Resilient Control of Fixed-Wing UAVs Under DoS Attacks and Actuator Faults via Intelligent Prediction Compensation

This paper proposes a resilient control framework for fixed-wing unmanned aerial vehicles (FUAVs) subject to denial-of-service (DoS) attacks and actuator faults. An extended Kalman filter with a temporal convolutional network (EKF–TCN) predictor is developed to forecast missing trajectory commands during DoS intervals. A tube-based model predictive control (Tube-MPC) scheme is then employed to generate guidance commands under explicit input constraints, while a quaternion-based velocity controller ensures full-attitude flight without Euler singularities. To enhance fault tolerance, a fixed-time observer and online quadratic programming-based allocation are integrated to handle actuator efficiency loss and bias. Hardware-in-the-loop (HIL) simulation results demonstrate that the proposed approach achieves accurate trajectory tracking with bounded errors under DoS attacks and actuator faults, while keeping control inputs within limits, thus enabling safe and resilient flight.