Review of Metamodeling Techniques for Product Design With Computation-Intensive Processes

Wang, G. Gary; Shan, Songqing

doi:10.24908/pceea.v0i0.3940

Cited by 12 publications

(8 citation statements)

References 118 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…They found the best performance corresponding to the model developed using the Latin hypercube design technique. Central composite design lacks in its efficacy when it comes to a design problem having high-dimensional data [148]. In an attempt to achieve complete uniform curing along with reducing the residual stresses in the case of a fiberreinforced composite laminate, Shah et al optimal design and central composite design performs better compared to other design of experiment (DoE) methods for composite laminates [140,154].…”

Section: Sampling Techniques For Training Machine Learning Modelsmentioning

confidence: 99%

Advances in Computational Intelligence of Polymer Composite Materials: Machine Learning Assisted Modeling, Analysis and Design

Sharma

Mukhopadhyay

Rangappa

et al. 2022

Arch Computat Methods Eng

View full text Add to dashboard Cite

The superior multi-functional properties of polymer composites have made them an ideal choice for aerospace, automobile, marine, civil, and many other technologically demanding industries. The increasing demand of these composites calls for an extensive investigation of their physical, chemical and mechanical behavior under different exposure conditions. Machine learning (ML) has been recognized as a powerful predictive tool for data-driven multi-physical modeling, leading to unprecedented insights and exploration of the system properties beyond the capability of traditional computational and experimental analyses. Here we aim to abridge the findings of the large volume of relevant literature and highlight the broad spectrum potential of ML in applications like prediction, optimization, feature identification, uncertainty quantification, reliability and sensitivity analysis along with the framework of different ML algorithms concerning polymer composites. Challenges like the curse of dimensionality, overfitting, noise and mixed variable problems are discussed, including the latest advancements in ML that have the potential to be integrated in the field of polymer composites.Based on the extensive literature survey, a few recommendations on the exploitation of various ML algorithms for addressing different critical problems concerning polymer composites are provided along with insightful perspectives on the potential directions of future research.

show abstract

Section: Sampling Techniques For Training Machine Learning Modelsmentioning

confidence: 99%

Advances in Computational Intelligence of Polymer Composite Materials: Machine Learning Assisted Modeling, Analysis and Design

Sharma

Mukhopadhyay

Rangappa

et al. 2022

Arch Computat Methods Eng

View full text Add to dashboard Cite

show abstract

“…A comparative study of various metamodeling techniques has been provided by Jin et al (2001) using multiple modeling criteria and multiple test problems. And a whole review of their potential use for computation intensive processes is given by Wang and Shan (2005).…”

Section: Surrogate Models Construction Techniquesmentioning

confidence: 99%

Surrogates based multi-criteria predesign methodology of Sodium-cooled Fast Reactor cores – Application to CFV-like cores

Fabbris

Dardour

Blaise

et al. 2016

Nuclear Engineering and Design

View full text Add to dashboard Cite

The Sodium-cooled Fast Reactor (SFR) core predesign process is commonly realized on the basis of expert advices and local parametric studies. As such, in-deep knowledge of physical phenomena avoids an important number of expensive simulations. However, the study space is explored only partially. To ease the computational burden metamodels, or surrogate models, can be used, to quickly evaluate the performances of a wide set of different cores, individually defined by a set of parameters (pellet diameter, fissile height...), in the study space. This paper presents the development of a simplified neutronics ERANOS reference core calculation scheme that is then implemented in the construction of the Design of Experiment (DOE) database. The surrogate models for SFR CFV-like cores performances are developed, biases and uncertainties are quantified against the CFV-v1 version. Global Sensitivity Analysis also allowed highlighting antagonist performances for the design and to propose two alternative core configurations. A broadened application of the method with an optimization of a CFV-like core is also detailed. The Pareto front of the seven selected performance parameters has been studied using eleven surrogate models, based on Artificial Neural Network (ANN). The optimization demonstrates that the CFV-v1, designed using Best Estimate codes, under given performance constraints, is Pareto optimal: no other configuration is highlighted from the Multi-Objective Optimization (MOO) study. Further MOO analysis, including a specific study on impact of new degrees of freedom, such as five Pu enrichments compared to two, or different pellet diameters have been performed. Additional configurations are then found by the surrogate models, improving simultaneously all performances of the CFV-v1 configuration.

show abstract

“…In addition, the framework of these interactions defines the nature of non-homogeneity of process variance and typifies the design of parameters. The application of robust design optimization has been contributed by the great number of researchers in quality improvement of a various process or product design in practice, and several appropriate studies have been reviewed the application of Taguchi methodology in real case studies, see [15], [22], [24]- [27].…”

Section: Strength and Weakness Viewpoints Over Rdmentioning

confidence: 99%

Robust Product Design: A Modern View of Quality Engineering in Manufacturing Systems

Parnianifard¹,

AHMAD²,

Ariffin³

et al. 2019

Preprint

View full text Add to dashboard Cite

One of the main technological and economic challenges for an engineer is designing high-quality products in manufacturing processes. Most of these processes involve a large number of variables included the setting of controllable (design) and uncontrollable (noise) variables. Robust Design (RD) method uses a collection of mathematical and statistical tools to study a large number of variables in the process with a minimum value of computational cost. Robust design method tries to make high-quality products according to customers’ viewpoints with an acceptable profit margin. This paper aims to provide a brief up-to-date review of the latest development of RD method particularly applied in manufacturing systems. The basic concepts of the quality loss function, orthogonal array, and crossed array design are explained. According to robust design approach, two classifications are presented, first for different types of factors, and second for different types of data. This classification plays an important role in determining the number of necessity replications for experiments and choose the best method for analyzing data. In addition, the combination of RD method with some other optimization methods applied in designing and optimizing of processes are discussed.

show abstract

Review of Metamodeling Techniques for Product Design With Computation-Intensive Processes

Cited by 12 publications

References 118 publications

Advances in Computational Intelligence of Polymer Composite Materials: Machine Learning Assisted Modeling, Analysis and Design

Advances in Computational Intelligence of Polymer Composite Materials: Machine Learning Assisted Modeling, Analysis and Design

Surrogates based multi-criteria predesign methodology of Sodium-cooled Fast Reactor cores – Application to CFV-like cores

Robust Product Design: A Modern View of Quality Engineering in Manufacturing Systems

Contact Info

Product

Resources

About