Neuroadaptive Fault-Tolerant Control of Quadrotor UAVs: A More Affordable Solution

• Published in: IEEE transaction on neural networks and learning systems Vol. 30; no. 7; pp. 1975 - 1983
• Main Authors: Song, Yongduan, He, Liu, Zhang, Dong, Qian, Jiye, Fu, Jin
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.07.2019; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Actuation; Actuation faults; Actuators; Adaptive algorithms; Adaptive control; Aerodynamics; Artificial neural networks; Attitude control; Basis functions; Computer simulation; Failure analysis; Fault tolerance; fault-tolerant control (FTC); Heuristic algorithms; indirect neuroadaptive; Information systems; Mathematical models; Neural networks; Nonlinear systems; Parameter estimation; Radial basis function; Robustness (mathematics); Rotors; Tracking control; Uncertainty; unmanned aerial vehicle (UAV); Unmanned aerial vehicles; Unmanned helicopters; Vehicle dynamics; Virtual networks
• ISSN: 2162-237X; 2162-2388; 2162-2388
• DOI: 10.1109/TNNLS.2018.2876130

This paper investigates the position and attitude tracking control problem of a quadrotor unmanned aerial vehicle subject to modeling uncertainties and actuator failures. A comprehensive mathematical model reflecting the nonlinearity and state-space coupling of the dynamics as well as actuation faults and external disturbances is derived. By combining the radial basis function neural networks (NNs) with virtual parameter estimating algorithms, an indirect NN-based adaptive fault-tolerant control scheme is developed, which exhibits several attractive features as compared with most existing methods: 1) it is not only robust and adaptive to nonparametric uncertainties but also tolerant to unexpected actuation faults; 2) it ensures stable tracking without the need for precise information on system model; and 3) it only involves one lumped parameter adaptation, thus is structurally simpler and computationally less expensive, rendering the resultant scheme less demanding in programming and more affordable for onboard implementation. The effectiveness and benefits of the proposed method are confirmed via computer simulation.