CN-121983015-A - Noise elimination method and device, electronic equipment and vehicle

CN121983015ACN 121983015 ACN121983015 ACN 121983015ACN-121983015-A

Abstract

The application provides a noise elimination method, a device, electronic equipment and a vehicle, wherein the method comprises the steps of collecting time sequence data of environmental noise in the vehicle, analyzing the time sequence data, and determining main noise type and noise intensity spatial distribution; the method comprises the steps of activating a target noise reduction channel of a target partition in a vehicle through noise intensity spatial distribution, determining an original reference signal of a reference microphone in the target noise reduction channel through a main noise type, carrying out secondary path estimation and forward path calculation of a preset adaptive filter on the original reference signal to obtain a first noise reduction driving signal of a loudspeaker, obtaining a historical reference signal of the reference microphone and a training noise signal of an error microphone, inputting the historical reference signal and the training noise signal of the error microphone into a preset depth active noise control network to output a second noise reduction driving signal, carrying out adaptive weighting on the first noise reduction driving signal and the second noise reduction driving signal to obtain a target noise reduction driving signal, and eliminating the environmental noise of the vehicle through the target noise reduction driving signal, so that user experience is improved.

Inventors

TANG HONGYING
REN PING
LI ZHONGTONG
LI JINHUAI
WU HAIYIN

Assignees

赛力斯汽车有限公司

Dates

Publication Date: 20260505
Application Date: 20251231

Claims (10)

1. A method of noise cancellation, the method comprising: collecting time sequence data of environmental noise in the vehicle through a plurality of microphones; analyzing the time sequence data and determining the main noise type and noise intensity spatial distribution of the environmental noise; activating a target noise reduction channel of a target partition in a vehicle through the noise intensity spatial distribution, and determining an original reference signal of a reference microphone in the target noise reduction channel through the main noise type; performing secondary path estimation and forward path calculation of a preset adaptive filter on the original reference signal to obtain a first noise reduction driving signal of a loudspeaker in a target noise reduction channel; Acquiring a historical reference signal of the reference microphone and a training noise signal of an error microphone in a target noise reduction channel, inputting the historical reference signal and the training noise signal into a preset depth active noise control network, and outputting a second noise reduction driving signal; And carrying out self-adaptive weighting on the first noise reduction driving signal and the second noise reduction driving signal to obtain a target noise reduction driving signal, and completing the elimination of the environmental noise of the vehicle through the target noise reduction driving signal.
2. The method of claim 1, comprising, after the collecting of the time series data of the environmental noise in the vehicle by the plurality of microphones: Performing band-pass filtering and wavelet transformation on the time sequence data to obtain wavelet coefficients; Identifying a dominant noise type of the ambient noise from the timing data; If the main noise type is a low-frequency stable main noise type, processing the wavelet coefficient by using a soft threshold algorithm to obtain a first screening wavelet coefficient; If the main noise type is a high-frequency main noise type or a burst main noise type, processing the wavelet coefficient by using a hard threshold algorithm to obtain a second screening wavelet coefficient; performing wavelet inverse transformation on the first screening wavelet coefficient or the second screening wavelet coefficient to obtain denoising time sequence data; and carrying out normalization processing and self-adaptive gain control on the denoising time sequence data to obtain preprocessed time sequence data.
3. The method of claim 2, wherein the analyzing the time series data to determine a dominant noise type and noise intensity spatial distribution of the ambient noise comprises: Dividing the preprocessed time sequence data into continuous time domain frame data; Performing discrete Fourier transform on the time domain frame data to obtain a frequency domain complex spectrum corresponding to the time domain frame data; Determining a mel energy spectrum of the environmental noise, a time difference between arrival of the environmental noise at any two microphones and a spatial spectrum power between the environmental noise and each microphone through the frequency domain complex spectrum; Determining a mel frequency cepstrum coefficient of the environmental noise through the mel energy spectrum; determining a dominant noise type and noise intensity spatial distribution of the environmental noise through the mel energy spectrum, the mel frequency cepstrum coefficient, the time difference and the spatial spectrum power.
4. The method of claim 1, wherein said activating a target noise reduction channel of a target zone in a vehicle by said noise intensity spatial distribution further comprises: determining noise intensities of a plurality of subareas of the vehicle through the noise intensity spatial distribution; And screening out target subareas with noise intensity larger than the intensity threshold value, and activating target noise reduction channels of the target subareas in the vehicle.
5. The method of claim 1, wherein determining the original reference signal of the reference microphone within the target noise reduction channel by the primary noise type comprises: Carrying out correlation analysis on a plurality of reference microphones in the target noise reduction channel through the main noise type to obtain a correlation coefficient; Setting weights for each reference microphone in the target partition through the correlation coefficient, wherein the larger the correlation coefficient is, the larger the weights are set; and weighting time sequence data acquired by the reference microphone through weights to obtain an original reference signal of the reference microphone in the target noise reduction channel.
6. The method of claim 1, further comprising, after the performing secondary path estimation and forward path calculation of a preset adaptive filter on the original reference signal to obtain a first noise reduction driving signal for a speaker in a target noise reduction channel: Driving a loudspeaker in the noise reduction channel to generate a first reverse sound wave through the first noise reduction driving signal; Identifying a current residual noise signal received by an error microphone in a target noise reduction channel, wherein the current residual noise signal is generated after linear superposition of environmental noise and a first reverse sound wave in a sound field; And adjusting the weight of a preset adaptive filter through the current residual noise signal.
7. The method of claim 1, further comprising, after adaptively weighting the first noise reduction drive signal and the second noise reduction drive signal to obtain a target noise reduction drive signal: driving a loudspeaker in the noise reduction channel to generate second reverse sound waves through the target noise reduction driving signal; identifying an actual residual noise signal received by an error microphone in a target noise reduction channel, wherein the actual residual noise signal is generated after linear superposition of environmental noise and a second reverse sound wave in a sound field; Calculating the power of the actual residual noise signal; and if the power of the actual residual noise signal is larger than a power threshold value, adjusting the weight for self-adaptive weighting.
8. A noise cancellation device, the device comprising: the first acquisition module is used for acquiring time sequence data of environmental noise in the vehicle through a plurality of microphones; The first determining module is used for analyzing the time sequence data and determining the main noise type and the noise intensity spatial distribution of the environmental noise; The second determining module is used for activating a target noise reduction channel of a target partition in the vehicle through the noise intensity spatial distribution and determining an original reference signal of a reference microphone in the target noise reduction channel through the main noise type; the calculation module is used for carrying out secondary path estimation and forward path calculation of a preset adaptive filter on the original reference signal to obtain a first noise reduction driving signal of a loudspeaker in a target noise reduction channel; The input/output module is used for acquiring a historical reference signal of the reference microphone and a training noise signal of the error microphone in the target noise reduction channel, inputting the historical reference signal and the training noise signal into a preset depth active noise control network, and outputting a second noise reduction driving signal; And the weighting module is used for carrying out self-adaptive weighting on the first noise reduction driving signal and the second noise reduction driving signal to obtain a target noise reduction driving signal, and the elimination of the vehicle environment noise is completed through the target noise reduction driving signal.
9. An electronic device, comprising: A processor; A memory for storing processor-executable instructions; Wherein the processor is configured to execute the instructions to implement the noise cancellation method of any one of claims 1 to 7.
10. A vehicle comprising the noise canceling device of claim 8.

