基于滤波方法的瓦塞克模型估计研究  

Research on estimations of Vasicek model based on the approach of filters

作　　者：杨宝臣[1] 苏云鹏[1] 

机构地区：[1]天津大学管理学院,天津300072

出　　处：《河北工程大学学报（自然科学版）》2007年第3期80-83,共4页Journal of Hebei University of Engineering:Natural Science Edition

基　　金：国家自然科学基金资助项目(70471051)

摘　　要：将扩展卡尔曼滤波和无损卡尔曼滤波应用于Vasicek模型的参数估计上,并通过对两种滤波方法的估计效果、运算速度等方面进行对比,探讨了EKF和UKF的特点、适用范围以及性能优劣,并得出虽然UKF的计算量要大于EKF,但UKF无论在估计效果上还是在稳定性和鲁棒性上,均明显优于EKF。因此,UKF在利率期限结构模型估计方面具有巨大的应用潜力。EKF and UKF were used to estimate parameters of Vasicek model; the characters of the two fil- ters include applicability scopes and superiority was discussed by contrasting their estimation results in estimarion effects, computing speed and so on. Then the conclusion was derived that although the computation of UKF is larger than EKF, UKF plays a much more excellent performance than EKF at the aspects of stability, robustness and so on, which demonstrates the great potential of UKF in application to model estimation of the term structure of interest rates.

关 键 词：扩展卡尔曼滤波 无损卡尔曼滤波 利率期限结构 VASICEK模型 参数估计 

分 类 号：O212[理学—概率论与数理统计]