Implementation Code for Paper "Application of kernel POD pre-image recovery methods to nonintrusive nonlinear reduced order model of steady-state velocity magnitude field in the RPV lower plenum"

Published: 11 June 2025| Version 1 | DOI: 10.17632/2nvftwv9ts.1
Contributor:
he shang

Description

This dataset contains the implementation code for constructing non-intrusive nonlinear reduced order model (ROM) using kernel proper orthogonal decomposition (POD) combined with three different pre-image recovery methods proposed by Mika, Rathi, and Kwok. The code corresponds to the methodology proposed in paper and can be used to reproduce the kernel POD-based nonlinear ROM results.

Files

Steps to reproduce

By running the main program located in the codes folder, you can obtain the result data files named U_91_KPCA.mat, U_95_KPCA.mat, and U_99_KPCA.mat. Then, run the program TH_validation.m to verify the performance of the reduced-order model on the test dataset. The program has been tested and can be run using either MATLAB 2021 or Octave.

Institutions

  • Harbin Engineering University

Categories

Nuclear Engineering, Computational Fluid Dynamics

Licence