针对含异常观测值的非线性系统滤波问题，以Huber损失函数替代推导滤波器最大后验准则中观测误差的l2范数，构造出了一种新的优化准则函数，从而给出了一种对异常值鲁棒的非线性后验线性化滤波器。分析表明：由于Huber损失函数兼具l1和l2范数的性质，从而使得由这个新准则推导出的滤波器，不仅具有l2范数的低误差拟合性，也具备l1范数对异常值的鲁棒性。而当观测噪声的分布未知时，通过引入箱线图法检测异常值，并对噪声统计分布的参数进行估计，进一步提出了对异常值和未知观测噪声分布鲁棒的非线性后验线性化滤波器。仿真实验验证了分析结果的有效性，并表明本文算法的性能优于现有文献报道的非线性滤波算法。
An outlier-robust posterior linearization filter is presented to address the filtering problem of nonlinear systems with observation outliers. Analysis shows that a new optimization criterion function can be constructed when the l2 norm of the observation errors in a maximum posterior criterion in filter derivation is replaced by the Huber loss function. Since the Huber loss function has the properties of both l1 and l2 norms, the filter derived from the new criterion presented in this paper not only has a low fitting error as the l2 norm does, but is robust to outliers as the l1 norm. When the distribution of the observation noise is unknown, a boxplot method is introduced to detect the outliers in the observations and estimate the statistical distribution parameters of the observation noise. Thus, an outlier and unknown observation noise-robust posterior linearization filter is further proposed. Simulation verifies the analytical results and further shows that the performance of the proposed algorithms surpasses that of the nonlinear filtering algorithms reported in the literature.
[1] ZHANG G D, LI C L. An UKF target tracking method with effective suppression of angular glint[C]//2013 10th IEEE International Conference on Control and Automation (ICCA). Piscataway:IEEE Press, 2013:956-960.
[2] PICHÉ R, SÄRKKÄ S, HARTIKAINEN J. Recursive outlier-robust filtering and smoothing for nonlinear systems using the multivariate student-t distribution[C]//2012 IEEE International Workshop on Machine Learning for Signal Processing. Piscataway:IEEE Press, 2012:1-6.
[3] LI H W, WANG J. Particle filter formanoeuvring target tracking via passive radar measurements with glint noise[J]. IET Radar, Sonar & Navigation, 2012, 6(3):180.
[4] DONG P, JING Z L, LEUNG H, et al. The labeled multi-bernoulli filter for multitarget tracking with glint noise[J]. IEEE Transactions on Aerospace and Electronic Systems, 2019, 55(5):2253-2268.
[5] SARKKA S. Bayesian filtering and smoothing[M].Cambridge:Cambridge University Press, 2009.
[6] ZHONG S K, CHIRARATTANANON P. Direct visual-inertial EGO-motion estimation via iterated extended Kalman filter[J]. IEEE Robotics and Automation Letters, 2020, 5(2):1476-1483.
[7] JULIER S J, UHLMANN J K. Unscented filtering and nonlinear estimation[J]. Proceedings of the IEEE, 2004, 92(3):401-422.
[8] ARASARATNAM I, HAYKIN S. Cubature Kalman filters[J]. IEEETransactions on Automatic Control, 2009, 54(6):1254-1269.
[9] ARASARATNAM I, HAYKIN S, ELLIOTT R J. Discrete-time nonlinear filtering algorithms using gauss-hermite quadrature[J]. Proceedings of the IEEE, 2007, 95(5):953-977.
[10] MURATA M, KAWANO I, INOUE K. Extended, unscented Kalman, and Sigma point multiple distribution estimation filters for nonlinear discrete state-space models[J]. IEEE Control Systems Letters, 2020, 4(4):982-987.
[11] GUO R X, WEI Z L, WEI Y. State estimation for the electro-hydraulic actuator based on particle filter with an improved resampling technique[J]. Proceedings of the Institution of Mechanical Engineers, Part O:Journal of Risk and Reliability, 2020, 234(1):41-51.
[12] GARCÍA-FERNÁNDEZ Á F, SVENSSON L, MORELANDE M R, et al. Posterior linearization filter:Principles and implementation using sigma points[J]. IEEE Transactions on Signal Processing, 2015, 63(20):5561-5573.
[13] LIU X, QU H, ZHAO J H, et al. Maximum correntropy unscented Kalman filter for spacecraft relative state estimation[J]. Sensors, 2016, 16(9):1530.
[14] YIN L J, DENG Z H, HUO B Y, et al. Robust derivative unscented Kalman filter under non-Gaussian noise[J]. IEEE Access, 2018, 6:33129-33136.
[15] MARONNA R. Robust statistical methods[M]//International encyclopedia of statistical science. Berlin, Heidelberg:Springer, 2011:1244-1248.
[16] KARLGAARD C D, SCHAUB H. Huber-based divided difference filtering[J]. Journal of Guidance, Control, and Dynamics, 2007, 30(3):885-891.
[17] WANG X, CUI N, GUO J. Huber-based unscented filtering and its application to vision-based relative navigation[J]. IET Radar, Sonar & Navigation, 2010, 4(1):134.
[18] KARLGAARD C D, SCHAUB H. Adaptive nonlinear Huber-based navigation for rendezvous in elliptical orbit[J]. Journal of Guidance, Control, and Dynamics, 2011, 34(2):388-402.
[19] SCARDUA L A, DA CRUZ J J. Complete offline tuning of the unscented Kalman filter[J]. Automatica, 2017, 80:54-61.
[20] CILDEN-GULER D, RAITOHARJU M, PICHE R, et al. Nanosatellite attitude estimation using Kalman-type filters with non-Gaussian noise[J]. Aerospace Science and Technology, 2019, 92:66-76.
[21] DOVOEDO Y H, CHAKRABORTI S. Boxplot-based outlier detection for the location-scale family[J]. Communications in Statistics-Simulation and Computation, 2015, 44(6):1492-1513.
[22] HERSHEY J R, OLSEN P A. Approximating the kullback leibler divergence between Gaussian mixture models[C]//2007 IEEE International Conference on Acoustics, Speech and Signal Processing-ICASSP'07. Piscataway:IEEE Press, 2007:IV-317-IV-320.
[23] GREWAL M S, ANDREWS A P. Kalman filtering[M]. Hoboken:John Wiley & Sons, Inc., 2014.
[24] 李伟. 鲁棒自适应滤波算法及在飞行器技术中的应用研究[D]. 上海:上海交通大学, 2014:29-30. LI W. Research on robust adaptive filter algorithms and their applications to flight vehicle technology[D]. Shanghai:Shanghai Jiaotong University, 2014:29-30(in Chinese).
[25] 耿修林. 方差推断时样本容量的确定[J]. 统计与决策, 2008(16):23-25. GENG X L. Determination of sample size when inferring variance[J]. Statistics and Decision, 2008(16):23-25(in Chinese).
CJA