Adaptive Kalman filter based on variational Bayesian approach for one-step randomly delayed measurements

Document Type : Research Article

Authors

Department of Electrical Engineering, National Institute of Technology Durgapur, West Bengal, 713209, India

Abstract

This article addresses the state estimation problem for dynamic systems with linear models wherein covariance matrices of the process and measurement noise are unknown and one step delay randomly occurs in the measurements. Due to network congestion, limited bandwidth during transmission of sensor data to the central processing unit the probability of measurements getting randomly delayed is high and this phenomenon is ignored for conventional adaptive Kalman filters. A new algorithm for Adaptive Kalman filter with one step randomly delayed measurements is proposed here wherein the randomly delayed measurements are modelled using Bernoulli’s distribution. The adaptation algorithm has been mathematically derived for such situations following the variational Bayesian approach and subsequently a recursive algorithm for variational Bayesian adaptive delayed Kalman filter is formulated. Monte Carlo simulation demonstrates the excellence of the proposed filter over the conventional Kalman filter for the estimation problem addressed in this work. The comparative study with the competing maximum likelihood estimation variant also reveals the superiority of the proposed filter. To exemplify the effectiveness of the proposed algorithm for real world applications validation with the real measurement data has been carried out for offline harmonics estimation which ensures satisfactory estimation results.

