使用自回归滑动平均(ARMA)和广义自回归条件异方差(GARCH)过程对金融数据建模是经济学常用手段。文中结合ARMA 过程和GARCH 过程的非线性化扩展模型,将其扩展到复数域,适合于海杂波建模应用。相比传统的海杂波模型及原始的GARCH 模型,文中提出的模型在概率密度函数拟合上具有明显的优势。此外,新模型还可准确地捕获相邻海杂波中存在的强相关性。实际雷达海杂波数据验证了该模型的准确性和有效性。

Autoregressive moving-average ( ARMA ) process and generalized autoregressive conditional heteroskedasticity (GARCH) process are common tools for modeling financial data in economics. In order to make it suitable for modeling sea clutters, we combine the nonlinear expansion models of ARMA and GARCH, and extend it to the complex domain. Compared with the traditional sea clutter models and the original GARCH model, the proposed model has obvious advantages in probability density function fitting. Besides, the new model can accurately capture the strong correlation in adjacent sea clutters. Empirical studies on actual sea clutter data show the accuracy and effectiveness of our model.

TN957.9