基于无偏灰色模型的煤矿百万吨死亡率预测

Application of Unbiased Grey-forecasting Model in Prediction of Million Tons Death Rate of Coal Mine

为预测中国煤炭行业安全生产状况发展趋势,构建无偏灰色预测模型对我国煤矿百万吨死亡率数据进行模拟预测。通过对我国2005—2010年度煤炭行业的百万吨死亡率有关数据进行模拟,并与传统的灰色GM(1,1)模型的预测模拟结果进行对比分析。计算结果表明:无偏灰色模型消除了传统GM(1,1)模型本身固有偏差,预测精度较高,分析结果可靠,其中,平均绝对误差为0.030 6%,平均相对误差为2.71%,均低于传统GM(1,1)模型。预测数据显示近年来我国煤矿百万吨死亡率正逐步下降,2011年和2012年分别降至0.536%和0.411%,符合煤矿安全生产"十二五"规划要求。

In order to accurately predict the development trend of work safety status of China＇s coal industry,the unbiased grey-forecasting model was built to simulate the death rate of million tons of coal mines during 2005 and 2010.Compared with the result gotten by traditional gray GM（1,1） model,the results show that the unbiased gray model eliminates the inherent bias of the GM（1,1） model and the prediction accuracy is higher and the error is relatively low.The average absolute error is 0.030 6% and the average relative error is 2.71%,which are all lower than those by the traditional GM（1,1） model.The forecast data shows that the death rate of million tons in China＇s coal mines is gradually decreasing in recent years with being down to 0.536% in 2011 and 0.411% in 2012,respectively.The predictions satisfy the requirements of the Twelfth Five-Year Guideline for work safety of coal mine.