Research On The Algorithm For Active Control Of Vehicle Interior Noise Based On Broadband And Narrowband Hybrid Structure

With the development of the automobile industry and the increasingly fierce market competition,the NVH（Noise,Vibration ＆ Harshness）problem has been paid more and more attention by automobile manufacturers and scientific research institutes.The traditional passive noise control（PNC）technology can effectively suppress the vehicle interior noise in mid-and high-frequency through sound absorption,sound insulation,etc.,but its noise reduction effect on low-frequency components is minimal.Conversely,the active noise control（ANC）technology based on the principle of destructive interference of acoustic waves can effectively control low-frequency noise through acoustic silencing.Mostly,the vehicle interior lowfrequency noise is dominated by engine order noise and road booming noise,which exhibits a characteristic of a mixture of broadband and narrowband components.Note that the hybrid ANC system that combines the advantages of the broadband ANC system and the narrowband ANC system has a stronger ability for attenuating such noise.Therefore,this paper focuses on the algorithm for active control of vehicle interior noise based on the hybrid structure.The basic physical structure and adaptive filtering algorithm principle of the traditional broadband ANC system,the narrowband ANC system,and the hybrid ANC system are studied progressively,and then the following core research work is completed:To improve the performance of the broadband ANC system,a normalized weighted signal filtered-x least mean square（NWSFXLMS）algorithm is proposed in this paper.It introduces a signal weighting vector into the FXLMS algorithm to give a higher weight to the samples close to the current moment in the reference signal vector and the filtered reference signal vector,and adopts a normalization step size adjustment mechanism,leading to enhanced convergence performance and noise reduction performance.Aiming at the problems of insufficient convergence and high computational complexity of the traditional narrowband ANC system,a smoothed filtered-e least mean square（SFELMS）algorithm is proposed.In this algorithm,smooth processing is introduced for the engine speed signal,and the FELMS algorithm is used for adaptive filtering,which improves the stability,noise reduction performance,and computational efficiency of the system.Subsequently,by combining the strengths of the NWSFXLMS algorithm and the SFELMS algorithm,an improved hybrid ANC（IHANC）algorithm is proposed to enhance the ability to attenuate the noise mixed with broadband and narrowband components.To guarantee the optimal performance of the system,this paper develops a parameter matching method of ANC algorithm based on improved particle swarm optimization（IPSO）.This method utilizes the performance evaluation indicators of the ANC system to establish the objective function that considers the convergence speed and steady-state error simultaneously.It is of both higher efficiency and higher accuracy compared with the traditional trial-and-error method.To verify the superiority of the proposed algorithms,we have conducted ANC simulations of the algorithms in the MATLAB software platform using the measured interior noise of a vehicle under different constant-speed conditions.The simulation results demonstrate that,compared with the traditional ANC algorithms,the IHANC algorithm can eliminate the target order noise and the rest of the broadband noise more effectively,exhibiting a faster convergence speed and improved steady-state noise reduction.Finally,an active noise control test platform is built to verify the effectiveness of the IHANC algorithm,and noise control tests are carried out for the real vehicle interior noise collected under constant-speed conditions.The IHANC system achieves 19.3d B～23.0d B noise reduction at the second-order frequency,and reduces the total sound pressure level by 4.3d B.In general,the research work of this paper provides new ideas for the improvement and optimization of the adaptive algorithm for active control of vehicle interior noise and has a certain reference value.