Intelligent Prediction and Modeling Method for Rock Drillability Based on Support Vector Machine
Chengtao Lv, Meng Cui, Yan Ding, Ziyun Zhao, Shuo Li, Fei Zhao, Reyu Gao, Ge Wang · 2025
The data-driven intelligent modeling method for the drillability of formation rocks is an approach that utilizes machine learning and artificial intelligence technologies to predict and evaluate the drillability of formation rocks by establishing a relationship model between the drillability of formation rocks and various logging parameters. This paper proposes a data-driven intelligent modeling method for the drillability of formation rocks. It uses the Extreme Learning Machine (ELM) algorithm to establish a one-dimensional model for predicting the rock drillability in the Yu well block. A software named Drillability 1.0, developed in Matlab, was used to predict the rock drillability for five wells, establishing a high-precision model of the relationship between the drillability grade of the rocks and some important influencing factors. This improves the accuracy and efficiency of predicting the rock drillability grade and lays a data foundation for the subsequent establishment of a three-dimensional model for predicting the drillability of formation rocks.