A component-based coding-decoding approach to set-membership filtering for time-varying systems under constrained bit rate

This paper is concerned with the set-membership filtering (SMF) problem for a class of discrete time -varying systems with unknown-but-bounded noises under bit rate constraints. The communication between sensor nodes and filters is implemented through a wireless digital communication network with limited bandwidth. A bit rate constraint is first established to quantify the extent to which the network is constrained. A component-based coding-decoding procedure is proposed that enables individual decoder to decode messages from different components scattering in different physical locations. Based on this procedure, a decoded-measurement-based recursive SMF scheme with a prediction-correction structure is put forward. The desired parameters of the set-membership filter can be calculated recursively by the proposed recursive SMF scheme. Furthermore, the co-design issue of the bit rate allocation protocol and the filter gain is converted into the mixed-integer nonlinear programming problem that is solved by means of the particle swarm optimization and the recursive filtering algorithms. Finally, numerical simulations on two scenarios are conducted to validate the effectiveness of the proposed SMF approach.(c) 2023 Elsevier Ltd. All rights reserved.

Automated summary

This paper addresses the set-membership filtering problem for discrete time-varying systems under bit rate constraints and unknown-but-bounded noises. The communication between sensor nodes and filters is implemented through a wireless digital communication network with limited bandwidth. A component-based coding-decoding procedure is proposed to decode messages from different components scattered in different physical locations. A recursive set-membership filtering scheme is developed, and the co-design issue of the bit rate allocation protocol and the filter gain is solved using particle swarm optimization and recursive filtering algorithms. Numerical simulations validate the effectiveness of the proposed approach.