摘要
卡尔曼(Kalman)滤波为线性最优递推滤波算法,但鲁棒性差,无法实时精确跟踪系统突变状态。为此,设计了一款双渐消因子调节的自适应Kalman滤波器。算法剖析了状态扰动环境下,不精准的先验预测及定量滤波增益对最优估计的影响。在标准Kalman滤波器的基础上,引入双渐消因子,实时激活滤波增益,调节先验估计及量测新息在状态估计中的权重。基于新息正交性定理,依据Sage开窗估计原理与加权最小二乘准则,建立了双渐消因子的函数解析式。借鉴滤波发散判据,构造了函数边界条件。实例研究表明,相较于抗差Kalman滤波器,自适应Kalman滤波器鲁棒性强,状态收敛速度快,稳态跟踪精度提升了44.76%。
Kalman filtering is linear optimal recursive filtering algorithm.However,it has poor robustness.It is impossible to accurately track breaking state of the system in real time.An adaptive Kalman filter with double fading factor adjustment was designed.About the algorithm,the influence of inaccurate priori prediction and quantitative filtering gain on the optimal estimation under state disturbance environment was analyzed,based on standard Kalman filtering,then double fading factors were introduced.Filtering gain was activated in real-time and regulated variable measure contribution degree of innovation in the state estimation.It was worth to use orthogonality theorem of innovation for reference,based on fenestration estimation principle and weighted least squares,and functional analytic formula of double fading factors was constructed.Using criterion of filtering divergence,this study analyzed the relationship between reserve coefficient and covariance of innovation,and function boundary condition was constructed.The case study shows that compared with the anti-error Kalman filter,adaptive Kalman filter has strong robustness and fast convergence speed.The steady-state tracking accuracy is improved by 44.76%.
作者
朱文超
何飞
Zhu Wenchao;He Fei(The 38th Research Institute of China Electronics Technology Group Corporation,Hefei 230041,China;Department of Electronic Engineering and Information Science,University of Science and Technology of China,Hefei 230027,China;Institute of Intelligent Machines,Chinese Academy of Science,Hefei 230031,China)
出处
《石家庄铁道大学学报（自然科学版）》
2019年第4期66-72,共7页
Journal of Shijiazhuang Tiedao University(Natural Science Edition)
基金
国家自然科学基金(No.61473272)
