论文部分内容阅读
据１９８３～１９９３年吉林省各地区森林火灾资料计算出森林火灾发生率（Ｙ１）、受害率（Ｙ２）、控制率（Ｙ３）、燃烧率（Ｙ４）；同期的森林覆被率（Ｘ１）、湿润系数（Ｘ２）、人口密度（Ｘ３）、单位了望台控制面积（Ｘ４）等为制约因素，经相关和回归计算建立了适合各地使用的多元回归预测预报方程；此方程是以Ｘ１，Ｘ２，Ｘ４为自变量，以发生率Ｙ１为应变量的三元二次方程：Ｙ＝Ａ０＋Ａ１Ｘ２１＋Ａ２Ｘ２２＋Ａ３Ｘ２４＋Ａ４Ｘ１Ｘ２＋Ａ５Ｘ１Ｘ４＋Ａ６Ｘ２Ｘ４＋Ａ７Ｘ１＋Ａ８Ｘ２＋Ａ９Ｘ４其模拟值与实际值误差很小，是一个较理想的林火发生率的预测预报数学模型。
According to the forest fire data from 1983 to 1993 in Jilin Province, the incidence of forest fire (Y1), victimization rate (Y2), control rate (Y3) and combustion rate (Y4) (X2), population density (X3) and control area of observatory (X4), the multiple regression forecasting forecasting equation is established by correlation and regression calculation. The equation is based on X1, X2 , X4 is an independent variable, the incidence of Y1 as the dependent variable of the quadratic quadratic equation: Y = A0 + A1X21 + A2X22 + A3X24 + A4X1X2 + A5X1X4 + A6X2X4 + A7X1 + A8X2 + A9X4 its analog value and the actual value of the error is small, is a more ideal forest fire incidence prediction mathematical model.