Abstract:Optimization constrained by high-fidelity computational models has potential for transformative impact. However, such optimization is frequently unattainable in practice due to the complexity and computational intensity of the model. An alternative is to optimize a low-fidelity model and use limited evaluations of the high-fidelity model to assess the quality of the solution. This article develops a framework to use limited high-fidelity simulations to update the optimization solution computed using the low-fidelity model. Building off a previous article [22], which introduced hyper-differential sensitivity analysis with respect to model discrepancy, this article provides novel extensions of the algorithm to enable uncertainty quantification of the optimal solution update via a Bayesian framework. Specifically, we formulate a Bayesian inverse problem to estimate the model discrepancy and propagate the posterior model discrepancy distribution through the post-optimality sensitivity operator for the low-fidelity optimization problem. We provide a rigorous treatment of the Bayesian formulation, a computationally efficient algorithm to compute posterior samples, a guide to specify and interpret the algorithm hyper-parameters, and a demonstration of the approach on three examples which highlight various types of discrepancy between low and high-fidelity models.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: in the case where the high - fidelity model is computationally complex and time - consuming, how to use the limited high - fidelity simulation data to update the optimized solution obtained from the low - fidelity model and perform uncertainty quantification on the updated solution. Specifically, the paper proposes a framework to update the optimized solution calculated based on the low - fidelity model using limited high - fidelity simulations. By introducing hyper - differential sensitivity analysis (HDSA) and combining it with the Bayesian framework, the paper addresses the following issues: 1. **Infeasibility of the optimization problem**: Due to the complexity and computational intensity of the high - fidelity model, it is usually unrealistic to directly optimize it. 2. **Impact of model differences**: Although the low - fidelity model is convenient for optimization, its solution is sub - optimal due to modeling errors. 3. **Uncertainty quantification**: Previous HDSA methods can only provide updates to the average optimal solution and cannot quantify the uncertainty of the solution. To solve these problems, the main contributions of the paper include: - Proposing a new prior and noise model formulation to address the deficiencies in previous formulations. - Developing mathematical tools to make posterior sampling more efficient. - Introducing a new formula for projecting the optimal solution update, which solves theoretical problems and improves computational efficiency. - Providing explanations and guidance on hyper - parameters (parameters that define the prior distribution) to help users set these parameters correctly. In summary, the paper aims to provide a practical method to improve the optimized solution and quantify its uncertainty by combining the ease of use of the low - fidelity model and the accuracy of the high - fidelity model. This is of great significance in multi - fidelity optimization, especially when the computational cost of the high - fidelity model is extremely high.

Hyper-differential sensitivity analysis with respect to model discrepancy: Posterior Optimal Solution Sampling

Hyper-differential sensitivity analysis with respect to model discrepancy: Optimal solution updating

Hyper-differential sensitivity analysis with respect to model discrepancy: Mathematics and computation

Enabling hyper-differential sensitivity analysis for ill-posed inverse problems

Hyper-Differential Sensitivity Analysis of Uncertain Parameters in PDE-Constrained Optimization

Variance-based sensitivity of Bayesian inverse problems to the prior distribution

Hyper-Differential Sensitivity Analysis for Inverse Problems Constrained by Partial Differential Equations

Surrogate-Based Inverse Modeling of the Hydrological System: An Adaptive Approach Considering Surrogate Structural Error.

Sensitivity Analysis for Computationally Expensive Models using Optimization and Objective-oriented Surrogate Approximations

Emulation methods and adaptive sampling increase the efficiency of sensitivity analysis for computationally expensive models

An Uncertainty Quantification Method for Inexact Simulation Models

Optimal design of large-scale Bayesian linear inverse problems under reducible model uncertainty: good to know what you don't know

Computationally Efficient Sampling Methods for Sparsity Promoting Hierarchical Bayesian Models

Model Uncertainty in Evolutionary Optimization and Bayesian Optimization: A Comparative Analysis

Discrepancy Measures for Global Sensitivity Analysis

A Fully Bayesian Approach to Assessment of Model Adequacy in Inverse Problems

Uses of Sub-sample Estimates to Reduce Errors in Stochastic Optimization Models

Generalized Bayes Approach to Inverse Problems with Model Misspecification

On Uncertainty Quantification for Near-Bayes Optimal Algorithms

Parametric/Stochastic Model Reduction: Low-Rank Representation, Non-Intrusive Bi-Fidelity Approximation, and Convergence Analysis

Sensitivity analysis using the Metamodel of Optimal Prognosis