Deep learning-based wind noise prediction study for automotive clay model

Abstract

Analyzing and mitigating wind noise in automobiles under high-speed conditions is a significant issue within the realm of Noise, Vibration, and Harshness (NVH). Due to the intricate nature of aeroacoustics generation mechanisms, current conventional methods for wind noise prediction have limitations. Hence, deep learning methods are introduced to investigate wind noise in the side window area of an automotive clay model.During aeroacoustic wind tunnel experiments, side window vibration data and noise data from the driver were collected under vehicle speed conditions of 100 km/h, 120 km/h, and 140 km/h, respectively. These data samples were obtained to be used for training and validation of the wind noise model. Convolutional Neural Networks (CNN) and Long Short-Term Memory Neural Network (LSTM) algorithms were separately employed to reveal the complex nonlinear relationship between wind noise and its influencing factors, leading to the establishment of a wind noise prediction model.Simultaneously, these two deep learning methods were compared with Backpropagation Neural Networks (BPNN), Extreme Learning Machines (ELM), and Support Vector Regression (SVR) methods. Our findings revealed that the LSTM wind noise prediction model not only exhibits higher accuracy but also demonstrates superior generalization capabilities, thereby substantiating the superiority of this method.