Keywords

References

References:
1.Kalman, R.E. “A new approach to linear filtering andprediction problems”, J. Basic Eng., 82(1), pp. 35-45(1960). http://doi.org/10.1115/1.3662552.
2.Dorostgan, M. and Taban, M.R. “Adaptive radar signaldetection in autoregressive interference using Kalman-based filters”, Sci. Iran., 28(6), pp. 3352-3362 (2019).http://doi.org/10.24200/sci.2019.50136.1534.
3.Bayat, S., Pishkenari, H.N., and Salarieh, H.“Observation of stage position in a two-axis nano-positioner using hybrid Kalman filter”, Sci. Iran., Trans.B, Mech. Eng., 28(5), pp. 2628-2638 (2021).http://doi.org/10.24200/sci.2021.54441.3752.
4.Sheikhbahaei, R., Vossughi, G., and Alasty, A. “Optimaltuner selection using Kalman filter for a real-timemodular gas turbine model”, Sci. Iran., 27(2), pp. 806-818 (2020). http://doi.org/10.24200/sci.2019.5341.1218.
5.Mohamed, A.H. and Schwarz, K.P. “Adaptive KalmanFiltering for INS/GPS”, J. Geod., 73(4), pp. 193-203(1999). http://doi.org/10.1007/s001900050236.
6.Mehra, R. “Approaches to adaptive filtering”, IEEETrans. Automat. Contr., 17(5), pp. 693-698 (1972).http://doi.org/10.1109/TAC.1972.1100100.
7.Sage, A.P. and Husa, G.W. “Adaptive filtering withunknown prior statistics”, Jt. Autom. Control Conf., 7,pp. 760-769 (1969).
8.Myers, K. and Tapley, B.D. “Adaptive sequentialestimation with unknown noise statistics”, IEEE Trans.Automat. Contr., 21(4), pp. 520-523 (1976).http://doi.org/10.1109/TAC.1976.1101260.
9.Maybeck, P.S. “Stochastic models, estimation, andcontrol”, Academic press (1982).
10.Bar-Shalom, Y., Li, X.R., and Kirubarajan, T. Estimationwith Applications to Tracking and Navigation: TheoryAlgorithms and Software, John Wiley and Sons (2004).
11.Sarkka, S. and Nummenmaa, A. “Recursive noiseadaptive Kalman filtering by variational Bayesianapproximations”, IEEE Trans. Automat. Contr., 54(3), pp. 596-600 (2009). http://doi.org/10.1109/TAC.2008.2008348.
12.Ma, J., Lan, H., Wang, Z., et al. “Improved adaptiveKalman filter with unknown process noise covariance,” in2018 21st International Conference on InformationFusion (FUSION), pp. 1-5 (2018).http://doi.org/10.23919/ICIF.2018.8455394.
13.Huang, Y., Zhang, Y., Wu, et al. “A novel adaptiveKalman filter with inaccurate process and measurementnoise covariance matrices”, IEEE Trans. Automat. Contr.,63(2), pp. 594-601 (2018).http://doi.org/10.1109/TAC.2017.2730480.
14.Huang, Y., Zhang, Y., Li, N., et al. “A novel robustStudent’s t-based Kalman filter”, IEEE Trans. Aerosp.Electron. Syst., 53(3), pp. 1545-1554 (2017).http://doi.org/10.1109/TAES.2017.2651684.
15.Wang, X., Liang, Y., Pan, Q., et al. “Gaussian filter fornonlinear systems with one-step randomly delayedmeasurements”, Automatica, 49(4), pp. 976-986 (2013).http://doi.org/10.1016/j.automatica.2013.01.012.
16.Schenato, L. “Optimal estimation in networked controlsystems subject to random delay and packet drop”, IEEETrans. Automat. Contr., 53(5), pp. 1311-1317 (2008).http://doi.org/10.1109/TAC.2008.921012.
17.Shen ,B., Wang, Z., Shu ,H., et al. “H∞ filtering fornonlinear discrete-time stochastic systems with randomlyvarying sensor delays”, Automatica, 45(4), pp. 1032-1037(2009). http://doi.org/10.1016/j.automatica.2008.11.009.
18.Hermoso-Carazo, A. and Linares-Pérez, J. “Extended andunscented filtering algorithms using one-step randomlydelayed observations”, Appl. Math. Comput., 190(2), pp.1375-1393 (2007).http://doi.org/10.1016/j.amc.2007.02.016.
19. Paul, A. Kamwa, I and Joos, G. “Centralized dynamic stateestimation using a federation of extended Kalman filterswith intermittent PMU data from generator terminals”,IEEE Trans. Power Syst., 33(6), pp. 6109-6119 (2018).http://doi.org/10.1109/TPWRS.2018.2834365.
20. Ray, A, Liou, L.W. and Shen. J.H. “State estimation usingrandomly delayed measurements”, J. Dyn. Syst. Meas.Control, 115(1), pp. 19-26, (1993).http://doi.org/10.1115/1.2897399.
21. Xu, B., Wang, X., Zhang, J. et al. “Maximum correntropydelay Kalman filter for SINS/USBL integratednavigation”, ISA Trans., 117, pp. 274-287 (2021). http://doi.org/10.1016/j.isatra.2021.01.055.
22. Xu, B., Zhang. J., and Razzaqi, A. “A novel robust filterfor outliers and time-varying delay on an SINS/USBLintegrated navigation model,” Meas. Sci. Technol., 32(1),pp. 15903 (2020). http://doi.org/10.1088/1361-6501/abaae9.
23. Xu, B., Wang, X., Guo, Y., et al. “A novel adaptive filterfor cooperative localization under time-varying delay andnon-gaussian noise”, IEEE Trans. Instrum. Meas., 70, pp.1-15 (2021). http://doi.org/10.1109/TIM.2021.3119130.
24. Wang, Y., Liu, Y., Fujimoto, H., et al. “Vision-based lateralstate estimation for integrated control of automatedvehicles considering multirate and unevenly delayedmeasurements”, IEEE/ASME Trans. Mechatronics, 23(6),pp. 2619-2627 (2018).http://doi.org/10.1109/TMECH.2018.2870639.
25.Larsen, T.D., Andersen, N.A., Ravn, O., et al.“Incorporation of time delayed measurements in adiscrete-time Kalman filter”, In Proceedings of the IEEEConference on Decision and Control,4, pp. 3972–3977(1998). http://doi.org/10.1109/CDC.1998.761918.
26. Tiwari, R.K., Bhaumik, S., Date, P., et al. “Particle Filterfor Randomly Delayed Measurements with UnknownLatency Probability”, Sensors, 20(19), 5689 (2020). http://doi.org/10.3390/s20195689.
27. Hermoso-Carazo, A. and Linares-Pérez, J. “Unscentedfiltering algorithm using two-step randomly delayedobservations in nonlinear systems”, Appl. Math. Model.,33(9), pp. 3705–3717 (2009). http://doi.org/10.1016/j.apm.2008.12.008.
28.Jia, G., Zhang, Y., Bai, M., et al. “A novel robust Student’s t-based Gaussian approximate filter with one-steprandomly delayed measurements”, Signal Processing,171, 107496 (2020).http://doi.org/10.1016/j.sigpro.2020.107496.
29. Jiang, Z., Zhou, W., Chen, C., et al. “A novel robustKalman filter with adaptive estimation of the unknowntime-varying latency probability”, Signal Processing, 189,108290 (2021).http://doi.org/10.1016/j.sigpro.2021.108290.
30. Wang, Z., Huang, Y., Zhang, Y., et al. “An improvedKalman filter with adaptive estimate of latencyprobability”, IEEE Trans. Circuits Syst. II Express Briefs,67(10), pp. 2259-2263 (2020). http://doi.org/10.1109/TCSII.2019.2952090.
Scientia Iranica
Volume 32, Issue 1
Transactions on Computer Science & Engineering and Electrical Engineering
January and February 2025 Article ID:6243

Files

History

  • Receive Date: 01 December 2021
  • Revise Date: 10 June 2022
  • Accept Date: 29 August 2022

Share

How to cite

Statistics