Artificail neural networks for predicting rockburst in deep mining

Abstract：Rockburst is a kind of dynamic instability phenomenon for surrounding rock mass in deep mining, which has complicated nonlinear relationship between rockburst and its factors.Based on the analysis of various factors of rockburst, select the mining depth H, ratio of rock's maximal tangential stress and rock's uniaxle compressive strength, ratio of rock's uniaxle compressive strength and rock's uniaxle tensile strength, and elastic energy index as judging indexes of rockburst.A rockburst prediction model is proposed with the use of artificial neural network.Through analysing and computing real deep mining and rock underground project examples at home and abroad, the results show that it is feasible and appropriate to select mining depth H as a main factor, the model is valid of predicting rockburst in deep mining.

Keyword：deep mining;rockburst;nonlinear;mining depth;artificial neural network;