Noise suppression method and device for loudspeaker

Abstract

The invention is suitable for the technical field of audio processing, and provides a noise suppression method and device for a loudspeaker, and the method comprises the steps: obtaining a current vehicle speed and a current position type, and carrying out the matching of the current vehicle speed and a driving noise feature corresponding to the current position type; collecting vehicle exterior environment audio data through a microphone module, and identifying and inhibiting audio data corresponding to the driving noise characteristics from the vehicle exterior environment audio data to obtain a target noise audio; generating a reverse audio corresponding to the target noise audio, and playing the reverse audio through a loudspeaker; the reverse audio is used for noise suppression. According to the scheme, the noise component which is most required to be suppressed in the current environment can be locked more accurately, so that the depth and the breadth of noise reduction are greatly improved.

Description

Technical Field

[0001] This invention belongs to the technical field of audio processing, and particularly relates to a noise suppression method and apparatus for a loudspeaker. Background Technology

[0002] With the rapid development of the automotive industry and the increasing demands of consumers for driving and riding experiences, the performance and sound quality of in-car audio systems have become one of the important indicators for measuring vehicle quality. A good in-car acoustic environment can provide passengers with an immersive music experience, clear call quality, and a pleasant entertainment experience. However, vehicles inevitably generate various noises during operation, such as wind noise, tire noise, and engine noise. These driving noises can intrude into the cabin, severely interfering with the audio signals played by the in-car speakers, leading to a decrease in sound quality and loss of detail. This forces passengers to increase the volume, which, over time, not only affects the auditory experience but may also exacerbate auditory fatigue.

[0003] To address this issue, active noise cancellation technology has been introduced into the automotive field. Traditional in-vehicle active noise cancellation systems are typically based on pre-established noise models related to vehicle speed. This technology has a good suppression effect on engine harmonic noise and resonance noise at specific frequencies.

[0004] However, existing technical solutions have significant limitations. First, their noise models are typically only strongly correlated with vehicle speed, failing to adequately consider the external acoustic environment in which the vehicle operates. Existing single models cannot accurately match these diverse dynamic noise scenarios, resulting in unsatisfactory noise reduction effects and poor adaptability. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a noise suppression method and apparatus for loudspeakers to solve the technical problem that existing single models cannot accurately match these diverse dynamic noise scenarios.

[0006] A first aspect of the present invention provides a noise suppression method for a loudspeaker, the noise suppression method for the loudspeaker comprising: The current vehicle speed and current location type are obtained, and the driving noise features corresponding to the current vehicle speed and current location type are matched; wherein, the location type includes at least one of non-road area, highway, urban road and tunnel; the driving noise features do not include prompt type audio features, and the prompt type audio features include at least one of the features corresponding to siren, horn and pedestrian prompt sound; The system collects ambient audio data from outside the vehicle using a microphone module, identifies and suppresses audio data corresponding to the driving noise characteristics from the ambient audio data, and obtains the target noise audio. The inverse audio corresponding to the target noise audio is generated and played through a speaker; the inverse audio is used for noise suppression.

[0007] Furthermore, the step of obtaining the current vehicle speed and current location type, and matching the driving noise features corresponding to the current vehicle speed and current location type includes: Obtain standard noise characteristics corresponding to multiple preset parameter combinations; wherein each preset parameter combination includes a preset vehicle speed and a preset position type; the standard noise characteristics include amplitude-frequency characteristic parameters and time-domain envelope parameters; the amplitude-frequency characteristic parameters include the peak amplitude, center frequency, bandwidth, and global spectral slope corresponding to each of the multiple resonance peaks; the time-domain envelope parameters include the constant RMS value corresponding to the steady-state component, the start-up time of the transient component, the decay time of the transient component, and the pulse repetition frequency of the transient component; Match multiple target standard noise features corresponding to the current location type; The preset vehicle speed and current vehicle speed corresponding to each of the multiple target standard noise features are sorted in descending order of value, and the first preset vehicle speed and the second preset vehicle speed adjacent to the current vehicle speed are extracted. Based on the current vehicle speed, interpolation is performed on the first standard noise feature corresponding to the first preset vehicle speed and the second standard noise feature corresponding to the second preset vehicle speed to obtain the current noise feature corresponding to the current vehicle speed. Construct the target amplitude-frequency response in the frequency domain and the target envelope function in the time domain based on the current noise characteristics; The target amplitude-frequency response and the target envelope function are used as the driving noise characteristics.

