Journal of Aeronautical Engineering

Journal of Aeronautical Engineering

Improvement of a Fault- and Disturbance-Tolerant Flight Control System Using Adaptive Neural Networks and a Higher-Order Sliding Mode Observer

Document Type : Original Article

Authors
1 PhD Student, Faculty of Aerospace Engineering, Malek Ashtar University of Technology, Tehran, Iran
2 Assistant Professor, Faculty of Aerospace Engineering, Malek Ashtar University of Technology, Tehran, Iran
3 Professor, Faculty of Aerospace Engineering, Malek Ashtar University of Technology, Tehran, Iran
10.22034/joae.2025.488363.1253
Abstract
In the field of flight control, developing systems that are resilient to faults and external disturbances is a fundamental challenge that directly contributes to the safety and stability of flight operations. This paper presents a novel strategy to enhance flight control performance in the presence of sensor faults and external disturbances. The proposed approach consists of three key stages:First, for disturbance rejection and estimation, a high-order sliding mode observer (HOSMO) is employed. This observer, in conjunction with a super-twisting controller, effectively isolates sensor noise and external disturbances from the angular velocity measurements.Second, to detect and isolate sensor faults, an adaptive neural observer is designed. This observer dynamically identifies unexpected variations and faults in the sensor data.Finally, in the third stage, a backstepping-based control framework is implemented, which utilizes the estimated fault information to apply smooth control commands for fault compensation in real time.Extensive nonlinear dynamic simulations conducted on the F-18A fighter aircraft model clearly demonstrate the superior fault-tolerant performance of the proposed control framework compared to conventional methods. The use of the HOSMO results in a 13.37% improvement in tracking accuracy and a 58.8% enhancement in estimation precision compared to the STA-based structure. Moreover, the system exhibits a high degree of adaptability under complex dynamic conditions key features that significantly improve the reliability and operational effectiveness of flight control systems, making this approach a promising candidate for real-world applications.
Keywords

Subjects


Volume 27, Issue 2
November 2025
Pages 154-177

  • Receive Date 12 November 2024
  • Revise Date 04 June 2025
  • Accept Date 09 July 2025