PPAC-Pilot: Prescribed-performance augmented control for fixed-wing autopilots

Abstract

This study introduces a Prescribed-Performance Augmented Control (PPAC) framework designed for fixed-wing Unmanned Aerial Vehicle (UAV) autopilots. The PPAC strategy aims to enhance, rather than replace, existing PID control loops in open-source autopilots. Although traditional autopilots effectively manage routine tasks in most applications, their reliance on meticulous tuning remains a limitation. To address this, PPAC leverages historical flight data, a frequently overlooked resource, to derive dynamic linearization models and control laws without requiring explicit UAV models. The PPAC framework is then integrated with the Total Energy Control System (TECS) for practical deployment in takeoff and cruising scenarios. Comprehensive numerical simulations and Hardware-in-the-Loop (HIL) tests validate the strategy by comparing baseline autopilot performance with PPAC-augmented systems. Results confirm that PPAC ensures prescribed performance bounds for altitude tracking errors across evaluated scenarios, demonstrating its effectiveness in augmenting autopilots with minimized redesign efforts.