Towards a Generalized Disturbance Model for Small UAVs: Concepts, Taxonomy, and Theoretical Foundations

Abstract

Small Unmanned Aerial Vehicles (UAVs) operate within the complex atmospheric boundary layer, yet current research often relies on simplistic additive white Gaussian noise or constant wind vectors that fail to reflect real-world airflow correlations. This conceptual paper addresses this gap by establishing the theoretical foundations for a Generalized Disturbance Model (GDM) specifically tailored for micro and small-scale aerial platforms. The GDM is predicated on a multi-tier taxonomy categorizing disturbances into three domains: stochastic atmospheric turbulence (refined von Kármán and Dryden models), deterministic spatial gradients (urban canyons and forest canopies), and platform-induced internal perturbations (sensor vibration and center-of-gravity shifts). We introduce the “Disturbance Influence Matrix” (DIM), a mathematical construct that maps environmental energy states to specific force and moment vectors acting on the UAV’s body frame. This matrix accounts for the vehicle’s geometry and aerodynamic coefficients, providing a deterministic link between the environment and resulting flight path deviations. Furthermore, the framework explores “Disturbance Observability,” theorizing how onboard IMU and GPS data can be inverted to estimate the state of the surrounding atmosphere in real-time. By formalizing the mathematical structure used to describe aero-environmental interactions, this work provides a rigorous benchmarking framework for evaluating disturbance-rejection controllers and adaptive navigation algorithms. The theoretical insights aim to move the field toward “environmentally-aware” autonomy, where UAVs can actively exploit local wind patterns for energy-efficient flight in critical applications like search and rescue.