建筑安全事故灰色季节指数预测模型及应用

Grey seasonal index prediction model for construction safety accident and its application

为构建能针对建筑安全系统事故特征的事故预测模型,从气候、地域性、传统节日等方面分析我国房屋建筑及市政工程生产安全事故数据。基于灰色系统理论及季节变动预测方法,建立建筑安全事故灰色季节指数预测模型。用GM(1,1)模型作为季节指数预测的趋势方程,使不满1个整数周期的事故量统计数据参与建模,实现对模型的即时更新。结果表明:建筑安全事故时间序列有以指数变化为特征的趋势性和以年为周期的周期性;在建筑安全事故预测应用中,灰色季节指数预测模型平均预测准确度达到89%,比常规趋势方程季节指数预测方法预测精度提高5%以上。

To building an accident prediction model, which should consider characteristics of construc- tion safety system accident, data on housing construction and municipal engineering production safety acci- dents in China from 2008 January to 2013 May were analyzed from aspects of climate, region, traditional festivals and others. Based on grey system theory and seasonal variation forecasting method, a grey seasonal index prediction model was built for construction safety accident, in which GM （ 1,1 ） model was used as trend equation of seasonal index prediction, and accident data without integral period were involved in modeling to achieve instant updating the model. The model was used to predict construction safety accidents. The results show that, the average prediction accuraey of grey seasonal index prediction model is 89% , 5% at least better than that of seasonal index prediction model with conventional trend equation.