Lithology Identification of Uranium-Bearing Sand Bodies Using Logging Data Based on a BP Neural Network

• Published in: Minerals (Basel) Vol. 12; no. 5; p. 546
• Main Authors: Sun, Yuanqiang, Chen, Jianping, Yan, Pengbing, Zhong, Jun, Sun, Yuxin, Jin, Xinyu
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.05.2022
• Subjects: Accuracy; Algorithms; Back propagation networks; Boreholes; Classification; Coalbed methane; Core logging; Data logging; Efficiency; Entropy; Entropy (Information theory); Fault diagnosis; Geology; Identification; Lithology; Logging; Machine learning; Methods; Neural networks; Optimization; Parameter estimation; Predictions; Propagation; Regression analysis; Sandstone; Sedimentary rocks; Statistical analysis; Uranium; Well logging
• ISSN: 2075-163X; 2075-163X
• DOI: 10.3390/min12050546

Lithology identification is an essential fact for delineating uranium-bearing sandstone bodies. A new method is provided to delineate sandstone bodies by a lithological automatic classification model using machine learning techniques, which could also improve the efficiency of borehole core logging. In this contribution, the BP neural network model for automatic lithology identification was established using an optimized gradient descent algorithm based on the neural network training of 4578 sets of well logging data (including lithology, density, resistivity, natural gamma, well-diameter, natural potential, etc.) from 8 boreholes of the Tarangaole uranium deposit in Inner Mongolia. The softmax activation function and the cross-entropy loss function are used for lithology classification and weight adjustment. The lithology identification prediction was carried out for 599 samples, with a prediction accuracy of 88.31%. The prediction results suggest that the model is efficient and effective, and that it could be directly applied for automatic lithology identification in sandstone bodies for uranium exploration.