[0008] Further, the step of interpolating the first standard noise feature corresponding to the first preset vehicle speed and the second standard noise feature corresponding to the second preset vehicle speed based on the current vehicle speed to obtain the current noise feature corresponding to the current vehicle speed includes: Subtract the first preset speed corresponding to the first standard noise feature from the second preset speed corresponding to the second standard noise feature to obtain the first speed difference; wherein the order corresponding to the first preset speed is before the order corresponding to the second preset speed; Subtract the current vehicle speed from the second preset speed to obtain the second speed difference; Divide the first speed difference and the second speed difference to obtain the speed weighting factor; Extract two parameters of the same parameter type from the first standard noise feature and the second standard noise feature; Subtract the parameter corresponding to the first standard noise feature from the parameter corresponding to the second standard noise feature to obtain the parameter difference; Multiply the parameter difference by the speed weighting factor to obtain the current value; The parameters corresponding to the second standard noise feature are added to the current value to obtain the interpolation result; The interpolation results corresponding to each parameter type are used as the current noise feature corresponding to the current vehicle speed.

[0009] Furthermore, the step of constructing the target amplitude-frequency response in the frequency domain and the target envelope function in the time domain based on the current noise characteristics includes: An amplitude-frequency function is constructed based on the global spectral slope, the peak amplitudes of multiple resonance peaks, the center frequencies of multiple resonance peaks, and the bandwidths of multiple resonance peaks; wherein, the amplitude-frequency function is: H_target(f) = G * (f / f_ref) (S_slope) +Σ[ A_k * exp( -((f - f_k) / bw_k) 2 H_target(f) represents the noise energy at a frequency point, f represents the frequency point, G represents the overall gain, f_ref represents the reference frequency, S_slope represents the global spectral slope, A_k represents the peak amplitude of the k-th resonant, f_k represents the center frequency of the k-th resonant, and bw_k represents the bandwidth of the k-th resonant. The noise energy at each frequency point between 20 Hz and 20000 Hz is taken as the target amplitude-frequency response; Set the envelope value corresponding to the steady-state component to 1; The target envelope function corresponding to the transient component is set as: Env_target(t) =Σ[ u(t - n / RR) *exp( - (t - n / RR) / τ) * sin(2π* f_impact * (t - n / RR)) ]; where Env_target(t) represents the envelope value at time t, u(t - n / RR) represents the unit step function with respect to t - n / RR, RR represents the pulsation repetition frequency in the current noise feature, n represents the order of the filter, τ represents the decay time in the current noise feature, and f_impact represents the typical frequency of the impulse noise.

[0010] Furthermore, the step of acquiring ambient audio data from outside the vehicle via a microphone module, identifying and suppressing audio data corresponding to the driving noise characteristics from the ambient audio data, and obtaining the target noise audio includes: Audio data of the external environment is collected via a microphone module; The external ambient audio data is input into a gamma-ton filter bank to obtain sub-band signals output by each of the multiple gamma-ton filters. Calculate the adaptive weights corresponding to each sub-band signal; Based on the adaptive weights, the sub-band signals are weighted and merged to obtain the target noise audio.

[0011] Furthermore, the step of calculating the adaptive weights corresponding to each sub-band signal includes: Calculate the short-time energy of each sub-band signal; Obtain the frequency range of the sub-band signal, and perform integral processing on the amplitude-frequency function within the frequency range to obtain the predicted energy of the sub-band signal; Calculate the correlation coefficient between the short-time energy and the predicted energy; The correlation coefficient raised to the power of λ is used as the correlation weight; where λ is used to adjust the sensitivity. Calculate the energy difference between the short-time energy of the current sub-band signal and the adjacent sub-band signal; When the energy difference exceeds the energy threshold, the transient weight of the current sub-band signal is set to 0.1; When the energy difference does not exceed the energy threshold, the transient weight of the current sub-band signal is set to 1; The noise type is matched based on the current vehicle speed and current location type; wherein, the noise type includes at least one of wind noise, tire noise, engine noise and bump noise; Match the frequency range corresponding to the noise type, and extract the sub-band signal corresponding to the frequency range; Based on the noise type corresponding to the sub-band signal, the current vehicle speed is substituted into the weighting function corresponding to the noise type to obtain the speed adjustment weight of the sub-band signal; The adaptive weight is obtained by multiplying the correlation weight, the transient weight, and the velocity adjustment weight.

