Adaptive Practical Prescribed-Time Control Framework with Performance Guarantees for Unmanned Aerial Vehicles

Abstract

This paper develops an adaptive practical prescribed-time control framework for underactuated unmanned aerial vehicles (UAVs) with unknown mass. A prescribed-time scaling function embedded in the controller gains drives the position, velocity, and attitude tracking errors into a tunable residual set within a user-specified settling time for every admissible compact set of initial conditions, while keeping the control gains bounded, in contrast to exact prescribed-time approaches. To curb barrier-function reliance, a hierarchical prescribed performance constraint constructed from soft and hard performance functions is enforced by a novel Multiplicative Integral Barrier Lyapunov (MIBL) functional, whose smooth activation law keeps activating the barrier only as the error approaches the hard bound. The unknown mass is estimated online by a σ-modified adaptive mechanism, while two second-order linear systems are employed as low-pass filters and high-gain observers, simplifying the backstepping design while attenuating high-frequency disturbances. Rigorous Lyapunov analysis establishes practical prescribed-time convergence of all closed-loop signals together with constraint satisfaction for all time, and simulations under nominal and disturbed conditions corroborate the theoretical results.