NEURAL NETWORK ASSESSMENT OF ROCKBURST RISKS FOR DEEP GOLD MINES IN SOUTH AFRICA
(1.College of Resources and Civil Engineering,Northeastern University, Shenyang110006, P. R. China;
2.CSIR Mining Technology, Johannesburg, South Africa;
3.Department of Mining Engineering,University of the Witwatersrand, Johannesburg, South Africa)
2.CSIR Mining Technology, Johannesburg, South Africa;
3.Department of Mining Engineering,University of the Witwatersrand, Johannesburg, South Africa)
Abstract: A neural network modeling to assess rockburst risks for deep gold mines in South Africa has been described. About 200 cases of rockbursts from a database were used to train the neural network. The results from the test cases of VCR and Carbon Leader mining, for both stopes and tunnels, were presented. It was shown that, although it has the potential to assess rockburst risks, the proposed empirical approach is still highly dependent on the accuracy of the case records collected and the way the database is structured. Within the confines of the database used, various quantitative and qualitative features affecting rockbursts were identified and their integration of an expert system and neural networks was proposed.
Key words: neural network rockburst risk South Africa