High‐degree cubature Kalman filter for nonlinear state estimation with missing measurements

Zhang, Xing; Yan, Zhibin; Chen, Yunqi

doi:10.1002/asjc.2510

Cited by 12 publications

(8 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The estimates from the VB algorithm are obtained by minimizing the KL divergence such that the factored approximation as close as the targeted (Joint distribution) as given by (24).…”

Section: Comments On Convergencementioning

confidence: 99%

“…Satisfactory navigation performance has been demonstrated for the GPS/INS-integrated system with randomly delayed measurements. In a recent publication [24], high-degree CKF for nonlinear systems with missing or delayed measurements is advocated. The possibility of multi-step delay and chances of loss of measurements cannot be ruled out, which are reported in Singh et al [25].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Adaptive Gaussian filters for nonlinear state estimation with one‐step randomly delayed measurements

Poluri

Dey

2022

Asian Journal of Control

View full text Add to dashboard Cite

This paper proposes new algorithms of adaptive Gaussian filters for nonlinear state estimation with maximum one‐step randomly delayed measurements. The unknown random delay is modeled as a Bernoulli random variable with the latency probability known a priori. However, a contingent situation has been considered in this work when the measurement noise statistics remain partially unknown. Due to unavailability of the complete knowledge of measurement noise statistics, the unknown measurement noise covariance matrix is estimated along with states following: (i) variational Bayesian approach, (ii) maximum likelihood estimation. The adaptation algorithms are mathematically derived following both of the above approaches. Subsequently, a general framework for adaptive Gaussian filter is presented with which variants of adaptive nonlinear filters can be formulated using different rules of numerical approximation for Gaussian integrals. This paper presents a few of such filters, viz., adaptive cubature Kalman filter, adaptive cubature quadrature Kalman filter with their higher degree variants, adaptive unscented Kalman filter, and adaptive Gauss–Hermite filter, and demonstrates the comparative performance analysis with the help of a nontrivial Bearing only tracking problem in simulation. Additionally, the paper carries out relative performance comparison between maximum likelihood estimation and variational Bayesian approaches for adaptation using Monte Carlo simulation. The proposed algorithms are also validated with the help of an off‐line harmonics estimation problem with real data.

show abstract

“…The estimates from the VB algorithm are obtained by minimizing the KL divergence such that the factored approximation as close as the targeted (Joint distribution) as given by (24).…”

Section: Comments On Convergencementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Adaptive Gaussian filters for nonlinear state estimation with one‐step randomly delayed measurements

Poluri

Dey

2022

Asian Journal of Control

View full text Add to dashboard Cite

show abstract

“…The literature further witnesses some preliminary contributions [12,[21][22][23][24] to handle the missing measurements independently (ignoring the presence of delay). More specifically, [12] redesigns the conventional EKF for partially missing (fractionally received) measurements.…”

Section: Introductionmentioning

confidence: 99%

Improved Gaussian filtering for handling concurrent delayed and missing measurements

Kumar

Naik

Pachori

et al. 2023

Asian Journal of Control

View full text Add to dashboard Cite

This article addresses the Gaussian filtering problem under the environment of jointly occurring delayed and missing measurements. In this work, the former irregularity is incorporated (in the measurement model) using a Bernoulli random variable (BRV) and a geometric random variable, while the latter is subsumed with the help of the BRV; thereby, it enables to take account of large delay extents efficiently. Specifically, a modified measurement model, which incorporates the concerned irregularities, is introduced. Accordingly, the measurement‐related statistical parameters, that is, measurement estimate, covariance, and cross‐covariance, are rederived with respect to the modified measurement model. The rederived parameters replace the corresponding ones in the traditional Gaussian filtering algorithm, resulting in the proposed Gaussian filter. The simulation results conclude the superior performance of the proposed filter over the existing filters in handling the coexisting delay and missing measurement irregularities.

show abstract

“…In addition, when the initial noise statistics are inaccurately predefined, the convergence of the Kalman filter will be significantly reduced. Therefore, some improved Kalman filters such as unscented Kalman filter (UKF) (Julier and Uhlmann, 1997), cubature Kalman filter (CKF) (Arasaratnam and Haykin, 2009) and various improved CKF (Chandra et al , 2013; Zhang et al , 2020; Zhang et al , 2019a; Yang et al , 2019) are proposed to deal with the nonlinear state model. Furthermore, Sage Window is integrated to dynamically improve the noise statistics, which results in a windowing-based Kalman filter (Yang and Xu, 2003) and a windowing-based UKF (Gao et al , 2015).…”

Section: Introductionmentioning

confidence: 99%

External force estimation for robot manipulator based on a LuGre-linear-hybrid friction model and an improved square root cubature Kalman filter

Wang

Chen

Zhang

et al. 2022

View full text Add to dashboard Cite

Purpose The sensorless external force estimation of robot manipulator can be helpful for reducing the cost and complexity of the robot system. However, the complex friction phenomenon of the robot joint and uncertainty of robot model and signal noise significantly decrease the estimation accuracy. This study aims to investigate the friction modeling and the noise rejection of the external force estimation. Design/methodology/approach A LuGre-linear-hybrid (LuGre-L) friction model that combines the dynamic friction characteristics of the robot joint and static friction of the drive motor is proposed to improve the modeling accuracy of robot friction. The square root cubature Kalman filter (SCKF) is improved by integrating a Sage Window outer layer and a nonlinear disturbance observer (NDOB) inner layer. In the outer layer, Sage Window is integrated in the square root Kalman filter (W-SCKF) to dynamically adjust noise statistics. NDOB is applied as the inner layer of W-SCKF (NDOB-WSCKF) to obtain the uncertain state variables of the state model. Findings A peg-in-hole contact experiment conducted on a real robot demonstrates that the average accuracy of the estimated joint torque based on LuGre-L is improved by 4.9% in contrast to the LuGre model. Based on the proposed NDOB-WSCKF, the average estimation accuracy of the external joint torque can reach up to 92.1%, which is improved by 4%–15.3% in contrast to other estimation methods (SCKF and NDOB). Originality/value A LuGre-L friction model is proposed to handle the coupling of static and dynamic friction characteristics for the robot manipulator. An improved SCKF is applied to estimate the external force of the robot manipulator. To improve the noise rejection ability of the estimation method and make it more resistant to unmodeled state variable, SCKF is improved by integrating a Sage Window and NDOB, and a NDOB-WSCKF external force estimator is developed. Validation results demonstrate that the accuracy of the robot dynamics model and the estimated external force is improved by the proposed method.

show abstract

High‐degree cubature Kalman filter for nonlinear state estimation with missing measurements

Cited by 12 publications

References 41 publications

Adaptive Gaussian filters for nonlinear state estimation with one‐step randomly delayed measurements

Adaptive Gaussian filters for nonlinear state estimation with one‐step randomly delayed measurements

Improved Gaussian filtering for handling concurrent delayed and missing measurements

External force estimation for robot manipulator based on a LuGre-linear-hybrid friction model and an improved square root cubature Kalman filter

Contact Info

Product

Resources

About