Application of hybrid artificial intelligence model to predict coal strength alteration during CO2 geological sequestration in coal seams

CO2 geological sequestration in coal seams has gradually become one of the effective means to deal with the global greenhouse effect. However, the injection of CO2 into the coal seam can have an important impact on the physical and chemical properties of coal, which in turn affects the CO2 sequestration performance in coal seams and causes a large number of environmental problems. In order to better evaluate the strength alteration of coal in CO2 geological sequestration, a hybrid artificial intelligence model integrating back propagation neural network (BPNN), genetic algorithm (GA) and adaptive boosting algorithm (AdaBoost) is proposed. A total of 112 data samples for unconfined compressive strength (UCS) are retrieved from the reported studies to train and verify the proposed model. The input variables for the predictive model include coal rank, CO2 interaction time, CO2 interaction temperature and CO2 saturation pressure, and the corresponding output variable is the measured UCS. The predictive model performance is evaluated by correlation coefficient (R), root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE). The predictive results denote that the GA-BPNN-AdaBoost predictive model is an efficient and accurate method to predict coal strength alteration induced by CO2 adsorption. The simultaneous optimization of BPNN by GA and AdaBoost algorithm can greatly improve the prediction accuracy and generalization ability of the model. At the same time, the mean impact value (MIV) is used to investigate the relative importance of each input variable. The relative importance scores of coal rank, CO2 interaction time, CO2 interaction temperature and CO2 saturation pressure are 0.5475, 0.2822, 0.0373, 0.1330, respectively. The research results in this paper can provide important guiding significance for CO2 geological sequestration in coal seams. (C) 2019 Elsevier B.V. All rights reserved.