Active Noise Control System With Feedforward And Feedback Hybrid Structure In Vehicle And Its Application

With the acceleration of the modernization process,the vehicle parc in China has increased year by year,thus the noise pollution problem has become more and more serious.With the increasingly stringent laws and regulations and the increasing requirements of consumers for the ride comfort in vehicles,automobile manufacturers and research institutes have also paid more attention to the issues of automotive NVH(Noise,Vibration & Harshness).In the traditional automobile design processes,passive noise control(PNC)methods such as sound absorption and sound insulation are usually used to control the middle and high frequency noise in the vehicle cabins,but these methods are not obvious enough to suppress the low frequency noise.As a supplement to this method,active noise control(ANC)technology has a significant suppression effect on low frequency noise,and it has the characteristics of high integration,strong customization and can improve the sound quality in the vehicle cabins.Thus,the ANC technology has gradually become the first choice for low frequency noise control in the vehicle interior noise.Thus,this paper takes the low frequency narrowband noise as the object of cancellation,and focuses on the application of the feedforward and feedback hybrid structure ANC(HANC)algorithm in the vehicles.In order to verify the corresponding relationship between the engine speed and the frequency of the engine order noise,a real vehicle interior noise acquisition experiment under various driving conditions is carried out in this paper.After analyzing the frequency spectrum of the test data,it is found that the low frequency narrowband noise in the cabin not only includes the engine order noise but also contains other narrowband harmonics that are not significantly related to the engine speed,and the sound pressure level(SPL)of these harmonics is relatively close to that of the former.The commonly used notch LMS algorithm,feedback Fx LMS algorithm and the conventional HANC(CHANC)algorithm are analyzed,and the preponderances and drawbacks of the above algorithms are evaluated.Based on this,the modified HANC(MHANC)algorithm is proposed,which includes three subsystems: feedforward,feedback and error signal separation subsystem.The first two subsystems are used to control the engine order noise and other low-frequency narrowband noise in the cabin respectively,and the coupling effect between them is eliminated by the separation subsystem.For the sake of reducing the computation complexity and enhance the stability of the algorithm,the feedforward subsystem adopts the smooth notch delay LMS algorithm,while the adaptive filter in the separation subsystem is replaced by an adaptive LMS analyzer.The feedback subsystem adopts the Sfunction based modified variable step-size Fx LMS algorithm(MSVS-Fx LMS)to ensure fast convergence speed as improving the stability of the ANC system.Meanwhile,the input samples are clipped by some threshold values,which enhances the robustness of the proposed algorithm in the presences of the impulsive noise.The comparison of the computation complexity of the MHANC algorithm and the CHANC algorithm shows that the former algorithm only has a slight increase in the amount of multiplication,but the amount of addition of the above two algorithms are very close.Based on the MATLAB/Simulink platform,the simulation models of the CHANC algorithm and the MHANC algorithm are built.The synthesized noise and the sampled noise in a real vehicle cabin are used as the reference signals of the above two HANC algorithms,and corresponding simulations are carried out.The results show that the subsystems of the MHANC algorithm are more stable than that of the CHANC algorithm under the same conditions.Therefore,the convergence speed and noise attenuation effect of the proposed algorithm are better than the CHANC algorithm.Besides,the MHANC algorithm are less affected by the“waterbed effect” compared with CHANC algorithm.Finally,the real vehicle ANC tests are designed and the placement of equipments of the ANC system in the cabin is completed.The tests are carried out under three driving conditions:3800 rpm steady state(neutral gear),idle speed-3800 rpm acceleration(neutral gear)and 60km/h consistent speed driving.After analyzing the test data,it is found that the MHANC algorithm has a significant suppressive effect on the interior low frequency narrowband noise.The SPL of the second order engine noise is reduced about 30 d B(A)under two neutral driving conditions,and the SPL in the cabin is reduced by more than 3 d B(A).The ANC system also has a prominent suppression effect on the low frequency noise under 60 km/h consistent speed driving conditions,and the overall noise reduction is 1.35 d B(A).