Research Article

Coal Mine Safety Evaluation Based on Machine Learning: A BP Neural Network Model

Table 1

Normalization rules of neural network input data.

Evaluation indexNetwork input
RangeValueRangeValueRangeValue

Employee medical insurance rate (%)≤100.1(10, 100)Y/1001001
Per capita housing area of employees (m2/person) = 00(0, 20)Y/20≥201
Average education level of employeesBelow medium0.3–0.5Fairly good0.7Good0.9
Enrollment rate of employees' children (%)≤100.1(10, 100)(Y-80)/(120–80)1001
Reserve production ratio (%)≤800.1(80, 120)Y/100≥1201
Recovery ratio (%)≤100.1(10, 100)Y/1001001
Comprehensive utilization degree of resourcesBelow medium0.3–0.5Fairly good0.7Good0.9
Proportion of noncoal industry (%)≤100.1(10, 100)Y/1001001
Total efficiency of raw coal (t/person) = 00(0, 4)Y/4≥41
Per capita income of miners (yuan/month)≤10000.1(1000, 10000)Y/10000≥100001
Ratio of profits and taxes on funds (%) = 00(0, 20)Y/20≥201
Land reclamation rate of mining area (%)≤100.1(10, 100)Y/1001001
Standard rate of noise control in mining area (%)≤100.1(10, 100)Y/1001001
Standard rate of wastewater treatment in mining area (%)≤100.1(10, 100)Y/1001001
Standard rate of waste gas treatment in mining area (%)≤100.1(10, 100)Y/1001001
Standard rate of solid waste discharge in mining area (%)≤100.1(10, 100)Y/1001001
Million ton mortality (%) = 01(0, 1)(1-Y)/1≥10