-
摘要: Kalman滤波在系统控制、信号处理和飞行器导航等领域有广泛应用.众所周知，Kalman滤波是建立在一组递推计算的滤波算法.不准确的初始值设置是否影响状态滤波结果，这是实际应用Kalman滤波时必须关注和解决的问题.本文在推导滤波初始值与系统真实初态之间偏差对滤波结果影响的基础上，建立了初始值偏差对后续滤波影响的传递关系，以及滤波结果收敛的充分条件；通过设置多组不同初始值偏差，分别对某三阶可观测系统和不可观测系统进行了仿真计算及结果分析，验证了滤波初始值偏差会导致滤波结果发生明显偏离.研究结果揭示，即使是可观测的线性系统，采用Kalman滤波时也必须尽可能选取准确的初始值.Abstract: Kalman filter is widely used in system control, signal processing, spacecraft navigation and other fields. It has come to light that the Kalman filter is composed of a series of recursion algorithms. When the filter is used in practical projects, it is an open problem whether the biased initialization is infectious to Kalman filtering results or not, which must be paid much attention to and solved down. In this paper, the effect of initialization deviation on the recursive Kalman filter is discussed, a transmission formula is built to analyze the Kalman filtering difference which comes from initialization deviation, and a sufficient condition is deduced for convergence of the Kalman filter of state vectors. Setting several kinds of initial deviation values for simulation, series of simulation results are given for a 3rd-order observable system as well as a 3rd-order unobservable linear system. Simulation results show that deviation of initial value may result in distinct difference of the state filtering. The results given in this paper reveal that the initial value must be carefully selected even if the dynamic system is an observable linear time-invariant systems.
-
Key words:
- Kalman filter /
- initial value /
- initial deviation /
- convergence /
- eigenvalue
1) 本文责任编委 夏元清 -
图 1 初始值偏10%时的Kalman滤波比对差
Fig. 1 Difference of Kalman filtering states under 10% deviation from the initial state of dynamic system
图 2 初始值偏10%的滤波与实际状态比对差
Fig. 2 Difference of Kalman filtering and true states under 10% deviation from the initial state
图 3 初始值偏50%时的Kalman滤波比对差
Fig. 3 Difference of Kalman filtering states under 50% deviation from the initial state of dynamic system
图 4 初始值偏50%的滤波与实际状态比对差
Fig. 4 Difference of Kalman filtering and true states under 50% deviation from the initial state
-
[1] Kalman R. Phase-plane analysis of automatic control systems with nonlinear gain elements. American Institute of Electrical Engineers, Part Ⅱ: Applications and Industry, Transactions of the 1955, 73(6): 383-390 doi: 10.1109/TAI.1955.6367086 [2] 陈关荣.复杂动态网络环境下控制理论遇到的问题与挑战.自动化学报, 2013, 39(4): 312-321 http://www.aas.net.cn/CN/abstract/abstract17870.shtmlChen Guan-Rong. Problems and Challenges in Control Theory under Complex Dynamical Network Environments. Acta Automatica Sinica, 2013, 39(4): 312-321 http://www.aas.net.cn/CN/abstract/abstract17870.shtml [3] 陈杰, 方浩, 辛斌, 邓方.数字化陆用武器系统中的建模、优化与控制.自动化学报, 2013, 39(7): 943-962 http://www.aas.net.cn/CN/abstract/abstract18123.shtmlChen Jie, Fang Hao, Xin Bin, Deng Fang. Modeling, optimization and control in ground-based digital weapon systems. Acta Automatica Sinica, 2013, 39(7): 943-962 http://www.aas.net.cn/CN/abstract/abstract18123.shtml [4] 包为民.航天飞行器控制技术研究现状与发展趋势.自动化学报, 2013, 39(6): 697-702 http://www.aas.net.cn/CN/abstract/abstract18095.shtmlBao Wei-Min. Present situation and development tendency of aerospace control techniques. Acta Automatica Sinica, 2013, 39(6): 697-702 http://www.aas.net.cn/CN/abstract/abstract18095.shtml [5] Luke R, Wouters J. Kalman filter based estimation of auditory steady state response parameters. