Adaptive Observer‐Based Event‐Triggered Fault‐Tolerant Control of Multiagent Systems Subject to Output Quantization, Input Nonlinearities, and Unknown Actuator and Sensor Faults

• Published in: Mathematical methods in the applied sciences
• Main Authors: Beigizadeh, Maryam, Shahrokhi, Mohammad, Malek, Sayyed Alireza
• Format: Journal Article
• Language: English
• Published: 14.08.2025
• ISSN: 0170-4214; 1099-1476
• DOI: 10.1002/mma.70048

This paper proposes a novel distributed adaptive event‐triggered (ET) control strategy for achieving consensus in leader–follower multiagent systems (MASs) with unknown nonlinear dynamics in strict‐feedback form and external disturbances. To reduce the communication burden in MASs, a switching threshold–based ET controller and an output quantizer with hysteresis, logarithmic, and uniform properties are employed. This combination enables discrete communication both from controllers to actuators and from sensors to controllers, leading to notable reductions in communication load. Because actuator dynamics can include unknown and rate‐dependent nonlinear behavior, a first‐order filter preceding each actuator is considered. This filter protects the actuator against the ET controller signal and prolongs the actuator's operational lifespan. Moreover, sensor and actuator faults are likely to occur in MASs, which can significantly impact the performance of the control system. Therefore, it is crucial to account for these faults in the controller design process. To estimate the unmeasured state variables, a novel linear state observer is proposed that utilizes the quantized faulty output signals. The stability of the closed‐loop system is established via the Lyapunov theory and backstepping approach, incorporating an additional step due to the existence of a filter. To avoid the so‐called “explosion of complexity” in this methodology, neural networks are employed to approximate virtual control derivatives. It is shown that the Zeno behavior is avoided, and by selecting the design parameters appropriately, the tracking errors can be minimized. Finally, the satisfactory performance of the proposed controller is illustrated through two simulation examples.