正交级数方法与非平稳时间序列模型估计和检验的一些研究进展

董朝华, 高集体, 朱平芳

Recent Developments on Nonstationary Time Series Model Estimation and Testing Driven by Orthogonal Series Method

Chaohua DONG, Jiti GAO, Pingfang ZHU

表7 检验功效: $ m(x)=(1+x^2)\exp(-\theta_1 x^2)+\Delta_n(x), \theta_1=1$

n 检验水平1% 检验水平10% κ= 1/5 1/4.5 1/4 1/5 1/4.5 1/4 $ \alpha_0=0.1, \qquad a=0.02, \qquad b=-0.3$ 200 0.9968 0.9974 0.9978 0.9980 0.9988 0.9990 500 0.9972 0.9972 0.9974 0.9984 0.9992 0.9994 1200 0.9978 0.9976 0.9986 0.9990 0.9990 0.9996 $ \alpha_0=0.2, \qquad a=0, \qquad b=0.5$ 200 0.9846 0.9856 0.9858 0.9954 0.9964 0.9958 500 0.9860 0.9868 0.9870 0.9956 0.9958 0.9962 1200 0.9864 0.9860 0.9872 0.9986 0.9968 0.9968