A robust model predictive control-based method for fault detection and fault tolerant control of quadrotor UAV

In this study, first a model predictive control (MPC) solution is utilized to control the quadrotor model in both fault-free and faulty modes by employing a cost function. For this aim, the obtained non-linear continuous model has been discretized after linearization around an equilibrium point. Then, a fault detection and diagnosis (FDD) and fault tolerant control (FTC) strategy is proposed based on MPC in order to neutralize the effect of fault. The method of FDD and FTC designing is based on analytical redundancy relations (ARRs) which detects and tolerates the fault by producing a residual signal. Therefore, this method presents a complete FTC solution to fault issue without using observer. In addition, this solution has robustness against the uncertainties and disturbances which may challenge the model in practical situations. The discussed fault in this paper is a sensor fault which is modeled as an undesired function which affects the sensor of roll angle. However, the proposed FTC can tolerate all types of faults. The results of implementation of this method are validated under various simulation scenarios.

Automated summary

In this study, a model predictive control (MPC) solution is used to control a quadrotor model, and a fault detection and diagnosis (FDD) strategy based on analytical redundancy relations (ARRs) is proposed to detect and tolerate sensor faults.