Empirical models to describe UAV tonal dynamics during real operations

Abstract

Over the last decade, the aerospace industry has experienced a remarkable growth in the use of Unmanned Aerial Vehicles (UAVs) due to their low manufacturing and operational costs, adaptability, and scalability. In the context of climate change and desired emissions targets, UAVs are realistic options to complement the current freight transportation methods, and in the future, as an viable option for human mobility. UAVs are also starting to gain a place in emergency response, both in urban and rural areas, where aid or surveillance capabilities are required in short notice and a fast response is crucial. However, an important obstacle towards their broader use is the expected noise annoyance caused by UAV operations. In this study, we combine extensive outdoor measurements of drone noise with statistical modelling to predict acoustic metrics relevant to determining acoustic annoyance. Array-based measurements combined with GPS positioning information of an eVTOL fixed-wing UAV are studied under realistic operational conditions in an open field. The linear least squares theory is used to establish an empirical model and identify a set of control parameters, including acceleration and velocity, that serve as effective predictors of noise output. The results pave the way for the establishing a model that predicts drone noise for different UAV types and for different operational conditions.