Estimation of heterogeneous permeability using pressure derivative data through an inversion neural network inspired by the fast marching method

Engineering

• Root Mean Square Error 100%
• Fluid Flow 66%
• Time-of-Flight 66%
• Diffusive 66%
• Numerical Experiment 66%
• Subsurface 66%
• Model Parameter 66%
• Pressure Data 66%
• Transients 66%
• Transfer Learning 66%
• Quick Response 33%
• Proof-of-Concept 33%
• Graphics Processing Unit 33%
• Parameter Estimation 33%
• Computational Efficiency 33%
• Learning Approach 33%
• Forward Model 33%
• Iterative Procedure 33%
• Flow Behavior 33%
• Flow in Porous Medium 33%
• Convolutional Neural Network 33%
• Reservoir Heterogeneity 33%
• Solve Method 33%
• Neural Network Architecture 33%
• Partial Derivative 33%
• Demonstrates 33%
• Diffusivity Equation 33%
• Sparsity 33%
• Pressure Drop 33%
• Relative Error 33%
• Deep Learning 33%
• Natural Logarithm 33%

Keyphrases

• Reservoir Inversion 9%
• Permeability Inversion 9%
• Model Heterogeneity 9%