TY - JOUR
T1 - Evaluation of rockburst occurrence and intensity in underground structures using decision tree approach
AU - Ghasemi, Ebrahim
AU - Gholizadeh, Hasan
AU - Adoko, Amoussou Coffi
N1 - Publisher Copyright:
© 2019, Springer-Verlag London Ltd., part of Springer Nature.
PY - 2020/1/1
Y1 - 2020/1/1
N2 - Based on reported statistics, rockburst phenomenon is the main cause of many casualties and accidents occurred during the construction of deep underground structures. Therefore, its prediction in initial stages of design has a remarkable role on enhancement of safety. In this paper, two models have been developed for rockburst evaluation using the C5.0 decision tree classifier. The first model has been applied for prediction of rockburst occurrence and the second model for prediction of rockburst intensity. These models have been developed based on a database including 174 rockburst case histories. In both models, stress coefficient, rock brittleness coefficient, and the elastic strain energy index are the predictive variables. These models are easy to use and do not require extensive knowledge. Based on decision rules derived from these models, the rockburst occurrence and intensity can be evaluated easily. The results revealed that the proposed approach is a useful and robust technique for long-term prediction of rockburst.
AB - Based on reported statistics, rockburst phenomenon is the main cause of many casualties and accidents occurred during the construction of deep underground structures. Therefore, its prediction in initial stages of design has a remarkable role on enhancement of safety. In this paper, two models have been developed for rockburst evaluation using the C5.0 decision tree classifier. The first model has been applied for prediction of rockburst occurrence and the second model for prediction of rockburst intensity. These models have been developed based on a database including 174 rockburst case histories. In both models, stress coefficient, rock brittleness coefficient, and the elastic strain energy index are the predictive variables. These models are easy to use and do not require extensive knowledge. Based on decision rules derived from these models, the rockburst occurrence and intensity can be evaluated easily. The results revealed that the proposed approach is a useful and robust technique for long-term prediction of rockburst.
UR - http://www.scopus.com/inward/record.url?scp=85059867552&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85059867552&partnerID=8YFLogxK
U2 - 10.1007/s00366-018-00695-9
DO - 10.1007/s00366-018-00695-9
M3 - Article
AN - SCOPUS:85059867552
SN - 0177-0667
VL - 36
SP - 213
EP - 225
JO - Engineering with Computers
JF - Engineering with Computers
IS - 1
ER -