Robust Quadrotor Altitude Tracking Using Sliding-Mode Variants under Wind and Thrust Uncertainties

Wind gusts, thrust-efficiency variations, actuator limits, and modeling uncertainties challenge precise altitude control for a quadrotor near hover. We derive the standard small-angle approximation to a control-ready single-input singleoutput altitude channel with matched uncertainty from the 6DOF model. On this common plant, we implement and compare three sliding-mode variants: classic sliding-mode control (SMC), adaptive SMC (ASMC), and the super-twisting algorithm (STA). Five realistic scenarios are evaluated under identical reference and thrust saturation conditions. Lyapunov analyses on Filippov solutions establish finite-time reaching for SMC and supertwisting and boundedness with convergence to the sliding manifold for adaptive SMC, within the specified disturbance bounds. Simulations quantify performance smoothness trade-offs using IAE, control energy, a chattering proxy, and saturation statistics. In all scenarios, super-twisting reduces IAE by 46 54% compared to SMC and 21 35% compared to adaptive SMC, while lowering chattering by ≈ 98% or higher and keeping saturation below 8%. Meanwhile, adaptive SMC improves transient response compared to SMC, but at the cost of increased switching during reaching.