|
Author(s): |
Peixian Li, China University of Mining and Technology, Key Laboratory for Land Environment and Disaster Monitoring of SBSM, Xuzhou, China/China University of Mining and Technology, Jiangsu Key Laboratory of Resources and Environmental Information Engineering, Xuzhou,
Zhixiang Tan, China University of Mining and Technology, Key Laboratory for Land Environment and Disaster Monitoring of SBSM, Xuzhou, China/China University of Mining and Technology, Jiangsu Key Laboratory of Resources and Environmental Information Engineering, Xuzhou,
Lili Yan, China University of Mining and Technology, Key Laboratory for Land Environment and Disaster Monitoring of SBSM, Xuzhou, China/China University of Mining and Technology, Jiangsu Key Laboratory of Resources and Environmental Information Engineering, Xuzhou,
Kazhong Deng, China University of Mining and Technology, Key Laboratory for Land Environment and Disaster Monitoring of SBSM, Xuzhou, China/China University of Mining and Technology, Jiangsu Key Laboratory of Resources and Environmental Information Engineering, Xuzhou, |
|
Abstract: |
In order to obtain probability integration method parameters of surface movement after coal mining, based on analysis of mining and geological conditions, BP neural network model was built to back analysis the parameters with mining and geological conditions. Typical surface movement observation data in China were used as training and testing samples. Mean square error and mean absolute percentage error were used to evaluate the accuracy of the model. The calculated results show that model accuracy of fitting is goodness. Probability integration method parameters of 4 test samples were calculated by the inversion model, all mean square error of the results tested were less than 3 times of mean square error, and can meet the requirement of mining subsidence prediction, also show that the method to calculate probability integration method based on neural network inversion model is feasible. Various factors can be considered overall comprehensively with the BP neural network and nonlinear relationship between probability integration method parameters and mining and geological factors was established. The study provide basis to calculate mining subsidence prediction parameters for mining areas lack of actual observation data.
|