基于改进FX-NANC系统的车内主动噪声控制研究
Research on In-Vehicle Active Noise Control Based on the Improved FX-NANC System
刘志恩 1耿毅 1陈弯 1胡宇航 1孙孟雷2
作者信息
- 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;汽车零部件技术湖北省协同创新中心,武汉 430070
- 2. 东风汽车集团有限公司技术中心,武汉 430058
- 折叠
摘要
针对基于陷波FXLMS算法的窄带主动噪声控制(FX-NANC)系统计算量大、硬件成本高的问题,提出一种低计算量改进方案.该方案根据发动机转速信息生成余弦滤波参考信号并存入存储器中,之后通过对余弦滤波参考信号进行索引生成正弦滤波参考信号.以索引代替传统的卷积运算,以此减少系统计算量.为验证改进后系统的计算性能和降噪效果,对该系统进行数值仿真分析和实车双通道实验.结果表明:改进系统相比传统系统计算效率提升43.48％,计算速度显著提高.实车测试中稳态工况下主、副驾驶位置处的2、4、6阶噪声实现30.00 dB(A)、36.85 dB(A)、19.37 dB(A)和21.85 dB(A)、26.63 dB(A)、9.38 dB(A)的降噪量,非稳态工况两位置最大降噪量也能达到5.00 dB(A)以上,改进系统在保证原有降噪性能的前提下,具有更高的计算效率.
Abstract
In order to solve the problem of heavy computation and high hardware cost of narrowband active noise con-trol(FX-NANC)system based on notch FXLMS algorithm,this thesis proposes a low computation improvement scheme.According to the information of engine speed,the cosine filter reference signal is generated and stored in the memory,and then the sine filter reference signal is generated by indexing the cosine filter reference signal.The traditional convolution operation is replaced by index,so as to reduce the computation amount of the system.In order to verify the calculation per-formance and noise reduction effect of the improved system,numerical simulation analysis and real vehicle dual-channel experiment are carried out.The results show that compared with the traditional system,the calculation efficiency of the improved system is increased by 44.37％,and the calculation speed is significantly improved.In the real car test under steady condition,the noise of order 2,4 and 6 in the main and passenger positions can achieve noise reduction of 30.00 dB(A),36.85 dB(A),19.37 dB(A)and 21.85 dB(A),26.63 dB(A)and 9.38dB(A).The maximum noise reduc-tion of the two positions can also reach more than 5.00 dB(A)under the unsteady condition,and the improved system has higher computational efficiency under the premise of ensuring the original noise reduction performance.
关键词
窄带主动噪声控制/滤波参考信号/计算复杂度Key words
narrowband active noise control/filter the reference signal/computational complexity引用本文复制引用
基金项目
国家自然科学基金面上项目(52175111)
出版年
2024
NETL
NSTL
万方数据