Learning-based Adaptive Estimation for AOA Target Tracking with Non-Gaussian White Noise

In angle-of-arrival (AOA) target tracking, the target states are estimated using the noisy angular measurements. The non-Gaussian measurement noise is common in practical applications and will decrease the estimation accuracy significantly. This paper is to reduce the negative impacts of the non-Gaussian white noise (NGWN), which is modeled by multiple mixture Gaussian white noises, and guarantee the estimation accuracy. A learning-based adaptive extended Kalman filter (EKF) for AOA target tracking is developed in the NGWN environment. In the proposed method, the extreme learning machine (ELM) using a three-layer neural network is applied to identify the characters of measurement noise at each sampling time. Consequence, the EKF will consider the noise characters by the trained ELM, and make the corresponding real-time adjustments to improve the tracking performance. The effectiveness of the proposed method is verified by simulation examples.

[1]  Henry Leung,et al.  State Estimation in Unknown Non-Gaussian Measurement Noise using Variational Bayesian Technique , 2013, IEEE Transactions on Aerospace and Electronic Systems.

[2]  Nello Cristianini,et al.  An Introduction to Support Vector Machines and Other Kernel-based Learning Methods , 2000 .

[3]  E. Olson,et al.  Robust Range-Only Beacon Localization , 2004, IEEE Journal of Oceanic Engineering.

[4]  S. Kay.  Fundamentals of statistical signal processing: estimation theory , 1993 .

[5]  Hadi Sadoghi Yazdi,et al.  Kalman filtering based on the maximum correntropy criterion in the presence of non-Gaussian noise , 2016, 2016 Annual Conference on Information Science and Systems (CISS).

[6]  Sheng Xu,et al.  Optimal Sensor Placement for 3-D Angle-of-Arrival Target Localization , 2017, IEEE Transactions on Aerospace and Electronic Systems.

[7]  Y. Oshman,et al.  Optimization of observer trajectories for bearings-only target localization , 1999 .

[8]  Neil J. Gordon,et al.  A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking , 2002, IEEE Trans. Signal Process..

[9]  Kutluyil Dogançay,et al.  3D Pseudolinear Target Motion Analysis From Angle Measurements , 2015, IEEE Transactions on Signal Processing.

[10]  Yaakov Bar-Shalom,et al.  Multitarget-Multisensor Tracking: Applications and Advances , 1992 .

[11]  Kutluyil Dogançay,et al.  Distributed pseudolinear estimation and UAV path optimization for 3D AOA target tracking , 2017, Signal Process..

[12]  Dan Simon,et al.  Kalman filtering based on the maximum correntropy criterion in the presence of non-Gaussian noise , 2016 .

[13]  H. Leung,et al.  Maximum Likelihood State Estimation of Semi-Markovian Switching System in Non-Gaussian Measurement Noise , 2010, IEEE Transactions on Aerospace and Electronic Systems.

[14]  Kutluyil Dogançay,et al.  3D pseudolinear Kalman filter with own-ship path optimization for AOA target tracking , 2016, 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[15]  Mark R. Morelande,et al.  Sensor scheduling for multiple target tracking and detection using passive measurements , 2008, 2008 11th International Conference on Information Fusion.

[16]  Kumpati S. Narendra,et al.  Identification and control of dynamical systems using neural networks , 1990, IEEE Trans. Neural Networks.

[17]  H. Sorenson,et al.  Nonlinear Bayesian estimation using Gaussian sum approximations , 1972 .

[18]  J. Tabrikian,et al.  MMSE-Based Filtering in Presence of Non-Gaussian System and Measurement Noise , 2010, IEEE Transactions on Aerospace and Electronic Systems.

[19]  Peng Shi,et al.  Robust Kalman Filters Based on Gaussian Scale Mixture Distributions With Application to Target Tracking , 2019, IEEE Transactions on Systems, Man, and Cybernetics: Systems.

[20]  Yaakov Bar-Shalom,et al.  An interacting multiple model approach for target tracking with glint noise , 1995 .

[21]  Hugh F. Durrant-Whyte,et al.  A new method for the nonlinear transformation of means and covariances in filters and estimators , 2000, IEEE Trans. Autom. Control..

[22]  Chee Kheong Siew,et al.  Extreme learning machine: Theory and applications , 2006, Neurocomputing.

[23]  Jeffrey K. Uhlmann,et al.  Corrections to "Unscented Filtering and Nonlinear Estimation" , 2004, Proc. IEEE.

[24]  Xinyu Wu,et al.  Robot trajectory tracking control using learning from demonstration method , 2019, Neurocomputing.

[25]  F. Gustafsson,et al.  Mobile positioning using wireless networks: possibilities and fundamental limitations based on available wireless network measurements , 2005, IEEE Signal Processing Magazine.

[26]  C. Masreliez.  Approximate non-Gaussian filtering with linear state and observation relations , 1975 .

[27]  A. Ali,et al.  A comparative study of target tracking with Kalman filter, extended Kalman filter and particle filter using received signal strength measurements , 2015, 2015 International Conference on Emerging Technologies (ICET).

[28]  Y. Bar-Shalom,et al.  An interacting multiple model approach for target tracking with glint noise , 1994, Proceedings of 1994 IEEE National Radar Conference.

[29]  Michael A. Goodrich,et al.  Supporting wilderness search and rescue using a camera‐equipped mini UAV , 2008, J. Field Robotics.