... using the recurrent residual-U-Net surrogate model. Two cases, involving different amounts of historical data, are considered. We show that both MADS-RML and ES-MDA provide history matching results in general agreement with those from RS. MADS-RML is more accurate, however, and ES-MDA can display...

... Abstract History matching, a highly non-unique inverse problem, is critical to calibrate model parameters in many scientific applications. A typical approach to history matching is to start with a uniform sampling of the high-dimensional parameter space and employ a surrogate modeling based...

... powerful supercomputer facilities available today. Consequently, there has been an increasing attempt to develop physics-based surrogate models that could eventually replace the job of high-fidelity simulations in an accurate and consistent way within optimization and uncertainty quantification workflows...

... simulation Artificial Intelligence machine learning base point surrogate model parameter point history matching prediction gmm distribution uncertainty quantification Upstream Oil & Gas proposal sample point algorithm Bayesian Inference gp surrogate model parameter production history...

... derived from data analytics, Depending on the degree of physics involved in surrogate modeling workflows, non-intrusive surrogate models have wider applicability to deal with different data and physical models than intrusive model approaches, and, As timely sources of data are becoming more...

... Compared with intrusive ROMs, non-intrusive ROMs have the advantage that no knowledge of the governing equations or under- lying physics is required. These ROMs build a surrogate model to approximate the input-output response based on a set of preliminary simulation results known as samplings...