Carbonate mound–shoal complexes, despite their complex pore structures and pronounced heterogeneity, represent one of the most productive reservoir units within carbonate formations. Accurately predicting key physical properties—such as porosity, permeability, and flow zone index—from well log data remains a significant challenge for conventional empirical methods. This study investigates the application of machine learning algorithms for optimizing the prediction of reservoir properties in hill-and-plain carbonate bodies. Six machine learning approaches—Support Vector Machines (SVM), Backpropagation Neural Networks (BPNN), Long Short-Term Memory Networks (LSTM), K-Nearest Neighbors (KNN), Random Forests (RF), and Gaussian Process Regression (GPR)—are systematically evaluated and compared. The analysis employed flow zone indices, geological data, and well log curves to classify porosity–permeability types. Seven logging parameters were used as input features: spectral gamma ray (SGR), uranium-free gamma ray (CGR), photoelectric absorption cross-section index (PE), bulk density (RHOB), acoustic travel time (DT), neutron porosity (NPHI), and true resistivity (RT). These features were paired with measured physical property values to train and validate the predictive models. Results demonstrate distinct algorithmic advantages for specific properties. The RF model achieved superior performance in permeability prediction, yielding an R² of 0.6824, whereas the GPR model provided the highest accuracy for porosity estimation, with an R² of 0.7342 and an Accuracy Index (ACI) of 0.9699. Despite these improvements, machine learning models still face limitations in accurately characterizing low-permeability zones within highly heterogeneous hill–terrace reservoirs. To address this challenge, the study integrates geological prior knowledge into the machine learning framework and applies cross-validation techniques to optimize model parameters, thereby providing a practical and robust approach for detailed assessment of mound–hoal carbonate reservoirs.

Formation pore pressure is a critical parameter controlling drilling safety and wellbore stability, and its prediction in ultra-deep carbonate formations is challenging due to extreme temperature–pressure conditions, complex geological settings, and strong lithological heterogeneity. This study develops a multi-parameter while-drilling process that integrates drilling engineering parameters, mud logging gas measurements, cuttings-based elemental logging data, and drilling fluid performance indicators to investigate the processes governing abnormal overpressure and its real-time responses. By combining elemental logging parameters, rock drillability indices, and gas logging responses, a predictive framework for detecting abnormal pore pressure is established. The results show that structural position strongly controls overpressure distribution; secondary fault zones preferentially host abnormal overpressure; synchronous enrichment of S and Sr in cuttings-derived elemental data provides precursor signals; and gas logging indicators, including total hydrocarbon peaks and hydrocarbon migration velocity, are highly sensitive to overpressure. Application of the proposed while-drilling process to three ultra-deep wells (Fudong-101, Hade-18, and TKe-1) generated nine real-time pressure warnings, eight of which were confirmed, yielding a prediction accuracy of 88.89%. These results demonstrate that the proposed process effectively improves real-time identification of abnormal overpressure in ultra-deep carbonate formations, enhancing drilling safety and operational efficiency.