Iterative Method for Tuning Complex Simulation Code

TL;DR

This paper introduces an iterative algorithm for tuning complex simulation codes by repeatedly updating parameters and Gaussian process metamodels, leading to improved calibration over existing methods.

Contribution

The paper proposes an iterative tuning algorithm that updates parameters and metamodels alternately, enhancing calibration accuracy compared to previous approaches.

Findings

The iterative method outperforms the ANLS method in toy-model tests.

The approach effectively reduces bias and improves fit in simulation calibration.

Application to nuclear fusion code demonstrates practical utility.

Abstract

Tuning a complex simulation code refers to the process of improving the agreement of a code calculation with respect to a set of experimental data by adjusting parameters implemented in the code. This process belongs to the class of inverse problems or model calibration. For this problem, the approximated nonlinear least squares (ANLS) method based on a Gaussian process (GP) metamodel has been employed by some researchers. A potential drawback of the ANLS method is that the metamodel is built only once and not updated thereafter. To address this difficulty, we propose an iterative algorithm in this study. In the proposed algorithm, the parameters of the simulation code and GP metamodel are alternatively re-estimated and updated by maximum likelihood estimation and the ANLS method. This algorithm uses both computer and experimental data repeatedly until convergence. A study using toy-models including inexact computer code with bias terms reveals that the proposed algorithm performs better than the ANLS method and the conditional-likelihood-based approach. Finally, an application to a nuclear fusion simulation code is illustrated.