改进的散粒噪声模型及其应用

Improved shot noise models and its applications

对Ｇ．Ｗｅｉｓｓ提出的单散粒噪声模型中的复合泊松过程作了改进，把齐次的泊松过程改为强度函数为λ（ｔ）的非齐次泊松过程，把伴随的随机变量序列改为与时间有关，并对随机事件赋予新的定义，进而用多项式函数根据最小二乘原理拟合参数函数λ＊（ｔ）和θ＊（ｔ）。最后就我国一些河川流量资料举例说明此模型的应用．

In this paper, the compound Poisson process of shot noise models proposed by G. Weiss in 1970s is replaced by their non-homogeneous versions. The parametric functions of this model are fitted using the least square principle by using polynomial functions. The application of this model to streamflow data is illustrated with case studies for some rivers in China.