This content is not included in your SAE MOBILUS subscription, or you are not logged in.
Time Domain Full Vehicle Interior Noise Calculation from Component Level Data by Machine Learning
ISSN: 0148-7191, e-ISSN: 2688-3627
To be published on September 30, 2020 by SAE International in United States
Event: 11th International Styrian Noise, Vibration & Harshness Congress: The European Automotive Noise Conference
Computational models directly derived from data gained increased interest in recent years. Data-driven approaches have brought breakthroughs in different research areas such as image-, video- and audio-processing. Often denoted as Machine Learning (ML), today these approaches are not widely applied in the field of vehicle Noise, Vibration and Harshness (NVH). Works combining ML and NVH mainly discuss the topic with respect to psychoacoustics, traffic noise, structural health monitoring and as improvement to existing numerical simulation methods. Vehicle interior noise is a major quality criterion for today’s automotive customers. To estimate noise levels early in the development process, deterministic system descriptions are created by utilizing time-consuming measurement techniques. This paper examines whether pattern-recognizing algorithms are suitable to conduct the prediction process for a steering system. Starting from operational measurements, a procedure to calculate the sound pressure level in the passenger cabin is developed and investigated. Component time domain data serves as basis for the computation. The important inputs are determined by a correlation analysis. Input selection is followed by data reduction. After preprocessing, a supervised learning environment is established. By performing a multivariate regression, the noise transfer between the component acceleration levels and the passenger seat sound pressure level is approximated. Artificial Neural Networks (ANNs) are used to fit the function among the parameters. Different ANN architectures are trained by backpropagating errors. Investigations range from 29 to 275969 trainable network parameters. The machine learned system descriptions are then used to predict the full vehicle system response. The work concludes by evaluating the ANN performances in comparison with an additionally measured validation dataset.