[0012] Furthermore, the weighting function corresponding to the noise type is: ; ; .

[0013] A second aspect of the present invention provides a noise suppression device for a loudspeaker, comprising: The acquisition unit is used to acquire the current vehicle speed and the current location type, and match the driving noise features corresponding to the current vehicle speed and the current location type; wherein, the location type includes at least one of non-road area, highway, urban road and tunnel; the driving noise features do not include prompt type audio features, and the prompt type audio features include at least one of the features corresponding to siren sound, horn sound and pedestrian prompt sound; The identification unit is used to collect audio data of the external environment through the microphone module, identify and suppress the audio data corresponding to the driving noise feature from the audio data of the external environment, and obtain the target noise audio. The generation unit is used to generate the inverse audio corresponding to the target noise audio and play the inverse audio through a speaker; the inverse audio is used for noise suppression.

[0014] A third aspect of the present invention provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the noise suppression method for a speaker described in the first aspect.

[0015] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the noise suppression method for a loudspeaker described in the first aspect.

[0016] The beneficial effects of this invention compared to existing technologies are as follows: Traditional technologies rely solely on vehicle speed, have a single model, and poor adaptability. This solution, however, obtains the current location type and combines it with vehicle speed to match corresponding driving noise characteristics, achieving contextualized and refined noise reduction strategies. The noise spectrum, sound pressure level, and main sources of vehicles vary significantly in different environments such as highways, urban roads, and tunnels. This solution pre-establishes a noise feature library for these specific scenarios, enabling the system to select the most suitable noise reduction target according to local conditions. For example, in tunnels, the system prioritizes suppressing strong reverberation and low-frequency resonance; on urban roads, it may focus more on mid-to-high frequency tire noise and surrounding vehicle noise. Compared to traditional methods, this solution can more accurately pinpoint the noise components most in need of suppression in the current environment, thereby significantly improving the depth and breadth of noise reduction and providing users with a more stable and pure acoustic environment. Through technical means, harmful noise and safety warning sounds are isolated at the audio signal level, ensuring that the noise reduction process does not endanger driving safety. This fundamentally eliminates the risk of important external warning sounds being masked by the active noise cancellation system. The driver can clearly hear danger signals from outside the vehicle, thus enabling them to react in a timely manner. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A schematic flowchart of a noise suppression method for a loudspeaker provided by the present invention is shown; Figure 2 A schematic diagram of a noise suppression device for a loudspeaker according to an embodiment of the present invention is shown; Figure 3 A schematic diagram of a terminal device provided in an embodiment of the present invention is shown. Detailed Implementation

[0019] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0020] This invention provides a noise suppression method and apparatus for loudspeakers to solve the technical problem that traditional technologies cannot fully reflect the performance degradation of loudspeakers during actual use.

[0021] First, this invention provides a noise suppression method for a loudspeaker. Please see below. Figure 1 , Figure 1 A schematic flowchart of a noise suppression method for a loudspeaker provided by the present invention is shown. Figure 1 As shown, the noise suppression method for this loudspeaker may include the following steps: Step 101: Obtain the current vehicle speed and current location type, and match the driving noise features corresponding to the current vehicle speed and current location type; wherein, the location type includes at least one of non-road area, highway, urban road and tunnel; the driving noise features do not include prompt type audio features, and the prompt type audio features include at least one of the features corresponding to siren, horn and pedestrian prompt sound; Vehicle speed is one of the main sources of noise. At different speeds, the frequency and intensity distribution of tire noise, wind noise, and engine noise are quite different (for example, engine noise is dominant at low speeds, while wind noise and tire noise are dominant at high speeds).

[0022] Location type is used to introduce environmental context. Highways: Expect continuous, high-frequency wind and tire noise. Urban roads: Expect complex noise composition, potentially including more low-to-mid-frequency engine noise and ambient noise, with significant intensity variations. Tunnels: Expect strong resonant and reflected noise, with noise energy concentrated in specific frequency bands. Non-road areas (e.g., parking lots): Expect very low noise levels, potentially requiring no activation or operating noise reduction at minimal power.

