Protocol-Based Unscented Kalman Filtering in the Presence of Stochastic Uncertainties

• Published in: IEEE transactions on automatic control Vol. 65; no. 3; pp. 1303 - 1309
• Main Authors: Liu, Shuai, Wang, Zidong, Chen, Yun, Wei, Guoliang
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.03.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Approximation algorithms; Communication protocols; Computer simulation; Data transmission; Filter design (mathematics); Kalman filtering (KF); Kalman filters; Nonlinear systems; Protocols; Stochastic processes; stochastic uncertainties; Time-varying systems; Uncertainty; unscented transformation (UT)
• ISSN: 0018-9286; 1558-2523
• DOI: 10.1109/TAC.2019.2929817

In this paper, the unscented Kalman filtering (UKF) problem is investigated for a class of general nonlinear systems with stochastic uncertainties under communication protocols. A modified unscented transformation is put forward to account for stochastic uncertainties caused by modeling errors. For preventing data collisions and mitigating communication burden, the round-robin protocol and the weighted try-once-discard protocol are, respectively, introduced to regulate the data transmission order from sensors to the filter. Then, by employing two kinds of data-holding strategies (i.e., zero-order holder and zero input) for those nodes without transmission privilege, two novel protocol-based measurement models are formulated. Subsequently, by resorting to the sigma point approximation method, two resource-saving UKF algorithms are developed, where the impact from the underlying protocols on the filter design is explicitly quantified. Finally, compared with the protocol-based extended Kalman filtering algorithms, a simulation example is presented to demonstrate the effectiveness of the proposed protocol-based UKF algorithms.