摘要
在目标被动式跟踪中广泛应用的伪量测变换估计器(PLE)具有良好的误差收敛性。然而由于等价噪声和状态的相关性,该估计器的估计是有偏的。提出的强跟踪滤波器(STF)通过强制白化残差具有自适应地校正估计偏差和迅速跟踪状态变化的能力。STF已经在非线性系统时滞估计、故障诊断与容错控制方面取得了很好的效果。结合PLE和STF,提出了一种自适应伪量测变换估计器(APLE)。APLE减小了状态的估计偏差,同时对目标的初始值具有极强的鲁棒性。计算机仿真验证了APLE算法的有效性。
The pseudomeasurement line ar estimator (PLE), extensively used in passive tracking due to its good convergen ce of estimate errors, just supplies the biased state estimates because its meas urement matrix is a function of estimate error. The strong tracking filters (STF ) we proposed before can reduce adaptively estimate bias and thus has ability to track abrupt changes in nonlinear systems. Its principle is to identify the cov ariances of predicted state error so that the orthogonality among filter residua ls at different times can be ensured. Now the STF has achieved satisfactory resu lts in the joint estimation of time-delay and parameter, fault diagnosis, fault -tolerant control in nonlinear systems. Here an adaptive PLE (APLE) is proposed through combining the PLE with the STF. Compared with the PLE, the APLE can eff ectively decrease state bias and also has strong robustness to initial state est imate errors in computer simulations.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2003年第7期880-882,886,共4页
Journal of Tsinghua University(Science and Technology)
基金
国家自然科学基金资助项目(60274059
60025307)
教育部博士点基金及中国博士后基金。
