摘要
传统降噪方法对高品质车型降噪效果有限,导致乘用车对主动降噪控制(ANC)产生迫切需求,但传统前馈式ANC对随机非平稳、非线性信号存在时滞性的技术瓶颈,从而优化与应用ANC已成为降低车内噪声的必要手段。本文结合非负Tucker3分解(NTD)算法理论,提出非负约束型ANC(NCANC)的降噪方法。通过对多通道式NCANC的输入端进行非负约束使子特征信号稀疏化,在进入CPU前完成分块处理,以使处理器可进行并行计算,提高处理效率;同时,次级通道在频率内增加误差补偿和反馈,降低振幅能量,最终起到降噪效果。仿真和工程试验表明:在随机信号处理时NCANC降低近84.4%计算内存,在3ms内达到96%以上的计算精度,而且在工程上实现93dB(A)以上的相关性,最终降低5.21 dB(A)的声压级,提高车内声品质水平。
Traditional noise reduction methods have limited effects on noise reduction of high quality vehicles,so passenger vehicles have an urgent need for active noise control(ANC).However,the traditional feedforward mode ANC has a technical bottleneck for stochastic non-stationary and nonlinear signals processing.The optimization and application of ANC have become the necessary means to reduce vehicle interior noise.Based on the theory of non-negative Tucker3 decomposition(NTD)algorithm,a non-negative constraint ANC(NCANC)denoising method is proposed in this paper.By adding the non-negative constraints to the input of multi-channel NCNAC,the sub-feature signal is scatted and partitioned before entering the CPU,so that the processor can perform a parallel computation and improve processing efficiency.At the same time,the error compensation and feedback in frequency is increased and the amplitude energy is reduced in the secondary channel so that a good noise reduction effect is achieved.Simulation and engineering test show that the NCNAC can save 84%computational memory for random signals processing and achieve 96%computational accuracy in 3 ms.It also achieves 93 dB correlation in engineering.Finally,the sound pressure level is reduced by 5.21 dB(A),and the interior sound quality is improved.
作者
赵红飞
王海军
韦宁
ZHAO Hongfei;WANG Haijun;WEI Ning(SAIC–GM-Wuling Automotive Co.,Ltd.,Liuzhou 545007,Guangxi China)
出处
《噪声与振动控制》
CSCD
2019年第6期83-88,共6页
Noise and Vibration Control
