Prediction of roof subsidence in metal mine goaf based on BP neural network
Metal mine is quite different from coal mine in aspects of mineralization mechanism and ore bodys stratigraphic configuration, etc. The present mature law of overlying strata movement and ground surface subsidence mechanism in coal mine are not adaptable to metal mine. The roof subsidence of goaf in metal mine is the primary factor which leads to overlying strata movement and ground surface deformation. This paper focused on the study of this issue based on BP neural network with its characteristic that it could approach to any non-linear mapping. Firstly, the principle and features of BP neural network were briefly introduced. Then time series prediction model was established, which was based on the measured data of roof subsidence in the goaf of metal mine, and the tested sample data were trained and tested by many times. Finally the predicted value of the BP neural network was compared with the measured value, which showed that the prediction model achieved good accuracy and high precision, and could be adopted in the engineering application. The model could be used for the trend prediction of future development on roof subsidence of goaf in metal mines.
K. Zhao J.A. Wang S.N. Chen
School of Resource and Environmental Engineering, Jiangxi University of Science and Technology. Ganz School of Civil and Environment Engineering, University of Science and Technology Beijing, Beijing, School of Resource and Environmental Engineering, Jiangxi University of Science and Technology, Ganz
国际会议
The 2nd ISRM International Young ScholarsSymposium on Rock Mechanics(第二届国际岩石力学青年学者研讨会)
北京
英文
557-560
2011-10-14（万方平台首次上网日期，不代表论文的发表时间）