煤矿安全是当前安全生产工作的重中之重。为掌握煤矿安全生产情况,降低事故损失,保证中国煤炭工业健康、快速、可持续发展,本文在传统GM(1,1)模型的基础上,建立了关于煤矿百万吨死亡率的尾部残差修正GM(1,1)模型。将该方法应用于2001—2011年全国煤矿百万吨死亡率分析,并以此为基础对2012、2013年的煤矿百万吨死亡率进行预测,与传统GM(1,1)模型的预测结果进行对比分析。研究结果表明,传统的GM(1,1)模型精度较差,最大误差达到14.35%,经修正的尾部残差GM(1,1)模型预测结果可靠,实际值与预测值平均相对误差1.14%,最大相对误差3.81%,各项指标均明显优于传统的GM(1,1)预测模型,为政府、矿山企业制定安全生产目标、政策以及建立科学高效的安全管理机制提供理论依据。

Coal mine safety is the top priority of the current production safety work, in order to master the mine safety production status in China, decrease the accident loss, and ensure a healthy, rapid, and sustainable development of China's coal industry, a modified empennage residual error GM (1, 1) model is built on the basis of the traditional GM (1,1) model to analyze the Death Rate Per Million Ton (DRPMT) of coal mines from the year of 2001 to 2011, and to predict the DRPMT of coal mines in the year of 2012 and 2013, respectively. The result shows that the accuracy of traditional GM (1, 1) model is quite poor with the maximum error of 14.35%. Compared with the result given by traditional GM (1, 1) model, the proposed method is easy to operate with reliable results. The average relative error is 1.14% and the maximum relative error is 3.81%, which are all better than that by using the GM (1, 1) model. The method provides the theory guidance for governments and mining enterprises to set safe production targets and policies, establish scientific and efficient security management mechanism. And the method has the great application value in practice.