Abstract:The structure of coal indicates the degree of its fragmentation after tectonic movement, which affects the exploration and development of coalbed methane (CBM). Although coal core observations are the most convenient and intuitive way of identifying the coal structure, they are not applicable for use in unexplored coal seams without CBM wells, and they are also very time-consuming. In comparison, geophysical-logging interpretation of the coal structure is more efficient and economical. However, although qualitative methods, such as principal component analysis (PCA), can be used to identify the coal structure with geophysical logging, the interpretation is limited by the calculation ability, and improvements are required based on the structure of an empirical model. Multinomial logistic regression (MLR), random forest (RF), and deep fully connected neural network (DNN) are effective machine learning methods more accurate than the traditional method that with model-aided identification. In this respect, the MLR method is a classical method based on mathematical linear regression, and it has a low construction cost; RF is an ensemble learning algorithm based on a decision tree and use of a bagging algorithm; and DNN is a deep learning model based on self-built feature engineering that has high classification accuracy under a large amount of data training and provides obvious advantages in visual coal classification problems. In this work, the three machine learning methods, MLR, RF, and DNN, were used to identify the coal structure. Two sets of logging data comprising different quantities from the Anze Block of the southern Qinshui Basin, North China, were selected to quantitatively compare the accuracy of coal structure identification with partial coal core observation. The results showed that for 210 and 840 samples, respectively, the accuracy was 76% and 77% for MLR, 83% and 86% for RF, and 82% and 86% for DNN. These results show that the MLR and DNN methods are superior for use with minimal and maximum amounts of data, respectively, and the RF method provides overall accuracy. Furthermore, an algorithmic classification of the coal structure was established, and the geological factors controlling the predicted structure, such as geostress, coal seam thickness, and burial depth, were distinguished.

Interpreting coal component content in logging data by combining grey relational analysis and hybrid neural network

Intelligent classification of coal structure using multinomial logistic regression, random forest and fully connected neural network with multisource geophysical logging data

Mining Technology Evaluation for Steep Coal Seams Based on a GA-BP Neural Network

Selection of machine learning algorithms in coalbed methane content predictions

Maximum burning rate and fixed carbon burnout efficiency of power coal blends predicted with back-propagation neural network models

Maximum solid concentrations of coal water slurries predicted by neural network models

An ANN-based method for predicting Zhundong and other Chinese coal slagging potential

Research on the Method of Methane Emission Prediction Using Improved Grey Radial Basis Function Neural Network Model

Prediction of Coal Slurry Concentration Based on Artificial Neural Networks

Intelligent Proximate Analysis of Coal Based on Near Infrared Spectroscopy and Multi-output Deep Learning

Research on Intelligent Comprehensive Evaluation of Coal Seam Impact Risk Based on BP Neural Network Model

Main Controlling Factors Analysis and Prediction of Coalbed Gas Content

A Physics-Constrained Data-Driven Workflow for Predicting Coalbed Methane Well Production Using A Combined Gated Recurrent Unit and Multi-Layer Perception Neural Network Model

The Investigation of Carbon Content in Fly Ash for a BP Neural Network-based Pulverized Coal-fired Boiler

MSA-Net: A Precise and Robust Model for Predicting the Carbon Content on an As-Received Basis of Coal

Intelligent Prediction of Coal Mine Gas Emission Based on Deep Neural Network

Coal Structure Recognition Method Based on LSTM Neural Network

Coal Mine Safety Evaluation Based on Machine Learning: A BP Neural Network Model

Product quality prediction in dense medium coal preparation process based on recurrent neural network

Real-time in situ monitoring of ash content in coal via dual-energy X-rays and an MLP neural network

Optimized Random Forest Method for 3D Evaluation of Coalbed Methane Content Using Geophysical Logging Data