摘要
提出一类有色噪声干扰随机系统的辅助模型最小二乘递推辨识算法.此算法结合辅助模型辨识思想和递推增广最小二乘理论,用辅助模型的输出代替辨识模型信息向量中未知中间变量,用估计残差代替信息向量中不可测噪声项,从而可以运用递推辨识思想来估计系统所有参数,包括噪声模型参数.仿真例子说明提出算法的有效性.
An auxiliary model and least squares based recursive identification algorithm was presented for stochastic systems with colored noises. The basic idea is, by means of the auxiliary model identification principle and extended least squares identification theory, to replace the unknown variables in the information vector with the outputs of the auxiliary model and to replace the unmeasurable noise terms with the estimated residuals, and then to propose a recursive identification algorithm to estimate all parameters of the systems, including the parameters of the noise models. The simulation results indicate that the algorithm is effective.
出处
《北京航空航天大学学报》
EI
CAS
CSCD
北大核心
2008年第8期935-939,共5页
Journal of Beijing University of Aeronautics and Astronautics
基金
国家自然科学基金资助项目(60673101)
关键词
递推辨识
参数估计
最小二乘
随机系统
recursive identification
parameter estimation
least squares
stochastic systems
