Saturated High-Gain Observer for Quadrotor Trajectory Tracking Under Sampled Outputs

This paper proposes a saturated high-gain observer for a class of state-affine nonlinear systems with coupled structures under sampled-output measurements. The study is motivated by state estimation for quadrotor trajectory tracking, where output measurements are available only at large sampling intervals and are corrupted by noise. Although continuous-discrete high-gain observers are effective for reconstructing unmeasured states, large correction gains may induce the peaking phenomenon during transient stages, producing aggressive estimated signals that can affect the closed-loop response. To address this issue, the proposed observer introduces saturated correction terms that explicitly bound the transient injection signals while preserving the convergence properties of the high-gain framework. This strategy enables aggressive observer gain selection to accelerate convergence dynamics while completely suppressing transient peaking in the estimated states. The proposed observer is applied to a quadrotor trajectory-tracking problem, and numerical simulations are conducted under large sampling periods and measurement noise.