Fault-tolerant learning control of air-breathing hypersonic vehicles with uncertain parameters and actuator faults

This paper investigates fault-tolerant learning control for air-breathing hypersonic vehicles (AHVs) subject to parametric uncertainties, external disturbances, and actuator faults. By treating the flexible dynamics as equivalent disturbance, the AHV model can be decomposed into velocity subsystem and altitude subsystem. An intelligent fault diagnosis approach is deployed to estimate the actuator fault information for each subsystem sequentially. Furthermore, combined with a fixed-time neural disturbance observer, composite learning control is deployed to construct control commands using the historical stack to improve the learning accuracy, while compensating for lumped disturbances, including fault diagnosis errors and external disturbances. The designed controller addresses the fault-tolerant tracking problem by combining offline and online learning strategies, which is a significant advantage over other existing AHV controllers. Simulation results validate the effectiveness of the fault-tolerant learning control.