Citation

Bi J, Li X, Chen L. China Saf. Sci. J. 2019; 29(12): 103-109.

Copyright

(Copyright © 2019, China Occupational Safety and Health Association, Publisher Gai Xue bao)

DOI

10.16265/j.cnki.issn1003-3033.2019.12.017

PMID

unavailable

Abstract

In order to improve prediction accuracy of hazard levels of rock burst with a small sample, an R-type factor analysis-Fisher discriminant prediction model is proposed. Firstly, with Yanshitai Coal Mine as an example, R-type factor analysis was used to process risk assessment index of rock burst, characteristic information of original index was extracted which were then replaced by a small number of main factors, and qualitative analysis of hazard levels was performed. Secondly, Fisher discriminant method was adopted to analyze R-type factor analysis results to determine distance between assessment set and different risk levels, and accuracy of judgment matrix was improved by rejudging training set. Finally, risk levels of rock burst were predicted according to judgment matrix. The results show that this model can weaken interaction between indicators and significantly improve predication accuracy for small samples. © 2019 China Safety Science Journal. All rights reserved.

Language: zh