Research And Implementation Of Adaptive Equalization Algorithm For Automotive Audio

With the popularity of cars,people have higher requirements for high-quality automotive audio,and automotive audio equalizers have gradually become a research hotspot.Since different automotive models have different seat materials and speaker configurations,resulting in different distortion levels of sound field of speakers,it is necessary to adjust the audio equalizer parameters for the sound field environment of a specific automotive model.However,the above process usually relies on manual adjustment by the tuner,which is not only inefficient but also costly.In addition,in order to solve the problem of limited computational resources of automotive audio equalizers,fewer filter resources need to be used.To address the above problems,this paper focuses on the design method of low-order audio equalizer to adaptively correct the sound field of automotive loudspeakers.The details are as follows:（1）For the problem of low efficiency,high cost and low accuracy of manual adjustment of the automotive audio equalizer parameters,an adaptive gain parallel graphic equalizer design method is proposed in the paper.Firstly,the effective frequency response range of the automotive loudspeaker is identified and the target gains of the designed filters are calculated adaptively.Then an adaptive gain calculation-high accuracy parallel graphic equalizer is designed based on this.Finally,the proposed method is tested and verified on MATLAB platform with the collected measurement data.The experimental results show that the number of band filters used in this method is reduced by about 20% on average compared with the classical methods,and the average maximum gain error of the fitted target gain is smaller and the average corrected spectral flatness measure is higher.（2）Limited by the cost and hardware resources,the number of equalization filters is desired to be as small as possible while ensuring the correction effect.To achieve this goal,the paper proposes an adaptive low-order parallel graphic equalizer design method.It contains the dynamic selection of filter target gain and adaptive adjustment of filter bandwidth,which can reduce the number of equalization filters without losing correction accuracy.The proposed method is proved and tested by the collecting data.The experimental results show that the correction effect of this method is comparable to the classical methods,and the number of band filters used is reduced by about 25.2% on average.（3）The graphic equalizer can only be adjusted at a limited number of fixed frequency points,which is limited by the fixed center frequency and bandwidth,which is less flexible enough to meet the user’s demand for high sound quality.To address this problem,the paper proposes an automatically adjustable multi-band parametric equalizer design method,which calculates the optimal parametric equalization filter bank coefficients based on the actual automotive loudspeaker audio signals collected by microphones.The effectiveness of the method is verified by the actual measured frequency response curves of the automotive loudspeakers.It can be seen that for the woofer,its average equalized spectral flatness measure is 0.986 by using 5 peak-notch filters,and for the full-range loudspeaker,its average equalized spectral flatness measure is 0.984 by using 20 peak-notch filters,i.e.,the method is more effective in correcting the sound field of the automotive loudspeaker.