4th July 2019
Deep Stochastic Inversion
With Lukas Mosser, Imperial College London
Numerous modelling tasks require the solution of ill-posed inverse problems where we seek to find a distribution of earth models that match observed data such as reflected acoustic waveforms or produced hydrocarbon volumes. We present a deep learning framework to create stochastic samples of posterior property distributions for ill-posed inverse problems using a gradient-based approach. The spatial distribution of petrophysical properties is created by a deep generative model and controlled by a set of latent variables. A generative adversarial network (GAN) is used to represent a prior distribution of geological models based on a training set of object-based models. Then we minimize the mismatch between observed ground-truth data and numerical forward-models of the generator output by first computing gradients of the objective function with respect to grid-block properties and then using neural network backpropagation to obtain gradients with respect to the latent variables. Synthetic test cases of acoustic waveform inversion and reservoir history matching are presented. For both synthetic cases, we show that deep generative models such as GANs can be combined in an end-to-end framework to obtain stochastic solutions to geophysical inverse problems.
Lukas Mosser is a 3rd year PhD Student at Imperial College London under supervision of Prof. Olivier Dubrule and Prof. Martin J. Blunt. He holds an MSc. in Petroleum Engineering from Imperial College London and has earned the BP Prize for best MSc. Thesis as well as the John S Archer Award for PhD Research Excellence from Imperial College London.
PESGB - Training Room