fidelity datasets.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Granularity Of Lf Datasets Mfsm Framework Type Multi-level N...mentioning

confidence: 99%

See 1 more Smart Citation

Extended Hierarchical Kriging Method for Aerodynamic Model Generation Incorporating Multiple Low-Fidelity Datasets

Pham,

Tyan,

Nguyen

et al. 2023

Aerospace

View full text Add to dashboard Cite

show abstract

“…least-angle regression (LAR) 58 , high-dimensional model representation (HDMR) 59 , are not suitable for general-purpose antenna modeling. Meanwhile, variable-resolution methods have been demonstrated to be beneficial in this context (co-kriging 60 , two-level GPR 61 ), also in combination with sequential sampling 62 – 65 .…”

Section: Introductionmentioning

confidence: 99%

Two-stage variable-fidelity modeling of antennas with domain confinement

Pietrenko‐Dabrowska

Kozieł

Golunski

2022

Sci Rep

View full text Add to dashboard Cite

Surrogate modeling has become the method of choice in solving an increasing number of antenna design tasks, especially those involving expensive full-wave electromagnetic (EM) simulations. Notwithstanding, the curse of dimensionality considerably affects conventional metamodeling methods, and their capability to efficiently handle nonlinear antenna characteristics over broad ranges of the system parameters is limited. Performance-driven (or constrained) modeling frameworks may be employed to mitigate these issues by considering a construction of surrogates from the standpoint of the antenna performance figures rather than directly geometry parameters. This permits a significant reduction of the model setup cost without restricting its design utility. This paper proposes a novel modeling framework, which capitalizes on the domain confinement concepts and also incorporates variable-fidelity EM simulations, both at the surrogate domain definition stage, and when rendering the final surrogate. The latter employs co-kriging as a method of blending simulation data of different fidelities. The presented approach has been validated using three microstrip antennas, and demonstrated to yield reliable models at remarkably low CPU costs, as compared to both conventional and performance-driven modeling procedures.

show abstract