Study on prediction of development height of water-conducting fracture zone in overlying strata under fully-mechanized caving in extremely thick coal seam
HAN Liang
WANG Shuaiqi
LIU Bo
ZHENG Yuchun
LIU Quansheng
GAO Heng
CAO Kuo
Abstract：To accurately predict the development height of the water-conducting fracture zone in super-thick coal seam of Zhaoxian Coal Mine,this paper controls the variation of single influencing factor parameters through numerical simulation,and quantitatively analyzes the correlation between the thickness of coal seam mining,the length of working face,the buried depth of coal seam and the development height of water-conduc-ting fracture zone.Combined with the existing research results,the influence of mining method and overburden type on the development height of water flowing fractured zone is qualitatively expounded.By systematically organizing field observation data from multiple mines in Huanglong Jurassic coalfield,two prediction models for the development height of the water-conducting fracture zone,suitable for the thick coal seam of Zhaoxian coal mine,are constructed:a multivariate nonlinear fitting model and a BP neural network prediction model.Based on the above analysis results of influencing factors and measured data,the fitting training and verification of the observation data of water-conducting fracture zones in multiple mines in Huanglong coalfield are conducted.The average relative error of the multivariate nonlinear fitting formula is 7.43%,and the overall correlation co-efficient R of the BP neural network prediction model is 0.96.The mining parameters of the 1303 and 2306 working faces in the Zhaoxian Coal Mine were selected for prediction and application.The results indicate that both prediction methods demonstrate high accuracy;specifically,the average relative error between the predic-tions generated by the BP neural network model and the measured site values is smaller,indicating superior prediction accuracy.The two prediction methods proposed in this paper are user-friendly and exhibit strong generalization capabilities,providing a reliable reference for the prevention and control of roof separation water disasters in fully mechanized caving mines within the Jurassic coalfield of Huanglong and similar regions.
Keywords：extremely thick coal seamprediction of water-conducting fractured zonenumerical simulationmultiple nonlinear regressionBP neural network
Publication Date：2025-12-30
Online Publishing Date：2025-12-19(First online date of this platform, not the publication date of the document)
Pages：9( 1-9 )