Research progress of machine-learning algorithm for formation pore pressure prediction

Formation pore pressure is one of the most important basic data in petroleum exploration and development. The traditional prediction model of formation pore pressure relies on artificial experience and is highly subjective, so it seldom considers the influence multiple variables characteristics of formation pore pressure. In recent years, machine learning has been applied in lithology, reservoir, complex drilling conditions, formation pore pressure, and other fields. This article introduces the research progress and general process of machine learning algorithm in formation pore pressure prediction in recent years. In this study, the intelligent optimization algorithm is used to optimize the machine learning model and realize the intelligent prediction method of formation pore pressure. The results show that support vector regression (SVR) obtained the best prediction performance with the determination coefficient of 0.996. At the same time, this study reflects the importance of intelligent optimization algorithm to machine learning model optimization accuracy. This method provides a new idea for using multiple variables formation pore pressure in the future.