基于回归分析方法的城市火灾预测——以天津市为例

Forecasting Model of the Fire Causes in the City Based on Linear Regression——Taking Tianjin City for Example

城市火灾预测对于火灾预防和消防部署具有一定的指导意义。本文采用回归分析方法建立了幂函数、指数函数等多种火灾预测数学模型,应用所建立的火灾预测模型和SPSS软件对天津市7年的火灾统计数据进行了曲线拟合,并对各种模型进行了显著性检验和比较,结果表明指数模型作为火灾起数随年份变化的曲线模型,预测的下一年火灾情况与实际情况基本吻合。

City fire forecasts have some guidance for fire prevention and fire fighting arrangements.This paper,by setting up fire prediction models such as power function,exponential function and others based on regression analysis,analyzes the data of fire of 7 years in Tianjin City.It also applies the models and SPSS software for statistical data curve fitting and makes the significance test and comparison of these models.The results show that the exponential function model,as the curve model that serves as the fire starting number and varies with the year,can achieve the prediction which is consistent with the actual fire conditions of the next year.