A robust gradient for long wavelength FWI updates

Proportionate gradient updates with PercentDelta

Deep Neural Networks are generally trained using iterative gradient updates. Magnitudes of gradients are affected by many factors, including choice of activation functions and initialization. More importantly, gradient magnitudes can greatly differ across layers, with some layers receiving much smaller gradients than others. causing some layers to train slower than others and therefore slowing ...

A Robust FWI Method for Model Updating in High Contrast Bodies

Policy-Gradient for Robust Planning

Real-world Decision-Theoretic Planning (DTP) is a very challenging research field. A common approach is to model such problems as Markov Decision Problems (MDP) and use dynamic programming techniques. Yet, two major difficulties arise: 1dynamic programming does not scale with the number of tasks, and 2the probabilistic model may be uncertain, leading to the choice of unsafe policies. We build h...

Regularizing velocity differences in time-lapse FWI using gradient mismatch information

We present a method for recovering time-lapse velocity changes using full waveform inversion (FWI). In a preprocessing step we invert for a single intermediate model by simultaneously minimizing the data misfit in the baseline and the monitor surveys. We record the individual FWI gradients corresponding to the baseline and the monitor datasets at each iteration of the inversion. Regions where t...

Variational Constitutive Updates for Strain Gradient Isotropic Plasticity

In the past decades, various strain gradient isotropic plasticity theories have been developed to describe the size-dependence plastic deformation mechanisms observed experimentally in micron-indentation, torsion, bending and thin-film bulge tests in metallic materials. Strain gradient plasticity theories also constitute a convenient device to introduce ellipticity in the differential equations...