Database Generation for Deep Learning Inversion of 2.5D Borehole Electromagnetic Measurements using Refined Isogeometric Analysis

TL;DR

This paper presents a method to generate large synthetic datasets for deep learning inversion of borehole electromagnetic measurements using refined isogeometric analysis, enabling rapid and accurate simulations of complex earth models.

Contribution

It introduces the use of refined isogeometric analysis for efficient 2.5D simulation and database generation in geophysical inversion tasks.

Findings

Generated 100,000 earth model measurements in 56 hours

Demonstrated rapid and accurate database creation for deep learning

Enabled large-scale synthetic data generation for geophysical inversion

Abstract

Borehole resistivity measurements are routinely inverted in real-time during geosteering operations. The inversion process can be efficiently performed with the help of advanced artificial intelligence algorithms such as deep learning. These methods require a large dataset that relates multiple earth models with the corresponding borehole resistivity measurements. In here, we propose to use an advanced numerical method --refined isogeometric analysis (rIGA)-- to perform rapid and accurate 2.5D simulations and generate databases when considering arbitrary 2D earth models. Numerical results show that we can generate a meaningful synthetic database composed of 100,000 earth models with the corresponding measurements in 56 hours using a workstation equipped with two CPUs.