Event-triggered adaptive control of hypersonic vehicles subject to actuator nonlinearities

• Published in: Aerospace science and technology Vol. 133; p. 108119
• Main Authors: An, Hao, Wang, Yiming, Wang, Guan
• Format: Journal Article
• Language: English
• Published: Elsevier Masson SAS 01.02.2023
• Subjects: Actuator nonlinearity; Adaptive control; Event-triggered control; Hypersonic vehicle; Neural allocation; Neural allocation; Hypersonic vehicle; Event-triggered control; Adaptive control; Actuator nonlinearity
• ISSN: 1270-9638; 1626-3219
• DOI: 10.1016/j.ast.2023.108119

This paper investigates the event-triggered adaptive control problem for air-breathing hypersonic vehicles (AHVs) subject to physical actuator nonlinearities. In contrast with abundant AHV control strategies belonging to the traditional sample-data framework, the presented adaptive control works with an event-triggered mechanism, which naturally reduces the executing frequency of scramjet fuel-to-air equivalency ratio (FER) and aerodynamic control surface (ACS). During the event-triggered controller design, practical actuator characteristics including magnitude-rate composite constraints and servo transient dynamics are synthetically considered. In the velocity loop, the built-in FER saturation, which is a fusion of physical FER magnitude and rate constraints, is handled by reshaping the velocity reference with the help of an adaptive reference governor. Meanwhile in the altitude loop, a neural control allocator directly generates feasible ACS commands via a neural network that is well trained by means of the optimal control allocation. As the result, ACS magnitude-rate composite constraints and servo transient dynamics can be accommodated without implementing any time-consuming optimization searching process. Theoretical analysis indicates the guaranteed closed-loop stability and tracking performance while excluding the undesired Zeno behavior. Simulations validate the effectiveness of the proposed event-triggered adaptive control.