基于离差最大化组合赋权的煤矿安全评价研究

针对组合赋权中权重分配不够客观准确的问题,提出基于离差最大化组合赋权的煤矿安全评价法。综合分析煤矿安全生产的风险原因,建立包含12个一级指标、50个二级指标的评价体系。采用熵权法与变异系数法计算各指标的客观权重,通过构建的离差最大化组合预警评价模型使两种赋权法的组合结果最佳,利用Spearman秩相关系数法检验其兼容度。以6所矿井为例进行实例验证,得到的风险等级量化值为{0.65,0.62,0.60,0.32,0.63,0.55},兼容度为0.8632。结果表明该模型充分融合了指标的权重信息,避免了主观分配法和平均分配法忽略指标数据实际意义的错误,验证了模型的可靠性与有效性。

COAL MINE SAFETY ASSESSMENT BASED ON MAXIMUM DEVIATION COMBINATION EMPOWERMENT

Aiming at the problem that the weight distribution in combination empowerment is not objective and accurate enough,a method of coal mine safety evaluation based on maximum deviation combination empowerment is put forward.After comprehensive analysis of the risk causes of coal mine safety production,an evaluation system including 12 first-level indicators and 50 second-level indicators was established.The objective weights of each index were calculated by using the method of entropy weight and coefficient of variation,and the best combination result of the two weighting methods was obtained by constructing the combined early warning evaluation model of maximum deviation.The compatibility of the two weighting methods was tested by Spearman rank correlation coefficient method.Six mines were taken as examples to verify the effectiveness of the proposed method.The quantified values of coal mine safety risk grade were{0.65,0.62,0.60,0.32,0.63,0.55}and the compatibility was 0.8632.The results show that the model fully integrates the weight information of indicators,avoids the errors of subjective allocation method and average allocation method neglecting the actual significance of indicators data,and verifies the reliability and validity of the model.