Zonotopic distributed fusion for 2-D nonlinear systems under binary encoding schemes: An outlier-resistant approach

• Published in: Information fusion Vol. 120; p. 103103
• Main Authors: Lan, Lan, Wei, Guoliang
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.08.2025
• Subjects: 2-D nonlinear systems; Binary encoding scheme; Dynamic quantization; Outlier; Zonotopic distributed fusion; Outlier; Zonotopic distributed fusion; Binary encoding scheme; Dynamic quantization; 2-D nonlinear systems
• ISSN: 1566-2535
• DOI: 10.1016/j.inffus.2025.103103

This study addresses the zonotopic distributed fusion problem for a class of 2-D nonlinear systems over sensor networks subject to unknown-but-bounded (UBB) noises, measurement outliers, and bandwidth limitations. A saturation function is integrated to mitigate the impact of outliers. Furthermore, a dynamic quantization strategy is deployed to compact the innovation into quantized forms. An encoding-decoding mechanism is implemented between nodes and the fusion center to accommodate the constraint of limited bits while ensuring data security. The main purpose of this paper is to design a zonotopic distributed fusion algorithm under the set-membership filtering framework. Subsequently, gain parameters are derived by minimizing the F-radius of the local zonotopes, and fusion weights are ascertained using the Lagrange multiplier method. Additionally, the upper and lower bounds of the F-radius of the fused zonotope are derived. It is demonstrated that the F-radius of the fused zonotope is less than those prior to fusion. Finally, the feasibility of the developed algorithm is verified using a model of water stream heating and gas absorption. •The distributed fusion estimation problem is studied for 2-D nonlinear systems.•The system is subject to unknown-but-bounded noises and outliers.•The filter gain is derived by minimizing the F-radius of the local zonotopes.•Weighted matrices are obtained based on the Lagrange multiple method.•The proposed zonotopic distributed fusion method is illustrated through an example.