[0023] Once the system obtains the current speed and location type (which can be obtained through GPS, map data, or vehicle sensors), it retrieves the expected noise from the database. This provides an accurate target profile for subsequent noise identification.

[0024] The noise to be canceled does not include important safety warning sounds, such as sirens, horns, and pedestrian alerts (e.g., low-speed warning sounds from electric vehicles). This avoids the active noise cancellation system canceling out important safety warning sounds from outside the vehicle, thus ensuring driving safety. When building the driving noise feature database, typical frequency bands and features of these safety warning sounds are deliberately avoided, or a whitelist is set in the algorithm to ensure that these features are not identified as noise that needs to be suppressed.

[0025] Specifically, step 101 includes steps 1011 to 1016: Step 1011: Obtain the standard noise features corresponding to each of the multiple preset parameter combinations; wherein, each preset parameter combination includes a preset vehicle speed and a preset position type; the standard noise features include amplitude-frequency characteristic parameters and time-domain envelope parameters; the amplitude-frequency characteristic parameters include the peak amplitude corresponding to each of the multiple resonance peaks, the center frequency corresponding to each of the multiple resonance peaks, the bandwidth corresponding to each of the multiple resonance peaks, and the global spectral slope; the time-domain envelope parameters include the constant RMS value corresponding to the steady-state component, the start-up time of the transient component, the decay time of the transient component, and the pulse repetition frequency of the transient component; This step builds the foundational database for the entire system. It defines how to fully describe a type of noise using a precise set of mathematical parameters.

[0026] The preset parameter combinations are sampling points that have been measured in advance in a laboratory or rigorous testing environment. For example, (60km / h, highway), (80km / h, highway), (100km / h, highway), etc. Each combination corresponds to a standard noise characteristic.

[0027] Standard noise characteristics include, but are not limited to, amplitude-frequency response parameters and time-domain envelope parameters.

[0028] A. Amplitude-frequency response parameters (describing the timbre of the noise): Formants are the frequency bands where noise energy is most concentrated, like a fingerprint of sound. The center frequency refers to the position of the formant on the frequency axis. The peak amplitude is the intensity of the formant. The bandwidth is the width of the formant, affecting the fullness of the timbre. The global spectral slope describes the trend of the overall noise energy as frequency changes (e.g., whether the high-frequency components decay rapidly). This effectively distinguishes between a low-pitched boom (gentle slope) and sharp wind noise (steep slope).

[0029] B. Time-domain envelope parameters (describing the dynamic changes of noise): The steady-state component (constant RMS value) represents the continuous and stable part of the noise (such as tire noise during constant speed driving). The RMS value refers to the average energy.

[0030] Transient components represent the sudden, rapidly changing parts of noise. Start-up time indicates the time required for transient noise to rise from zero to its peak value. Decay time indicates the time required for transient noise to decay from its peak value to a steady state. Pulse repetition frequency indicates the frequency at which transient noise (such as tires rolling over consecutive speed bumps) repeats. With these parameters, the system no longer stores a fixed audio segment, but rather a recipe for resynthesizing that noise. This lays the foundation for subsequent dynamic interpolation.

[0031] Step 1012: Match multiple target standard noise features corresponding to the current location type; Step 1013: Sort the preset vehicle speed and current vehicle speed corresponding to each of the multiple target standard noise features in descending order of value, and extract the first preset vehicle speed and the second preset vehicle speed adjacent to the current vehicle speed. Step 1014: Based on the current vehicle speed, interpolate the first standard noise feature corresponding to the first preset vehicle speed and the second standard noise feature corresponding to the second preset vehicle speed to obtain the current noise feature corresponding to the current vehicle speed; Steps 1012 to 1014 resolve the contradiction between continuously changing vehicle speed and a limited number of database sampling points. First, based on the current location type (e.g., highway), all standard noise feature samples belonging to that location type are identified from the database. These samples are then sorted by speed. Assuming the current speed is 95 km / h, and the database contains two sample points at (80 km / h) and (100 km / h), then 100 km / h is the first preset vehicle speed, and 80 km / h is the second preset vehicle speed.

