Active fault‐tolerant control system design with trajectory re‐planning against actuator faults and saturation: Application to a quadrotor unmanned aerial vehicle

Abstract

During the past 30 years, various fault-tolerant control (FTC) methods have been developed to address actuator or component faults for various systems with or without tracking control objectives. However, very few FTC strategies establish a relation between the post-fault reference trajectory to track and the remaining resources in the system after fault occurrence. This is an open problem that is not well considered in the literature. The main contribution of this paper is in the design of a reconfigurable FTC and trajectory planning scheme with emphasis on online decision making using differential flatness. In the fault-free case and on the basis of the available actuator resources, the reference trajectories are synthesized so as to drive the system as fast as possible to its desired setpoint without violating system constraints. In the fault case, the proposed active FTC system (AFTCS) consists in synthesizing a reconfigurable feedback control along with a modified reference trajectories once an actuator fault has been diagnosed by a fault detection and diagnosis scheme, which uses a parameter-estimation-based unscented Kalman filter. Benefited with the integration of trajectory re-planning using the flatness concept and the compensation-based reconfigurable controller, both faults and saturation in actuators can be handled effectively with the proposed AFTCS design. Advantages and limitations of the proposed AFTCS are illustrated using an experimental quadrotor unmanned aerial vehicle testbed.The reference trajectories F are determined as function of the initial conditions of the system and the final desired conditions as well as the initial time t 0 and the final time t f . The initial and final conditions and the initial time t 0 are known, and the sole unknown is the final time t f . Generally speaking, it is required to attain the desired objectives as fast as possible while respecting system When complete loss of actuators are not considered (i.e., w i < 1 or i > 0), 1 is invertible and 1 u > u and 1 u > u. Thus, for the remaining trajectories F rem .t / to travel, there is no AFTCS AND TRAJECTORY DESIGN AGAINST ACTUATOR FAULTS AND SATURATION P x f .t / D Ax f .t / C B.I m m W /u F T C .t / D Ax f .t / C B.I m m W /.u nom .t / C u add .t // D Ax f .t / C Bu nom .t / BW u nom .t / C B.I m m W /u add .t / (32)

Sliding mode fault‐tolerant control of an octorotor using linear parameter varying‐based schemes

“…However, despite this attention and the fact that some of the schemes are becoming reasonably 'mature', there are two factors which are significant obstacles to further testing of state-of-the-art schemes -especially for aerospace applications: these are cost and safety. In response to these two factors, there has been significant interest in applying state-of-the-art FTC schemes to multi-rotor UAVs -particularly to the so-called quadrotor (see for example [1], [2], [3], [4], [5]). This is largely due to their simple build, relatively low cost, and the convenience of undertaking tests in a controlled safe laboratory environment.…”

Integrated framework for constrained minimum‐time trajectory generation, fault detection and reconfiguration: A case‐study

“…The system (11) is also differentially flat, with w 2 as the flat output. The tank 1 level, w 1 and the single input, u 1w , can be described in terms of w 2 and its derivatives, and conversely, w 2 , P w 2 and R w 2 can be described in terms of the states and input.…”

Virtual actuator-based FTC for LPV systems with saturating actuators and FDI delays

“…These techniques usually lead to a redistribution of the control action among the healthy actuators, which can potentially become saturated. In order to avoid this potential actuator saturation, active FTC systems that include a trajectory replanning algorithm have been proposed [13]- [15].…”