Adaptive machine learning based surrogate modeling to accelerate PDE-constrained optimization in enhanced oil recovery

TL;DR

This paper presents an adaptive neural network surrogate modeling framework that accelerates PDE-constrained optimization in enhanced oil recovery, focusing on polymer flooding, by adaptively training on the optimization path and using true evaluations for certification.

Contribution

It introduces an adaptive training procedure for neural network surrogates tailored to PDE-constrained optimization in oil recovery, improving efficiency and accuracy.

Findings

Achieves high accuracy in heterogeneous five-spot benchmark

Reduces computational cost of PDE evaluations

Provides certified optimization results

Abstract

In this contribution, we develop an efficient surrogate modeling framework for simulation-based optimization of enhanced oil recovery, where we particularly focus on polymer flooding. The computational approach is based on an adaptive training procedure of a neural network that directly approximates an input-output map of the underlying PDE-constrained optimization problem. The training process thereby focuses on the construction of an accurate surrogate model solely related to the optimization path of an outer iterative optimization loop. True evaluations of the objective function are used to finally obtain certified results. Numerical experiments are given to evaluate the accuracy and efficiency of the approach for a heterogeneous five-spot benchmark problem.