Myong Nam Sin - Faculty of Mining Engineering, Kim Chaek University of Technology, Pyongyang, Democratic People’s Republic of Korea
Un Chol Han - School of Science and Engineering, Kim Chaek University of Technology, Pyongyang, Democratic People’s Republic of Korea
Hyon Hyok Ri - Faculty of Mining Engineering, Kim Chaek University of Technology, Pyongyang, Democratic People’s Republic of Korea
Sung Il Jon - Department of Applied Mathematics, Kim Chaek University of Technology, Pyongyang, Democratic People’s Republic of Korea

Anthracite coal seam of Democratic People’s Republic of Korea was broken into particles to be soft due to geological tectonic actions through several stages in the Mesozoic era. Because the folds and faults have excessively developed and the shape of coal seam is very complicated, it is impossible to extract the anthracite coal by longwall mining system, and coal has been mainly mined by entry caving mining system. The aim of this work is to assess effectiveness of new combination of flying squirrel search algorithm (SSA) and artificial neural-network (ANN) for back-analysis of time-depending mechanical parameters of anthracite coal based on timber loads and displacements measured in the coal face entry. The case study deals with a coal face entry in Sinchang Coal Mine located in the Unsan County, South Pyongan Province, DPR Korea. To verify the good performance of new combination of the SSA and ANN, the comparison studies between proposed back-analysis method and other methods with the same purpose, are conducted using data measured in coal face entry. The mean absolute error (MAE) of weighted error norm of ANN-SSA is relatively smaller in comparison with other methods, which is ۲.۴۹. The new back-analysis is the good method to determine the suitable time-dependent mechanical parameters of anthracite coal surrounding the entry in very soft coal seam.

SSA, displacement back analysis, Neural network, time-dependent mechanical parameter