SAIC Volkswagen internal project*

  • 📍Pre-R&D group, SAIC Volkswagen Automotive Co., Ltd. 🇨🇳
  • 📅 02/2019-06/2019
  • Primary developer in charge of algorithm development&deployment, and GUI development
  • Project manager and key participant of team of 3
  • Python, Pytorch, PyQT

Abstract

The Automotive Aeroacoustic (Whistle) Noise Prediction Software (ANPS) is a cross-platform software designed to predict and analyze automotive internal aeroacoustic noise. Leveraging machine learning and data science technologies , ANPS utilizes wind tunnel test data to evaluate the acoustic impact of sealing strips and identify any potential design deficiencies related to these strips. ANPS serves as a valuable tool for early-stage development and analysis of the wind noise performance of vehicles. Its key features and advantages include:

  • It can read the basic data of the experimental vehicle, such as model, height, width, length, configuration and status.
  • It can output a relative ML predicted Articulation Index (AI) value, which indicates speech intelligibility in noisy environments;
  • A line graph can be printed to show how different experimental states affect the AI value;
  • It can help engineers optimize the vehicle design for better wind-noise performance.

software panel

* Refer to pdf (Chinese) for more details.