Research on Prediction of Mudstone Breakthrough Pressure Based on Support Vector Machine in CO₂ Geological Storage

This study aims to use the Support Vector Machine (SVM) model to predict the breakthrough pressure of mudstone. By collecting data on porosity, permeability, specific surface area, and maximum throat radius from 55 sets of mudstone samples, using them as input factors and breakthrough pressure of mudstone as output factors, an SVM model was constructed and trained. The research results show that the established SVM model has high prediction accuracy and good generalization ability, and can accurately predict the breakthrough pressure of mudstone. Grid search and analysis of the penalty parameter C and kernel parameter γ in the SVM model revealed the existence of specific optimal parameter combinations that can improve model performance. This study provides an effective method for predicting the breakthrough pressure of mudstone, and also provides a scientific basis for a deeper understanding of mudstone permeability and its application in CO 2 geological storage.