基于时间序列与灰色拓扑的春节火灾损失预测

火灾每年给中国带来了巨大的损失,春节期间的火灾损失更是严重.根据1999-2010年春节期间火灾统计资料,火灾四项指标数据具有时序性以及随机波动性、模糊性.运用时间序列与灰色拓扑预测方法相结合预测春节期间火灾发生规律,且预测出未来3年内的火灾发生情况.结果表明,时间序列预测模型的平均绝对误差较小,且所建立的灰色拓扑预测模型的拟合精度都达到"好"的标准.因此,采用时间序列与灰色拓扑预测模型相结合对春节火灾发生情况进行预测,其结果合理可靠,可供理论研究和消防部门做出相应的预防措施参考,以达到有效控制和预防春节火灾的目的.

Prediction of Fire Damage in Spring Festival Based on Time Series and Grey Topology

Fire has caused great damages to China,especially when it happens during Spring Festivals.According to the fire accident statistics between 1999 and 2010,the four index data is characteristic of time sequence,volatility,and vagueness.Based on time series model and topology prediction,this paper aims to predict the rules,according to which fire happens during Spring Festivals,and moreover,to predict its other circumstances.It turns out that,the mean absolute error in the time series model is smaller,and the fitting precision of gray topology prediction model are established to achieve "good" standard.Therefore,the result is reasonable and reliable,if we based the fire damage prediction on a methodology which combines time series model with grey topology prediction.The result can be referred as precautions to fire damages for the fire department and related research projects.