[0032] The system does not abruptly switch between noise levels at 80 km / h and 100 km / h. Instead, it performs linear or nonlinear interpolation on all parameters (center frequency, amplitude, bandwidth, slope, RMS value, etc.) at these two sample points to calculate the parameters that should exist at 95 km / h. For example, the center frequency of a certain resonance peak at 95 km / h is the weighted average of that frequency at 80 km / h and 100 km / h.

[0033] This allows the current noise characteristics generated by the system to change smoothly and continuously with speed, completely avoiding the problem of abrupt changes in noise reduction effect caused by small changes in speed, thus improving comfort.

[0034] Specifically, step 1014 includes steps A1 to A8: Step A1: Subtract the first preset speed corresponding to the first standard noise feature from the second preset speed corresponding to the second standard noise feature to obtain the first speed difference; wherein the order of the first preset speed is before the order of the second preset speed. Step A2: Subtract the current vehicle speed from the second preset speed to obtain the second speed difference; Step A3: Divide the first speed difference and the second speed difference to obtain the speed weighting factor; Let's use a specific example to illustrate the entire explanation: First preset speed (V1): 80 km / h; Second preset speed (V2): 100 km / h; Current vehicle speed (Vc): 95 km / h; Interpolation parameter: center frequency (F) of the first resonance peak; At 80 km / h, F1 = 180 Hz; At 100 km / h, F2 = 220 Hz.

[0035] The purpose of steps A1 to A3 is to determine the position of the current velocity Vc relative to the two reference points V1 and V2, and to quantize it as a weighting factor between 0 and 1.

[0036] First speed difference (ΔV_total): V2 - V1 = 100 - 80 = 20 km / h; This represents the total velocity span across the entire interpolation interval.

[0037] Second speed difference (ΔV_current): Vc - V1 = 95 - 80 = 15 km / h; This represents the distance that the current velocity Vc is from the starting point V1.

[0038] Velocity weighting factor (α): α = ΔV_current / ΔV_total = 15 / 20 = 0.75; This factor of 0.75 is key. It means that the characteristic of a current speed of 95 km / h should have progressed by 75% from the characteristic of 80 km / h to the characteristic of 100 km / h.

[0039] Step A4: Extract two parameters of the same parameter type from the first standard noise feature and the second standard noise feature; Step A5: Subtract the parameter corresponding to the first standard noise feature from the parameter corresponding to the second standard noise feature to obtain the parameter difference; Step A6: Multiply the parameter difference by the velocity weighting factor to obtain the current value; Step A7: Add the parameter corresponding to the second standard noise feature to the current value to obtain the interpolation result; Steps A4 to A7 use the calculated weighting factor α to interpolate a specific parameter (such as the center frequency F).

[0040] The same parameter is extracted from the features of V1 and V2, namely F1 = 180 Hz and F2 = 220 Hz.

[0041] Parameter difference (ΔF): F2 - F1 = 220 - 180 = 40 Hz This means that when the speed increases from 80 km / h to 100 km / h, the frequency needs to increase by a total of 40 Hz.

[0042] Current value (Increment): ΔF * α = 40 Hz * 0.75 = 30 Hz; This represents the increment that the frequency should increase relative to the starting point V1 at the current speed of 95 km / h.

[0043] Interpolation result (Fc): F1 + Increment = 180 Hz + 30 Hz = 210 Hz; This is the final calculated center frequency of the resonance peak at the current speed of 95 km / h.

[0044] The general formula for the above calculation is: Current parameter value = + ( - ) * ( (current speed - ) / ( - ) ).

[0045] Step A8: Use the interpolation results corresponding to each parameter type as the current noise feature corresponding to the current vehicle speed.

[0046] The system needs to perform interpolation calculations cyclically, performing the above linear interpolation operation independently for each individual parameter.

[0047] Ultimately, the collection of all these interpolation results constitutes a complete noise characteristic specific to the current speed of 95 km / h. This newly generated characteristic perfectly transitions between the characteristics of 80 km / h and 100 km / h in both the frequency and time domains.

