Data-driven fault-tolerant control for unmanned aerial vehicles without using identification model

Abstract

Unmanned aerial vehicle (UAV)’s fault-control problem was studied in this paper, and data-driven fault-tolerant control scheme was developed for acceleration tracking control of UAV in order to cope with the uncertainties induced by aerodynamic damage. A linear UAV dynamic model was given with reasonable assumptions, and the acceleration tracking control for UAV was converted to solving an infinite-horizon optimal control problem. The augmented algebraic Riccati equation (ARE) is derived, and its solution stability is proved based on Lyapunov theory. The data-driven control algorithm is further derived for online solving of the augmented ARE with only using flight data. The proposed algorithm is based on experience replay of flight data rather than model knowledge, so it greatly reduces the effect of uncertainties induced by aerodynamic damage on the flight control system for UAVs. Finally, the effectiveness of developed algorithm is verified through the numerical simulations under different uncertainties induced by aerodynamic damage.

Koopman fault‐tolerant model predictive control

“…Although the work in [15] addresses non-linear FTC, it still requires training the meta-model using different types of faulty data during the training phase, which is not viable in many real-world cases. Moreover, the methods presented in [18,19] need a separate FDI unit as a prerequisite to perform FTC. Besides, the works in [20,21] require precise estimation and identification of the fault, hence, a separate fault estimator unit, in addition to the FDI unit, is required to address the FTC problem.…”

“…In [18], a mechanic and electric FTC methodology are introduced, aiming to reduce the effects of a wind turbine's broken rotor bar. The authors in [19] proposed an FTC approach based on data analysis for the acceleration tracking control of unmanned aerial vehicles to handle the uncertainties caused by aerodynamic damages. Moreover, in [20], authors focused on designing a data‐driven FTC controller for discrete‐time systems by considering control input information in a moving time window in the data model to obtain the FTC control law.…”

“…Moreover, despite some recent data‐driven FTC methods such as the ones in [16] and [17], this works assumes no constraint regarding the linearity or non‐linearity of the underlying dynamics and is applicable to the general class of systems. In addition, unlike in [18] and [19], there is no requirement for a separate FDI unit. As our method operates online, it will automatically update the controller to compensate for the effects of any faults as soon as they occur.…”