论文部分内容阅读
分析中国火灾的历年统计数据,发现中国火灾发生规律同时具有增长趋势性和周期波动性特征.借助于M ATLAB软件,根据2000-2006年中国火灾统计数据,分别建立了火灾24小时发生起数的(1)趋势灰色预测与周期波动指数相乘的组合预测模型和(2)二次曲线趋势与Fourier级数叠加的组合预测模型,两个模型(预测值与实际值)的平均相对误差都小于0.09.研究结论为消防研究、消防部门决策提供科学依据.
Based on the statistical data of Chinese fires over the years, it is found that the regularity of Chinese fires has the characteristics of both increasing trend and cyclical fluctuation.With M ATLAB software, based on the Chinese fire statistics from 2000 to 2006, (1) The combined forecasting model which multiplies trend gray forecast and periodic volatility index and (2) the combined forecasting model that the trend of quadratic curve and Fourier series stack up, the average relative errors of the two models (predicted value and actual value) are both less than 0.09. Research conclusions provide a scientific basis for fire research and fire department decision-making.