Active Control Of Road Noise Inside Electric Vehicle

Electric vehicles use electric motors that have no engine noise of conventional internal combustion engine vehicles,which makes the structural-borne road noise become very obvious at low-medium speed.This noise has the characteristics of low frequency,time-varying,broadband and random.Such low frequency noise is difficult to be controlled though conventional passive noise control(PNC)method,use active noise control(ANC)method to reduce the low frequency noise inside vehicles becomes the focus of research in the industry.One of the key issues of active road noise cancellation(RNC)is to select the high coherence reference signals to improve the noise reduction of system,the other is to improve system robustness.Aiming at these issues,multiple coherence analysis and operational transfer path analysis(OTPA)are both used to select high coherence reference signals,the leaky-NFx LMS(Normalized Filter-x Least Mean Square,NFx LMS)algorithm is also used to improve the system robustness.The simulation and road test are carried out on an electric vehicle RNC system.The results show that the proposed method can improve system robustness and increase system noise reduction.Firstly,in order to select the reference signals with high coherence for RNC system,the multiple coherence analysis and operational transfer path analysis are applied respectively.For multiple coherence analysis method,truncated singular value decomposition(TSVD)is used to improve the accuracy of calculation,and genetic algorithm is used to optimize the multi-coherence calculation results.For OTPA method,the TSVD is used to overcome the influence of noise,Multi-reference TPA is used to decouple the multiple reference issues.and the position of the component with the highest contribution is chosen for reference locations.The performance of the two mentioned methods is compared through simulation and road test,and the sensor placement scheme of batch production vehicle is finally determined.Then,aiming at the time-varying and stochastic characteristics of road noise,leakyNFx LMS algorithm is used to improve the robustness of RNC system.The multi-channel RNC system based on Fx LMS(Filter-x Least Mean Square,Fx LMS)algorithm and leaky-NFx LMS algorithm are established use MATLAB/ Simulink environment.By collect the real vehicle chassis vibration signal,passenger ear noise signal and secondary path information,the simulation analysis is carried out,and the parameters in the simulation model are optimized.Under the drive condition of steady-state and unsteadystate,the traditional Fx LMS algorithm and the leaky NFx LMS algorithm are compared.The results show that the two algorithms have similar control effects under steady-state working conditions.While the leaky-NFx LMS algorithm has better robustness under unsteady-state condition.Finally,to further verify the effectiveness of the reference signal and the robustness of the system,the road test of RNC system is carried out in ADSP 21489 hardware platform.The results show that the proposed reference signal selection method can improve the noise reduction of RNC system,and the leaky-NFx LMS algorithm can enhance the system robustness.In the experiment,A-weighted sound pressure level attenuations of 3-5 d B(A)were measured for frequency range up to about50-500 Hz,the maximum peak noise reduction reaches 18.95 dB(A).