Study on characteristics of unsafe behaviors in coal mines based on SCM and K-means clustering analysis

Abstract

In order to reduce the safety accidents caused by the unsafe behaviors in the managed operation of coal mines, we conducted a scientific classification study on the unsafe behaviors of personnel under such operation mode. The field investigation on the “three violations” behaviors of 1 996 coal miners in six coal mines in Inner Mongolia, Ningxia and Xinjiang from 2017 to 2018 was conducted. All the three violations were statistically classified from three aspects of time, job type and characteristics of SCM behavior generation. A data set of 8 indicators and 4 subcategories was established based on K-means clustering algorithm, and a visual clustering scatter diagram was drawn through PCA dimension reduction. The analysis shows that: using SCM and K-means clustering to calculate the proportion of four kinds of unsafe action classification, the same conclusion was obtained as manual analysis, in all unsafe action, lawless accounted for the biggest, errors accounted for the smallest proportion. The research results have certain guiding significance for reducing unsafe behaviors of coal mine employees and preventing safety accidents by classification.

Keywords

• coal mine safety
• scm model
• k-means clustering analysis
• pca dimension reduction
• unsafe behaviour