A Kriging and Stochastic Collocation ensemble for uncertainty quantification in engineering applications

Kaintura, Arun; Spina, Domenico; Couckuyt, Ivo; Knockaert, Luc; Bogaerts, Wim; Dhaene, Tom

doi:10.1007/s00366-017-0507-0

Cited by 15 publications

(6 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Advanced optimization methods such as Kriging and Stochastic Collocation , developed for mechanical or radio‐frequency (RF) design are now being introduced into photonic device design . These techniques provide a rigorous framework to treat performance variability, but also help to reduce the number of expensive simulations needed for an optimization.…”

Section: Challenges For An Integrated Photonic Design Flowmentioning

confidence: 99%

Silicon Photonics Circuit Design: Methods, Tools and Challenges

Bogaerts

Chrostowski

2018

Laser & Photonics Reviews

437

230

View full text Add to dashboard Cite

Silicon Photonics technology is rapidly maturing as a platform for larger‐scale photonic circuits. As a result, the associated design methodologies are also evolving from component‐oriented design to a more circuit‐oriented design flow, that makes abstraction from the very detailed geometry and enables design on a larger scale. In this paper, the state of this emerging photonic circuit design flow and its synergies with electronic design automation (EDA) are reviewed. The design flow from schematic capture, circuit simulation, layout and verification is covered. The similarities and the differences between photonic and electronic design, and the challenges and opportunities that present themselves in the new photonic design landscape, such as variability analysis, photonic‐electronic co‐simulation and compact model definition are discussed.

show abstract

Section: Challenges For An Integrated Photonic Design Flowmentioning

confidence: 99%

Silicon Photonics Circuit Design: Methods, Tools and Challenges

Bogaerts

Chrostowski

2018

Laser & Photonics Reviews

437

230

View full text Add to dashboard Cite

show abstract

“…There are many regression models that can be used to cheaply approximate the simulator, such as Bayesian neural networks, support vector regression, and random forest regression. Nevertheless, under the category of data-efficient machine learning, Gaussian Process (GP), also known as Kriging [34,35], is arguably the most prevalent probabilistic technique. GPs have been widely used as a regression method in many different areas including electronics engineering [36], computational fluid dynamics [37] and chemistry [38].…”

Section: Gaussian Process Regressionmentioning

confidence: 99%

Adaptive sampling with automatic stopping for feasible region identification in engineering design

Qing

Knudde

Garbuglia

et al. 2021

Engineering with Computers

View full text Add to dashboard Cite

Engineering design is a complex process to find a suitable trade-off among different, and sometimes conflicting, design specifications. In reality, these requirements can be often considered as constraints of the design problem, that can be defined in terms of performance measures or geometrical characteristics of the device under study. In this paper, a new design space exploration methodology is presented for discovering feasible regions in the design space, where the term feasible region indicates the set of all design configurations satisfying all constraints of the design problem. The proposed method is based on Gaussian Process metamodels to estimate the feasible region and leverages a information-based adaptive sampling technique to sequentially refine the prediction accuracy, which is applicable for multiple constraints problems. In order to efficiently stop the adaptive sampling process, a novel framework to estimate the metamodel's prediction accuracy is proposed. The efficiency, accuracy and robustness of the proposed approach are compared with state-of-art techniques on suitable benchmark problems and practical engineering examples.

show abstract

“…In this regard, an effective alternative is the class of nonparametric machine learning methods, for which the model complexity is not related to the problem dimensionality, but rather to the number of available training data 19 , 23 – 25 . One example is Gaussian process regression (GPR) 26 , also known as Kriging 27 . An advantage of nonparametric models is that they are purely data-driven, and therefore they can adapt better to the analysis of complex devices compared to other methods, like polynomial chaos, that assume a predefined model form.…”

Section: Introductionmentioning

confidence: 99%

Stochastic and multi-objective design of photonic devices with machine learning

Manfredi,

Waqas,

Melati

2024

Sci Rep

View full text Add to dashboard Cite

Compact and highly performing photonic devices are characterized by non-intuitive geometries, a large number of parameters, and multiple figures of merit. Optimization and machine learning techniques have been explored to handle these complex designs, but the existing approaches often overlook stochastic quantities. As an example, random fabrication uncertainties critically determines experimental device performance. Here, we present a novel approach for the stochastic multi-objective design of photonic devices combining unsupervised dimensionality reduction and Gaussian process regression. The proposed approach allows to efficiently identify promising alternative designs and model the statistic of their response. Incorporating both deterministic and stochastic quantities into the design process enables a comprehensive analysis of the device and of the possible trade-offs between different performance metrics. As a proof-of-concept, we investigate surface gratings for fiber coupling in a silicon-on-insulator platform, considering variability in structure sizes, silicon thickness, and multi-step etch alignment. We analyze 86 alternative designs presenting comparable performance when neglecting variability, discovering on the contrary marked differences in yield and worst-case figures for both fiber coupling efficiency and back-reflections. Pareto frontiers demonstrating optimized device robustness are identified as well, offering a powerful tool for the design and optimization of photonic devices with stochastic figures of merit.

show abstract

A Kriging and Stochastic Collocation ensemble for uncertainty quantification in engineering applications

Cited by 15 publications

References 51 publications

Silicon Photonics Circuit Design: Methods, Tools and Challenges

Silicon Photonics Circuit Design: Methods, Tools and Challenges

Adaptive sampling with automatic stopping for feasible region identification in engineering design

Stochastic and multi-objective design of photonic devices with machine learning

Contact Info

Product

Resources

About