Abstract
This paper investigates a tracking control problem of a flexible air-breathing hypersonic vehicle (FAHV), where prescribed performance, actuator fault, and aerodynamic uncertainty are considered. Based on the backgrounds of FAHV and for designing controller simply, an uncertain control-oriented model is reasonably decomposed into velocity subsystem and altitude subsystem. Then, the dynamic inversion controller and robust adaptive back-stepping controller are, respectively, designed for each subsystem. At the level of control design, the upper bound of the lumped uncertainty is not to be known in advance, and the uncertainty is handled by the adaptive technique that is also utilized to deal with actuator fault. A novel first-order filter is designed to solve the “explosion of terms” problem inherent in back-stepping control. Third, the prescribed performance on tracking error that features the transient performance constraint on the tracking error is resolved by transforming the constrained problem into an unconstrained problem. Furthermore, the detailed stability analysis of the closed-loop system is carried out within the framework of Lyapunov theory, in which the tracking error converges to an arbitrarily small neighborhood around zero with the prescribed performance. Finally, compared simulation result illustrates the property of the designed control strategy in tackling prescribed performance, actuator fault, and aerodynamic uncertainty.
Introduction
Hypersonic vehicles present a formidable challenge for control engineering. Their long, slender, and lightweight designs cause them to behave as flexible structures rather than rigid bodies. These flexible "modes" can interfere with the vehicle's aerodynamics and propulsion, leading to potential instabilities or structural failure. Furthermore, these vehicles operate in extreme environments characterized by unpredictable aerodynamic parameters (uncertainty) and high temperatures that can cause control surfaces, such as elevators, to lose effectiveness (actuator faults).
Existing controllers often require manual tuning or prior knowledge of the "worst-case" uncertainty. This research addresses the challenge of maintaining "prescribed performance"—ensuring tracking errors stay within specific bounds for both speed and smoothness—without requiring prior knowledge of uncertainty limits.
Control Methodology
The proposed strategy employs an Adaptive Prescribed Performance Fault Tolerant Control scheme. The complex vehicle model is first decomposed into velocity and altitude subsystems to simplify controller design.
Prescribed Performance Control (PPC)
Rather than merely aiming for zero error, the researchers utilize a performance function that defines a "funnel" within which the error must remain. A logarithmic transformation is applied to convert this constrained problem into an unconstrained one, facilitating a more straightforward mathematical solution.
Robust Adaptive Back-stepping
For the altitude subsystem, a back-stepping approach is used. This recursive design integrates adaptive laws that dynamically "learn" the unknown bounds of aerodynamic uncertainties and actuator faults. Actuator loss-of-effectiveness is treated as a type of time-varying uncertainty, allowing a unified adaptive mechanism to handle both physical modeling errors and hardware failures.
Mitigation of "Explosion of Terms"
A common drawback of back-stepping is the "explosion of terms," where calculating derivatives of virtual control laws becomes computationally prohibitive. This paper introduces a novel robust first-order filter to estimate these derivatives, significantly simplifying the controller implementation compared to traditional second-order filters.
Simulation and Results
The controller was validated using a standard Curve-Fitted Model (CFM) of a flexible air-breathing hypersonic vehicle, incorporating five rigid-body equations and three flexible-mode equations. To test robustness, a 20% uncertainty was added to nominal aerodynamic parameters.
The simulation evaluated three scenarios:
- Baseline: Aerodynamic uncertainty with Prescribed Performance.
- Fault Test: Uncertainty and Actuator Faults (without PPC).
- Comprehensive Test: Uncertainty, Actuator Faults, and Prescribed Performance.
Key Findings
- Robustness: The system maintained stable flight despite 20% parameter uncertainty.
- Fault Tolerance: Under significant actuator effectiveness loss, tracking errors remained within prescribed bounds. For instance, velocity error was maintained within a narrow band around 100 ft/s during climb maneuvers.
- Precision: Velocity and altitude settling times were approximately 65–80 seconds and 70 seconds, respectively, even under fault conditions. In contrast, the absence of the PPC mechanism led to significantly larger tracking errors.
Conclusion
The proposed adaptive control strategy effectively manages the complexities of FAHVs, including structural flexibility, aerodynamic uncertainty, and actuator faults. By transforming constrained performance requirements and utilizing dynamic adaptive laws, the controller ensures precise tracking without requiring prior knowledge of disturbance bounds.
Future work will address limitations such as the constant mass assumption—accounting for rapid fuel consumption in scramjets—and extending the control logic beyond the cruise phase to include takeoff, landing, and re-entry maneuvers. Additionally, the impact of real-world factors like sensor noise and communication delays on high-frequency adaptive laws warrants further investigation.
Summary updated