Quantized Observer‐Based Distributed Fault‐Tolerant Control for Discrete Multi‐Agent Systems With Limited Bandwidth

ABSTRACT This paper addresses the problem of fault‐tolerant control in discrete multi‐agent systems affected by actuator faults, sensor faults, and measurement noises with or without bandwidth limitation. Without bandwidth limitation, a distributed unknown input observer based on adjustable parameters is first developed to estimate augmented information (states, faults, and noises). Subsequently, leveraging the estimated information, a distributed fault‐tolerant controller is designed to achieve state‐bounded consensus within the finite horizon. The proposed dual‐distributed control strategy enhances the information‐sharing capability within the cluster, thereby augmenting both the local observation accuracy and the overall tolerance. With limited bandwidth, a novel dynamic quantization strategy, incorporating an encoding and decoding mechanism, is devised to efficiently compress the transmitted data. Afterwards, the compressed data is utilized to design a quantized observer for output estimation, and a fault‐tolerant controller for the average consensus achievement. Based on balancing data accuracy and bandwidth utilization, the proposed quantization scheme employs the time‐varying quantization factor to address the challenge of synchronous quantization while offering enhanced data protection. The simulation outcomes confirm the validity of the proposed algorithms under respective environmental conditions.