汽车工程

汽车工程 ›› 2025, Vol. 47 ›› Issue (8): 1607-1615.doi: 10.19562/j.chinasae.qcgc.2025.08.016

• • 上一篇    

基于改进无延迟频域算法的汽车风噪主动控制研究

冀慧俊1,2,卢炽华1,2,陈弯1,2(),刘志恩1,2,王颖1,2,王永亮3,孙孟雷3   

  1. 1.武汉理工大学，现代汽车零部件技术湖北省重点实验室，武汉 430070
    2.汽车零部件技术湖北省协同创新中心，武汉 430070
    3.东风汽车集团有限公司研发总院，武汉 430056
  • 收稿日期:2024-11-26 修回日期:2025-01-07 出版日期:2025-08-25 发布日期:2025-08-18
  • 通讯作者: 陈弯 E-mail:wch@whut.edu.cn
  • 基金资助:
    国家自然科学基金（52405127）和湖北省自然科学基金（2024AFB177）资助。

Research on Active Control of Automotive Wind Noise Based on Improved Delayless Frequency Domain Algorithm

Huijun Ji1,2,Chihua Lu1,2,Wan Chen1,2(),Zhien Liu1,2,Ying Wang1,2,Yongliang Wang3,Menglei Sun3   

  1. 1.Wuhan University of Technology，Hubei Key Laboratory of Advanced Technology for Automotive Components，Wuhan 430070
    2.Hubei Collaborative Innovation Center for Automotive Components Technology，Wuhan 430070
    3.Dongfeng Motor Group Co. ，Ltd. ，R&D Institute，Wuhan 430056
  • Received:2024-11-26 Revised:2025-01-07 Online:2025-08-25 Published:2025-08-18
  • Contact: Wan Chen E-mail:wch@whut.edu.cn

RichHTML

6

PDF (PC)

42

赞

摘要：

风噪是新能源汽车高速行驶车内噪声的主要来源之一，采用传统结构设计进行风噪控制存在控制难度大、降噪效果不明显等问题。本文基于次级通路均衡和频域分段变步长方法，提出了一种改进的无延迟频域滤波主动降噪算法。该算法相对于传统的滤波最小均方（FxLMS）算法具有计算复杂度低、系统降噪性能对频率的依赖性小的特点，能够在较宽频率范围内实现好的降噪效果。在SIMULINK中搭建了该算法双通道仿真模型，对宽带风噪的降噪控制效果进行了模拟，并在整车半消声室内搭建风噪主动控制试验台架进行测试验证。结果表明，基于高速工况下实车测得的车内风噪数据，本文的改进无延迟频域滤波算法的降噪效果明显优于传统FxLMS算法，在两个目标位置处可分别取得约9.42和8.81 dB（A）的降噪量。该研究成果可为汽车风噪主动控制和应用提供新思路和新方法。

关键词: 汽车风噪, 风噪主动控制, 无延迟频域算法, 次级通路均衡, 频域分段变步长

Abstract:

Wind noise is one of the main sources of interior noise in high-speed driving of new energy vehicles. Wind noise control by traditional structural design has problems such as difficulty in control and unclear noise reduction effect. In this article， an improved delayless frequency domain filtering active noise control algorithm based on secondary path equalization and frequency domain segmented variable step size method. This algorithm has the characteristics of lower computational complexity and less frequency dependence on system noise reduction performance compared to traditional filtering minimum mean square （FxLMS） algorithms， and can achieve good noise reduction effect over a wide frequency range. A dual channel simulation model of the algorithm is built in SIMULINK to simulate the noise reduction control effect of broadband wind noise， and a wind noise active control test bench is set up in a semi anechoic chamber for testing and verification. The results show that based on the wind noise data measured in real vehicles under high-speed conditions， the improved delayless frequency domain filtering algorithm proposed in this paper has significantly better noise reduction effect than the traditional FxLMS algorithm， with the noise reduction of about 9.42 dB（A） and 8.81 dB（A） at two target positions. This research result can provide new ideas and methods for active control and application of automotive wind noise.

Key words: automobile wind noise, active control of wind noise, delayless frequency domain algorithm, secondary path equalization, variable step size in frequency segmentation