Research On Active Noise Control Technology In The Cab Of Commercial Vehicles

With the increasingly stringent requirements of automotive noise regulations and the increasing sales of commercial vehicles in recent years,the noise problem in commercial vehicle cabs has gradually emerged.Effective control of the noise of the commercial vehicle cab and improvement of the NVH performance in the cab are gradually taken seriously.Traditionally,most of the noise control in commercial vehicle cab is realized by passive means such as sound absorption,sound insulation and vibration isolation.These methods are effective for dealing with high frequency noise in the cab,but they have little effect on low frequency noise.Active Noise Control(ANC)technology can effectively reduce low-frequency noise in the car.Therefore,researchs on commercial vehicle cab ANC system are significant to improve the acoustic design level of Chinese vehicles,and enhance the competitiveness of self-owned brand cars in the international markets.Based on the existing ANC algorithm,the notch two-channel filtering x-recursive least squares method(MFxRLS)and the notch two-channel affine projection method(MAP)are proposed,and the computational complexity of the two algorithms is analyzed.Based on MATLAB/simulink,the control algorithm model is built.The simulation model is used to determine the amplitude A of the reference noise,forgetting factor ? and real ? of the notch MFxRLS algorithm,and real ? of the notch MAP algorithm.The performance of these two kinds of ANC control algorithm is compared by simulation results.The development of commercial vehicle cab ANC system is divided into hardware design and software development: based on TMS320F28335 core chip,the minimum control system is designed,and supplemented by other peripheral circuits,the commercial vehicle cab ANC system hardware system is built;through the CCS software,system initialization,secondary path identification,CAN bus transceiver control,adaptive filtering algorithm and other programs are written and debugged,and ANC system software development is completed.Based on the developed ANC system,the actual layout in commercial vehicle cab is carried out.The noise signal before and after the ANC system worked is collected by the microphone.The test results show that the ANC system has obvious effect on the second-order peak noise control of the engine,and the maximum noise reduction of 10dB(A)can be obtained.A subjective evaluation model of sound quality based on GA-BP neural network is established to achieve subjective evaluation of sound quality of ANC system's effects.The sound quality evaluation model is trained by 61 sets of sound sample files.The simulation results show that the correlation degree of the subjective evaluation is as high as 95% or more.In order to verify the validity of the sound quality evaluation model,8 sets of sound sample files were selected for verification test.The results show that the maximum percentage error of the subjective scores of predicted sound quality does not exceed 6%,and most of the errors are within 1%.The sound quality evaluation model is used to analyze the noise signals before and after the ANC system worked.When the engine speed is about 2400r/min,the subjective evaluation score of the sound quality is increased from the initial 7.0413 to 7.3507;when the engine speed is about 3000r/min,the subjective evaluation score for sound quality increased from the initial 6.6203 to 6.7882.