Nonlinear adaptive control design and controller integrity monitoring for quadrotor UAVs

Abstract

This paper presents a nonlinear adaptive fault-tolerant control algorithm for accommodating actuator faults in quadrotor UAVs and an online controller integrity method for assuring safety of adaptive control system in the presence of controller software faults. The actuator faults considered in this paper are modeled as partial losses of effectiveness in one of the rotors. Using adaptive backstepping techniques, a nonlinear indirect adaptive controller is designed for the quadrotor altitude and attitude system. The adaptive control algorithm guarantees closed-loop system stability and asymptotic tracking performance even in the presence of actuator faults. However, the occurrence of faults in the embedded adaptive control software may lead to unstable adaptation behaviors and malfunctions of the control algorithm. Based on Lyapunov stability criterion, an online controller integrity monitoring method is developed to detect such controller software faults. Adaptive thresholds for software fault detection are derived, ensuring the robustness with respect to fault parameter approximation error. The proposed adaptive controller and the controller integrity monitoring algorithm are implemented and evaluated using a real-time indoor quadrotor test environment.