Physical and data combined driving oil reservoir seepage simulation method and system

• Main Authors: ZHAO XU, HAN JIANGXIA, LIU YUETIAN, XUE LIANG
• Format: Patent
• Language: Chinese; English
• Published: 11.10.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention discloses a physics and data combined drive oil reservoir seepage simulation method and system. The method comprises the steps that a physical model of a target oil reservoir area is determined; determining fluid high-pressure physical properties, reservoir attributes and a multiphase seepage mechanism of the target oil reservoir area; based on the physical model, the fluid high-pressure physical property, the reservoir attribute and the multiphase seepage mechanism, determining a multiphase seepage model of the target oil reservoir area; constructing an adaptive differentiable node neuron connection model, and determining a loss function of the adaptive differentiable node neuron connection model based on the multiphase seepage model and the well point production data; obtaining a space-time training data set of the physical model; training the self-adaptive differentiable node neuron connection model based on the well point production data and the multiphase seepage model to obtain a trained s