Hybrid method for analyzing air thermal conditions in underground mines

• Published in: Expert systems with applications Vol. 245; p. 123026
• Main Authors: Ihsan, Ahmad, Cheng, Jianwei, Widodo, Nuhindro Priagung, Wang, En-yuan, Waly, Fadli Zaka, Syachran, Satria Rum, Fadillah, Taruna, Khamidah, Halumi Nur
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.07.2024
• Subjects: ANFIS model; Mine air thermal; Numerical simulation; Prediction; Numerical simulation; Mine air thermal; ANFIS model; Prediction
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2023.123026

•A hybrid method is proposed to analyze air thermal conditions in underground mines.•Thermal conditions in underground mines are simulated numerically.•Two neuro-fuzzy models were developed to predict the Wet Bulb Globe Temperature.•An accurate predictor was obtained with coefficients of determination of 0.98 and 0.97. Thermal prediction of underground mine air is required to develop control measures against heat problems. Researchers have extensively studied this topic using numerical simulations; however, these require long processing times. Recently, researchers have begun using artificial intelligence algorithms; however, the predictive capabilities of the model are still limited because an intelligent system is formed from field measurement data. This study presents a fast and accurate prediction of the Wet Bulb Globe Temperature (WBGT) in underground mines using a hybrid integrated numerical method and an adaptive neuro fuzzy inference system (ANFIS) method. The mine air thermal conditions in various scenarios were analyzed by numerical simulation, and the results were utilized to develop intelligence for ANFIS model-based predictors. A case study was conducted in two underground gold mine areas to demonstrate the effectiveness of this method. The ANFIS model was trained and tested with 81 scenarios generated from numerical simulations. Accurate predictors were obtained, with a coefficient of determination (R2) of 0.98 and 0.97. In addition to predicting the WBGT, the developed ANFIS model optimized the selection of the auxiliary fan power, minimizing the power consumption while simultaneously providing a comfortable WBGT.