Description

Noise elimination method and device, electronic equipment and vehicle Technical Field The present application relates to the field of vehicle technologies, and in particular, to a method and apparatus for noise cancellation, an electronic device, and a vehicle. Background With the continuous improvement of the requirements of people on the riding comfort of automobiles, the noise management in automobiles becomes an important research topic of automobile design. The existing in-vehicle noise management technology mainly comprises passive noise reduction measures represented by optimization of sound insulation materials and structures, and active noise reduction technology based on an adaptive filtering algorithm. Passive noise reduction techniques attenuate noise propagation by physical blocking and absorption, while active noise reduction techniques achieve noise cancellation by generating inverted sound waves that are opposite in phase to the noise. However, the scheme has obvious defects that on one hand, passive noise reduction measures have limited effects on low-frequency noise and are difficult to cope with complex and changeable noise environments, on the other hand, active noise reduction technology has better performance under low-frequency stable noise, but the noise reduction effect is obviously reduced when high frequency, sudden or environment is rapidly changed, the adaptability is insufficient when nonlinear and non-stable noise is processed, the separation and offset effects on multi-source composite noise are poor, and the requirement of increasing the driving comfort of an automobile is difficult to be met. Disclosure of Invention In view of the above, the present application aims to provide a noise cancellation method, apparatus, electronic device and vehicle, which solve the problems that the noise reduction effect is significantly reduced when the current noise reduction technology is high-frequency, sudden or environmental fast changed, and the adaptability is insufficient when nonlinear and non-stationary noise is processed, the separation and cancellation effect on multi-source composite noise is poor, and the requirement of increasing the driving comfort of the automobile is difficult to meet, and the specific technical scheme is as follows: According to a first aspect of the present application, there is provided a noise cancellation method comprising: collecting time sequence data of environmental noise in the vehicle through a plurality of microphones; analyzing the time sequence data and determining the main noise type and noise intensity spatial distribution of the environmental noise; activating a target noise reduction channel of a target partition in a vehicle through the noise intensity spatial distribution, and determining an original reference signal of a reference microphone in the target noise reduction channel through the main noise type; performing secondary path estimation and forward path calculation of a preset adaptive filter on the original reference signal to obtain a first noise reduction driving signal of a loudspeaker in a target noise reduction channel; Acquiring a historical reference signal of the reference microphone and a training noise signal of an error microphone in a target noise reduction channel, inputting the historical reference signal and the training noise signal into a preset depth active noise control network, and outputting a second noise reduction driving signal; And carrying out self-adaptive weighting on the first noise reduction driving signal and the second noise reduction driving signal to obtain a target noise reduction driving signal, and completing the elimination of the environmental noise of the vehicle through the target noise reduction driving signal. Optionally, after the collecting the time sequence data of the environmental noise in the vehicle through the plurality of microphones, the method includes: Performing band-pass filtering and wavelet transformation on the time sequence data to obtain wavelet coefficients; Identifying a dominant noise type of the ambient noise from the timing data; If the main noise type is a low-frequency stable main noise type, processing the wavelet coefficient by using a soft threshold algorithm to obtain a first screening wavelet coefficient; If the main noise type is a high-frequency main noise type or a burst main noise type, processing the wavelet coefficient by using a hard threshold algorithm to obtain a second screening wavelet coefficient; performing wavelet inverse transformation on the first screening wavelet coefficient or the second screening wavelet coefficient to obtain denoising time sequence data; and carrying out normalization processing and self-adaptive gain control on the denoising time sequence data to obtain preprocessed time sequence data. Optionally, the analyzing the time sequence data, determining the main noise type and the noise intensity spatial distribution of the