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2017, 25(3): 196-204 doi: 10.1109/TNSRE.2016.2551302 [6] Wiberg D M, Powell T D, Ljungquist D. An online parameter estimator for quick convergence and time-varying linear systems. IEEE Transactions on Automatic Control, 2000, 45(10): 1854-1863 doi: 10.1109/TAC.2000.880986 [7] Cattivelli F S, Sayed A H. Diffusion strategies for distributed Kalman filtering and smoothing. IEEE Transactions on Automatic Control, 2010, 55(9): 2069-2084 doi: 10.1109/TAC.2010.2042987 [8] Lu X, Wang L L, Wang H X, Wang X H. Kalman filtering for delayed singular systems with multiplicative noise. IEEE/CAA Journal of Automatica Sinica, 2016, 3(1): 51-58 doi: 10.1109/JAS.2016.7373762 [9] Nikoukhah R, Campbell S L, Delebecque F. Kalman filtering for general discrete-time linear systems. IEEE Transactions on Automatic Control, 1999, 44(10): 1829-1839 doi: 10.1109/9.793722 [10] Ge Q B, Shao T, Duan Z S, Wen C L. Performance analysis of the Kalman filter with mismatched noise covariances. IEEE Transactions on Automatic Control, 2016, 61(12): 4014-4019 doi: 10.1109/TAC.2016.2535158 [11] Bjorn U. Asymptotic convergence properties of the extended Kalman filter using filtered state estimates. IRE Transactions on Automatic Control, 1980, 25(6): 1207-1211 doi: 10.1109/TAC.1980.1102518 [12] 邓自立, 郭一新.现代时间序列分析及其应用-建模、滤波、去卷、预报和控制.北京:知识出版社, 1989Deng Zi-Li, Guo Yi-Xin. The Modern Time Series Analysis and Its Application: Modeling, Filtering, Deconvolution, Forecast and Control. Beijing: Knowledge Press, 1989 [13] Yuki H, Takao S, Akira A, Tetsuya M. Application of the Kalman filter for faster strongcoupling of cardiovascular simulations. IEEE Journal of Biomedical and Health Informatics. 2015: 1-7 https://www.researchgate.net/publication/277413658_Application_of_the_Kalman_Filter_for_Faster_Strong_Coupling_of_Cardiovascular_Simulations [14] Jafar Z, Ehsan S. Convergence analysis of non-linear filtering based on cubature Kalman filter. IET Science, Measurement & Technology, 2015, 9(3): 294-305 https://www.researchgate.net/publication/273799698_Convergence_analysis_of_non-linear_filtering_based_on_cubature_Kalman_filter [15] Francesco A, Cangemi T, D'Ippolito F, Fagiolini A, Sferlazza A. Convergence analysis of extended Kalman filter for sensorless control of induction motor. IEEE Transactions on Industrial Electronics, 2015, 62(4): 2341-2352 doi: 10.1109/TIE.2014.2355133 [16] Martin W C, Stubberud A R. An uncoupling method for linear system identification. IEEE Transactions on Automatic Control, 1976, 21(4): 506-509 doi: 10.1109/TAC.1976.1101303 [17] Sangsuk-Iam S, Bullock T E. Analysis of discrete-time Kalman filtering under incorrect noise covariances. IEEE Transactions on Automatic Control, 1990, 35(12): 1304-1309 doi: 10.1109/9.61006 [18] Cinquemani E. A superposition principle for the Kalman filter. Systems & Control Letters, 2006, 55(1): 38-44 https://www.researchgate.net/publication/222355182_A_superposition_principle_for_the_Kalman_filter [19] Liu X, Escamilla-Ambrosio P J, Lieven N A J. Extended Kalman filtering for the detection of damage in linear mechanical structures. Journal of Sound and Vibration, 2009, 325(4-5): 1023-1046 doi: 10.1016/j.jsv.2009.04.005 -
下载:
图(5)
计量
- 文章访问数: 2360
- HTML全文浏览量: 187
- PDF下载量: 851
- 被引次数: 0
