Multiple-model self-calibration Kalman filter method
-
摘要:
基于自校准Kalman滤波方法和多模型估计理论,针对工程实际中未知输入(如突风、故障和未知系统误差等)对系统状态方程的影响问题,提出了一种多模型自校准Kalman滤波方法。该方法同时采用自校准Kalman滤波和标准Kalman滤波进行运算,并根据贝叶斯定理自动分配两种方法滤波值的权重,通过加权融合得到最终的滤波结果。与自校准Kalman滤波方法相比,多模型自校准Kalman滤波方法既能有效地补偿非零未知输入的影响,又明显改善了系统在未知输入为零时的滤波精度,大量数值仿真结果表明该方法精度提升可达10%以上,具有更强的适应性和鲁棒性。
Abstract:Based on the self-calibration Kalman filter (SKF) and the multiple-model estimation (MME), considering the influence of unknown inputs (such as gusts, faults, unknown system errors, etc.) on the system state equation in Engineering, the multiple-model self-calibration Kalman filter (MSKF) was proposed. According to the Bayes' theorem, this filtering method used the SKF and the standard Kalman filter (KF) whose weights were assigned automatically to obtain the final filtering result through weight-average way. Compared with the SKF, the MSKF can not only effectively compensate the effects of non-zero unknown inputs, but also improve the estimation accuracy when unknown inputs were zero. A large number of simulation results showed that accuracy can be improved by more than 10%, using the proposed method. In summary, the MSKF has stronger adaptability and robustness.
-
Key words:
- self-calibration filter /
- multiple-model estimation /
- Kalman filter /
- unknown input /
- fault diagnosis
-
表 1 方均根误差均值
Table 1. Mean of RMSEs
方法 方均根误差均值 MSKF 0.0372 SKF 0.0418 KF 0.1169 表 2 各维度方均根误差均值
Table 2. Mean of RMSEs in two dimensions
方法 方均根误差均值 维度一 维度二 MSKF 0.0335 0.0332 SKF 0.0355 0.0354 KF 0.0915 0.0920 -
[1] KALMAN R E. A new approach to linear filtering and prediction problems[J]. Journal of Basic Engineering,1960,82(1): 35-45. doi: 10.1115/1.3662552 [2] SUNAHARA Y. An approximate method of state estimation for nonlinear dynamical systems[J]. Journal of Basic Engineering,1970,92(2): 385-393. doi: 10.1115/1.3425006 [3] FUJIMOTO O,OKITA Y,OZAKI S. Nonlinearity compensation Extended Kalman filter and its application to target motion[J]. Oki Technical Review,1997,63(159): 1-12. [4] JULIER S J, UHLMANN J K. A new extension of Kalman filter to nonlinear systems[C]//Proceedings of 11th International Symposium Aerospace/Defense Sensing, Simulation and Controls. Orlando, US: Society of Photo-Optical Instrumentation Engineers (SPIE), 1997: 182-193. [5] JULIER S,UHLMANN J,DURRANT-WHYTE H F. A new method for the nonlinear transformation of means and covariances in filters and estimators[J]. IEEE Transactions on Automatic Control,2000,45(3): 477-482. doi: 10.1109/9.847726 [6] JULIER S J,UHLMANN J K. Unscented filtering and nonlinear estimation[J]. Proceedings of the IEEE,2004,92(3): 401-422. doi: 10.1109/JPROC.2003.823141 [7] DAN S. Optimal state estimation: Kalman H∞ and nonlinear approaches[M]. Hoboken, US: Wiley & Sons, 2006. [8] PITT M K,SHEPHARD N. Filtering via simulation: auxiliary particle filters[J]. Journal of the American Statistical Association,1999,94(446): 590-599. doi: 10.1080/01621459.1999.10474153 [9] BLANKE M, SCHRÖDER J. Diagnosis and fault-tolerant control[M]. 2nd ed. Berlin, Germany: Springer, 2006. [10] CHEN J, PATTON R. Robust model-based fault diagnosis for dynamic systems[M]. Boston, US: Kluwer Academic Publishers, 1999 [11] GILLIJNS S,DE MOOR B. Unbiased minimum-variance input and state estimation for linear discrete-time systems[J]. Automatica,2007,43(1): 111-116. doi: 10.1016/j.automatica.2006.08.002 [12] 傅惠民,吴云章,娄泰山,等. 自校准Kalman滤波方法[J]. 航空动力学报,2014,29(6): 1363-1368. doi: 10.13224/j.cnki.jasp.2014.06.015FU Huimin,WU Yunzhang,LOU Taishan,et al. Self-calibration Kalman filter method[J]. Journal of Aerospace Power,2014,29(6): 1363-1368. (in Chinese) doi: 10.13224/j.cnki.jasp.2014.06.015 [13] 傅惠民,娄泰山,肖强,等. 自校准扩展Kalman滤波方法[J]. 航空动力学报,2014,29(11): 2710-2715. doi: 10.13224/j.cnki.jasp.2014.11.023FU Huimin,LOU Taishan,XIAO Qiang,et al. Self-calibration extended Kalman filter method[J]. Journal of Aerospace Power,2014,29(11): 2710-2715. (in Chinese) doi: 10.13224/j.cnki.jasp.2014.11.023 [14] MAGILL D. Optimal adaptive estimation of sampled stochastic processes[J]. IEEE Transactions on Automatic Control,1965,10(4): 434-439. doi: 10.1109/TAC.1965.1098191 [15] RONG LI X. Hybrid estimation techniques[J]. Control and Dynamic Systems,1996,76: 213-287. [16] MAZOR E,AVERBUCH A,BAR-SHALOM Y,et al. Interacting multiple model methods in target tracking: a survey[J]. IEEE Transactions on Aerospace and Electronic Systems,1998,34(1): 103-123. doi: 10.1109/7.640267 [17] LI X R,BAR-SHALOM Y. Performance prediction of the interacting multiple model algorithm[J]. IEEE Transactions on Aerospace and Electronic Systems,1993,29(3): 755-771. doi: 10.1109/7.220926 [18] DAEIPOUR E,BAR-SHALOM Y. IMM tracking of maneuvering targets in the presence of glint[J]. IEEE Transactions on Aerospace and Electronic Systems,1998,34(3): 996-1003. doi: 10.1109/7.705913