基于Time Series-Markov模型的煤矿瓦斯事故起数预测

    The prediction of coal mine gas accident number based on Time Series-Markov model

    • 摘要: 本文以2001~2016年我国煤矿瓦斯事故起数为基础,利用时间序列预测模型及改进马尔科夫预测模型分别预测了2001~2010年、2001~2011年、…及2001~2015年中各年瓦斯事故起数,并计算了其相对误差。其中,TS分别计算的上述六组值的相对误差平均值在18.72%~23.4%之间,而TSM计算的对应值为5.79%~7.09%,且TSM的预测值的波动趋势更符合真实情况。将上述两种模型分别预测后计算的2011~2016各年瓦斯事故发生起数的相对误差进行线性拟合,发现TSM的预测精度更高。因此,用TSM预测煤矿瓦斯事故起数比用TS预测更可靠,这也间接反映了TSM比TS更多地考虑了因素的近期状况对预测值的影响。最后,用此法预测了2017~2020年我国煤矿瓦斯事故起数,其依次为6起、7起、6起及4起。

       

      Abstract: Based on the data of coal mine gas accidents in China from 2001 to 2016,this paper has predicted the number of accidents each year range from 2001 to 2010,2011,…,2015 with the Time Series model (TS) and the Time Series-Markov model (TSM),and the relative error is calculated.The average relative error of the number of accidents of the six groups with the calculation of TS is between 18.72% and 23.4%,while the corresponding value of TSM is between 5.79% and 7.09%,and the fluctuant trend of TSM’s predicted value is more in line with the real situation.With the help of the above two prediction models,Linear fitting the relative error of the number of gas accidents in each year from 2011 to 2016,It is found that the fitting curve of TSM is more accurate.Therefore,the use of TSM to predict the number of coal mine gas accidents is more reliable than TS’s,it also reflects that TSM has considered more about the affect of the forecast value by recent factors than TS.Finally,the number of coal mine gas accidents in China from 2017 to 2020 is predicted by the TSM, and the results of those were 6, 7, 6, 4.

       

    /

    返回文章
    返回