随机环境下非线性时间序列模型的渐近行为

The Asymptotic Behavior of Nonlinear Time Series Model under the Random Environment

推广了广泛用于金融经济领域的自回归条件异方差模型,提出了随机环境下的幂变换门限自回归条件异方差模型.新模型通过引入有限状态马氏链,增加了环境突变对模型的影响,使得模型具有了更为广泛的适应性.同时,得出了随机环境下的幂变换门限自回归条件异方差模型以几何速率收敛的充分条件.

The article proposes the power transformed threshold autoregressive conditional heteroskedastic model under the random environment which is the generalization of autoregressive conditional heteroskedastic model.By finite state Markov chain,the influence of the sudden change to the environment is promoted in the new modle,which makes the new model more flexible.Meanwhile,some sufficient conditions for convergence are obtained.