An LMI-Based Parameter Adjustment Descriptor Observer for Sensor Fault Estimation of a Quadrotor

In this paper, simultaneous fault estimation of attitude sensors in a quadrotor using a descriptor system observer specific to Lipschitz nonlinear systems is investigated. First, dynamical model of the quadrotor is described and converted to a nonlinear state space system with Lipschitz constraint. State vector of the system is augmented using fault term of sensors as auxiliary states and descriptor system is constituted. Nonlinear asymptotic estimator is designed using linear matrix inequality method to estimate simultaneously faults and states. By constituting dynamical equation of the error and Lyapunov's stability analysis, the ultimately bounded property of the fault and state estimation error is proved. The feasibility of linear matrix inequality in deterministic estimators based on LMI is an important problem and here, by considering an adjustment parameter, more freedom is created for solving the LMI problem. The Fault of sensors which is considered here might be in any form. Finally, the attitude sensor fault estimation of the quadrotor using the proposed method is investigated throughout the simulation.