Real-time prediction framework for frequency-modulated multirotor noise

Abstract

A comprehensive multirotor noise assessment framework is developed to predict the noise of rotational-speed-controlled rotor configurations in real-time. The key objectives are to synthesize the frequency-modulated multirotor noise and analyze the frequency modulation (FM) characteristics. The framework includes modules associated with the flight control, aerodynamics, time reconstruction, noise prediction, and time-frequency analysis (TFA). In addition to the hybrid blade element momentum model, the aerodynamics module contains a linear inflow model, a Beddoes wake model, and an unsteady aerodynamic correction model. The convective form and source-time dominant algorithms are used in the acoustic analogy for tonal noise prediction. The FM characteristics are identified using the synchrosqueezing-based high-resolution TFA for strongly non-stationary signals. The framework is verified through validation and verification studies for diverse rotor configurations and flight conditions. During the cruise flight of the multirotor, the tonal noise exhibits simultaneous frequency and amplitude modulations. In wind gust conditions, these modulations result from rotational speed variations, acoustic wave interference, and Doppler shifting. By clarifying the non-stationary noise signal in diverse flight environments, the proposed framework can facilitate noise assessment in the perception-influenced design stage of multirotor configurations.