[0048] In the embodiments corresponding to steps A1 to A8, the interpolation method can effectively combine the noise features at two preset speeds to generate a noise feature corresponding to the current vehicle speed. This method not only considers the change in vehicle speed, but also ensures the accuracy and continuity of the results through the introduction of weighting factors, thereby enhancing the effectiveness of the noise suppression method.

[0049] Step 1015: Construct the target amplitude-frequency response in the frequency domain and the target envelope function in the time domain based on the current noise characteristics; The set of scattered parameters (digits) obtained through interpolation needs to be reassembled into a form that can be directly used by signal processing algorithms.

[0050] Using the interpolated formant parameters (center frequency, bandwidth, amplitude) and global spectral slope, a corresponding digital filter frequency response curve is constructed in digital signal processing. This curve accurately describes the target noise in the frequency domain and will serve as a frequency domain template for noise identification in subsequent steps.

[0051] By using the interpolated time-domain parameters (steady-state RMS, transient onset / decay time, etc.), a function describing how the noise amplitude changes over time is generated. This will serve as a time-domain template for noise identification, helping the system distinguish between continuous noise and sudden impacts.

[0052] Ultimately, driving noise characteristics are no longer an abstract concept, but are concretized into two computable models: the target amplitude-frequency response for frequency domain filtering and matching, and the target envelope function for time domain analysis and event detection.

[0053] Specifically, step 1015 includes steps B1 to B4: Step B1: Construct an amplitude-frequency function based on the global spectral slope, the peak amplitude of each of the multiple resonance peaks, the center frequency of each of the multiple resonance peaks, and the bandwidth of each of the multiple resonance peaks; wherein, the amplitude-frequency function is: H_target(f) = G * (f / f_ref) (S_slope) +Σ[ A_k * exp( -((f - f_k) / bw_k) 2 H_target(f) represents the noise energy at a frequency point, f represents the frequency point, G represents the overall gain, f_ref represents the reference frequency, S_slope represents the global spectral slope, A_k represents the peak amplitude of the k-th resonant, f_k represents the center frequency of the k-th resonant, and bw_k represents the bandwidth of the k-th resonant. The overall gain (which can be set to 0.832) is used to adjust the overall volume of the synthesized template spectrum to match the noise level in the real physical world. The overall gain can be set based on the average of driving environment noise levels in multiple environments.

[0054] f_ref (which can be set to 1000 Hz) is a reference point used when calculating the spectral slope. The spectral slope S_slope represents the amount of attenuation relative to this reference frequency.

[0055] The function H_target(f) aims to accurately predict the energy (amplitude) of target noise at a given frequency f. It consists of two core components: the global spectral background and the local resonant structure. ① Global spectral background (G * (f / f_ref)^(S_slope)): G (overall gain) is a scaling factor that adjusts the overall volume. (f / f_ref)^(S_slope) is a model describing the overall color or slope of the noise. f_ref is an arbitrary reference frequency (e.g., 1000 Hz) used for normalization. S_slope (global spectral slope) is a key parameter. If S_slope is negative (e.g., -3), it indicates that energy decays as frequency increases, consistent with the characteristics of most noise (high-frequency sound waves dissipate more easily). If S_slope is 0, it represents a flat, white noise-like spectrum.

[0056] This section captures the underlying basis of noise, that is, what the overall profile of noise looks like, setting aside prominent peaks.

[0057] ② Local resonance peak structure (Σ[ A_k * exp( -((f - f_k) / bw_k)^2 ) ]): Σ represents the summation of the contributions of all resonants (k=1, 2, 3...). This means that the final spectrum is the result of all resonants superimposed on the global background. A_k * exp( -((f - f_k) / bw_k)^2 ) is a model describing a single resonant. f_k (center frequency) is the frequency point where the resonant energy is highest. A_k (peak amplitude) is the intensity of the resonant at that center frequency. bw_k (bandwidth) controls the thickness of the resonant. The narrower the bandwidth, the sharper the resonant; the wider the bandwidth, the flatter the resonant. exp( -((f - f_k) / bw_k)^2 ) is a Gaussian bell curve centered at f_k. This value drops rapidly as f deviates from f_k.

[0058] This section precisely captures the most prominent and characteristic tonal components of noise (such as a specific roar of an engine).

[0059] By calculating the value of H_target(f) at each discrete frequency point within the audible range (20Hz-20kHz), a complete digital array is obtained: the target amplitude-frequency response.

