Polynomial-based surrogate models for uncertainty quantification of passive electronic systems

Shangguan, Xingjian

Permalink

https://hdl.handle.net/2142/117839 Copy

Description

Title

Polynomial-based surrogate models for uncertainty quantification of passive electronic systems

Author(s)

Shangguan, Xingjian

Issue Date

2022-12-09

Director of Research (if dissertation) or Advisor (if thesis)

Chen, Xu

Department of Study

Electrical & Computer Eng

Discipline

Electrical & Computer Engr

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

M.S.

Degree Level

Thesis

Keyword(s)

• polynomial chaos
• rational polynomial chaos
• uncertainty quantification
• surrogate model
• electronics packaging
• principal component analysis.

Abstract

In this thesis, the state of the art of polynomial-based surrogate models for uncertainty quantification is reviewed. Different aspects of the surrogate models, including order selection, choice of polynomial basis, quadrature rules and regression-based coefficient obtain process, compression rate of principle component analysis and its relation to the order truncation, as well as the sampling of the design space, are discussed and compared. A new algorithm for sampling and a common denominator rational polynomial chaos formulation are proposed. Numerical results on an analytical distributed circuit model, a differential via structure, and a fan-out package are provided to demonstrate the applicability of the polynomial-based methods to work with passive electronic systems and structures. The proposed sampling method improved the $L_{\infty}$ norm of the error compared to the simulation. And the common denominator formulation provides speed-up to the rational polynomial chaos. And finally, a conclusion and directions for future work are presented.

Graduation Semester

2022-12

Type of Resource

Thesis

Copyright and License Information

Copyright 2022 Xingjian Shangguan

Owning Collections