Unmanned aerial vehicles, specifically quadrotor drones, are increasingly commonplace in community and workplace settings and are often used for photography, cinematography, and small parcel transport. The presence of these flying robotic systems has a substantial impact on the surrounding environment. To better understand the ergonomic impacts of quadrotor drones, a quantitative description of their acoustic signature is needed. While previous efforts have presented detailed acoustic characterizations, there is a distinct lack of high spatial-fidelity investigations of the acoustic field of a quadrotor hovering under its own power. This work presents an experimental quantification of the spatial acoustic pressure distribution in the near-field of a live hovering unmanned aerial vehicle. A large-aperture scanning microphone array was constructed to measure sound pressure level at a total of 1728 points over a 2 m × 3 m × 1.5 m volume. A physics-infused machine learning model was fit to the data to better visualize and understand the experimental results. The experimental data and modeling presented in this work are intended to inform future design of experiments for quadrotor drone acoustics, provide quantitative information on the acoustic near-field signature, and demonstrate the utility of optical motion tracking coupled with a custom microphone array for characterization of live acoustic sources.
Physics-Infused Machine Learning (PIML) architectures aim at integrating machine learning with computationallyefficient, low-fidelity (partial) physics models, leading to improved generalizability, extrapolability, and robustness to noise, compared to pure data-driven approximation models. End-uses of PIML include, but are not limited to, model-based optimization and model-predictive control. Recently a new PIML architecture was reported by the same authors, known as Opportunistic Physics-mining Transfer Mapping Architecture or OPTMA, which transfers the original inputs into latent features using a transfer neural network; the partial physics model then uses the latent features to generate the final output that is as close as possible to the high-fidelity output. While gradient-free solvers and back-propagation with supervised learning (where optimum labels are pre-generated) have been used to train OPTMA, that approach remains computationally inefficient, and challenging to generalize across different problems or exploit state-of-theart machine learning architectures. This paper aims to alleviate these issues by infusing the partial physics model inside the neural network, as described via tensors in the popular machine learning framework, PyTorch. Such a description also naturally allows auto-differentiation of the partial physics model, thereby enabling the use of efficient back-propagation methods to train the transfer network. The benefits of the upgraded OPTMA architecture with Automatic Differentiation (OPTMA-Net) is demonstrated by applying it to the problem of modeling the sound pressure field created by a hovering unmanned aerial vehicle (UAV). Ground truth data for this problem has been obtained from an indoor UAV noise measurement setup. Here, the partial physics model is based on the interference of acoustic pressure waves generated by an arbitrary number of acoustic monopole sources. Case studies show that OPTMA-Net provides generalization performance close to, and extrapolation performance that is 4 times better than, those given by a pure data-driven model.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.