[0060] Step B2: Use the noise energy at each frequency point between 20 Hz and 20000 Hz as the target amplitude-frequency response; Step B3: Set the envelope value corresponding to the steady-state component to 1; Step B4: Set the target envelope function corresponding to the transient component as: Env_target(t) =Σ[ u(t - n / RR) * exp( - (t - n / RR) / τ) * sin(2π* f_impact * (t - n / RR)) ]; where Env_target(t) represents the envelope value at time t, u(t - n / RR) represents the unit step function with respect to t - n / RR, RR represents the pulsation repetition frequency in the current noise feature, n represents the order of the filter, τ represents the decay time in the current noise feature, and f_impact represents the typical frequency of the impulse noise.

[0061] 'n' represents the filter order, typically set to 4. The typical frequency of impact noise can be roughly estimated by tapping the tire and analyzing the recording. For most car tires, this frequency is usually between 80 Hz and 250 Hz. For example, a common value is 180 Hz.

[0062] The purpose of `Env_target(t)` is to accurately describe how the amplitude of the noise changes with time `t`. It also consists of two parts: a steady-state envelope and a transient envelope. ① Steady-state envelope (value = 1): This represents continuous, stable noise (such as wind noise during constant-speed driving). Its envelope is a horizontal straight line with a value of 1, indicating constant amplitude.

[0063] ② Transient envelope (Σ[ u(t - n / RR) * exp( - (t - n / RR) / τ) * sin(2π * f_impact * (t - n / RR)) ]): This is a model with a very clear physical meaning, used to describe the thumping sound produced when a tire travels over an uneven road surface.

[0064] u(t - n / RR) (unit step function) is a switch. It ensures that the transient component does not exist (its value is 0) before time t = n / RR; it is triggered only at that time and afterwards. exp( - (t - n / RR) / τ) (exponential decay) simulates how the amplitude of the impact sound decays exponentially over time after it is generated. τ (decay time) controls the rate of decay. sin(2π * f_impact * (t - n / RR)) (high-frequency oscillation) is an impact that is not a pure snap; it contains a brief high-frequency vibration. f_impact is the typical frequency of this vibration. In Σ and n / RR, n is an integer (0, 1, 2...), and RR is the pulse repetition frequency. This means that the transient event occurs periodically at frequency RR.

[0065] This envelope function allows the system to identify not only the timbre of noise, but also its rhythm and dynamics. For example, the system can anticipate a transient event with specific attenuation characteristics and impact timbre that occurs at fixed intervals on a highway (depending on speed and road joint spacing). When the temporal envelope of the actual audio captured by the microphone closely matches this Env_target(t), the system can be confident that it has found the target noise.

[0066] In the embodiments corresponding to steps B1 to B4, the target amplitude-frequency response in the frequency domain and the target envelope function in the time domain are constructed based on the current noise characteristics. This method comprehensively considers both steady-state and transient noise components, providing comprehensive frequency and time characteristics for the noise suppression algorithm, making the suppression effect more accurate and effective.

[0067] Step 1016: Use the target amplitude-frequency response and the target envelope function as the driving noise features.

[0068] For example: Scenario: The vehicle is traveling on the highway at a speed of v_current = 77 km / h.

[0069] Determine the interpolation unit: The database contains two sets of data: (v_low=60, T_j) and (v_high=90, T_j). α_v = (77-60) / (90-60) = 17 / 30 ≈ 0.567; Interpolation calculation: At (60, high speed), the spectral slope of wind noise S_slope_low = -6 dB / Oct. At (90, high speed), the spectral slope of wind noise S_slope_high = -9 dB / Oct.

[0070] Calculation: S_slope_curr = -6 + 0.567 * (-9 - (-6)) = -6 - 1.7 = -7.7 dB / Oct.

[0071] At (60, high speed), one resonance peak of tire noise is at f_low = 180 Hz.

[0072] At (90, high speed), the resonance peak shifts to f_high = 200 Hz (due to the increased tire rotation speed).

[0073] Calculation: f_curr = 180 + 0.567 * (200-180) = 180 + 11.34 = 191.34 Hz.

