甘肃省火灾事故预测模型构建及应用研究

Investigation on prediction model and application of fire accident in Gansu province

针对甘肃省火灾事故预测方法误差大的问题,依据无偏灰色GM(1,1)理论,以2013-2017年全省火灾发生起数为样本数据,建立了全省火灾事故起数预测模型,并进行后验差检验。结果表明:无偏灰色预测模型预测值与实际值的平均误差为7.725%,能够精准预测火灾发生起数。

In view of the large error in the prediction methods of fire accidents in Gansu Province,based on the theory of unbiased grey GM (1,1),using the number of fires in Gansu province from 2013 to 2017 as sample data,a prediction model for the number of fire accidents in the province was established.And post-inspection test was made.The results show that the average error between predicted value and actual value of the unbiased gray prediction model is 7.725%,which shows that the prediction model can accurately predict the number of fire occurrences.