非负门限自回归模型NTAR（1）

NONNEGATIVE-THRESHOLD AUTOREGRESSIVE MODELS NTAR(1)

本文对一阶非负门限自回归模型ＮＴＡＲ（１）作了讨论，指出ＮＴＡＲ（１）模型有根强的实际背景，用马尔克夫理论证明了ＮＴＡＲ（１）模型的平稳性、几何遍历性，给出了矩的存在条件和参数估计。

The non-negative nolinear time series model of the first order(NTAR(1 ))isproposed which can be used to model such a series as river flow.The stationarity andergodicity of the model are obtained by using markoy chain theory developed recently byTweedle.The moment conditions of the model are also discussed,It is shown that theconditional least square estimators of the model parameters are consistent under certainConditions.