[0074] Build template: H_target(f) will be a curve with a prominent resonance peak near 191Hz and an overall slope decreasing at -7.7 dB / Oct.

[0075] Env_target(t) will be a pulse sequence that repeats at a specific frequency (which can be calculated based on the tire circumference) corresponding to a vehicle speed of 77 km / h.

[0076] In the embodiments corresponding to steps 1011 to 1016, a high degree of reproduction of complex noise is achieved through dual parameterization in both the frequency and time domains, far exceeding the method of simply recording a piece of noise. The introduction of interpolation algorithms enables the system to handle an infinite variety of actual vehicle speeds, achieving a seamless and smooth transition in noise reduction characteristics. Storing parameters instead of audio data consumes fewer resources; matching is achieved by constructing filters and envelope functions, resulting in a clear computational objective and high efficiency.

[0077] Step 102: Collect ambient audio data from outside the vehicle using a microphone module, identify and suppress audio data corresponding to the driving noise characteristics from the ambient audio data, and obtain the target noise audio. Microphones are placed outside the vehicle (such as in the rearview mirror or chassis) to collect real-time ambient sounds. This is a necessary step in obtaining information about the actual noise sources.

[0078] The system does not process all the acquired sounds, but instead uses the driving noise characteristics obtained in step 101 as a reference signal or filter template. Through digital signal processing algorithms, it accurately identifies audio components that match the driving noise characteristics in the complex environmental audio acquired in real time. These identified target noise components are then extracted or separated from the total environmental audio data. Here, suppression can be understood as filtering, with the aim of obtaining a relatively pure target noise audio that is primarily composed of the target driving noise.

[0079] After this step, the system has eliminated interference from non-target sounds such as music, human voices, and safety alerts, resulting in a very clean noise signal that needs to be canceled.

[0080] Specifically, step 102 includes steps 1021 to 1024: Step 1021: Collect ambient audio data from outside the vehicle using the microphone module; Step 1022: Input the external ambient audio data into a gamma-ton filter bank to obtain sub-band signals output by each of the multiple gamma-ton filters; Gamma-pass filters are a set of bandpass filters covering the audible frequency range of the human ear (20Hz-20kHz). Their characteristics (bandwidth, shape) are designed to mimic the frequency analysis mechanism of the basilar membrane in the human ear. They are characterized by a narrower bandwidth and higher frequency selectivity at low frequencies, and a wider bandwidth at high frequencies, consistent with the concept of the critical bandwidth of human hearing.

[0081] Because it mimics the human ear, the signal processed by it more accurately reflects the noise components actually heard by a person. It decomposes a broadband signal into dozens of sub-band signals. Each sub-band signal contains information within only a narrow frequency band, making it simpler to process. Decomposing the signal into individual sub-bands allows the system to independently analyze the energy and characteristics of each frequency band.

[0082] Step 1023: Calculate the adaptive weights corresponding to each sub-band signal; The system calculates a weight (a value between 0 and 1) for each sub-band signal output from the gamma-pass filter. It analyzes the characteristics of each sub-band signal (such as energy and spectral shape) in real time and compares them to the driving noise characteristics generated in the first step (especially its target amplitude-frequency response). If the characteristics of a sub-band signal highly match the prediction of the corresponding frequency band in the driving noise characteristics, then this sub-band signal is very likely to be the target noise and is therefore assigned a high weight (close to 1). If the characteristics of a sub-band signal do not match the target characteristics (for example, the characteristic frequency of a siren suddenly appears in the frequency band), it will be assigned a low weight (close to 0). The weights quantify the confidence that each frequency segment belongs to the target noise.

[0083] Specifically, step 1023 includes steps C1 to C11: Step C1: Calculate the short-time energy of each sub-band signal; The short-time energy of each subband signal is calculated using a first-order IIR smoothing filter.

[0084] For each sub-band signal, analyze its energy within a very short time window. This represents the actual measured noise intensity.

[0085] Step C2: Obtain the frequency range of the sub-band signal, and perform integral processing on the amplitude-frequency function within the frequency range to obtain the predicted energy of the sub-band signal; Step C3: Calculate the correlation coefficient between the short-time energy and the predicted energy; The correlation coefficient is calculated as follows:

[0086] Represents the correlation coefficient. This represents the short-time energy of the subband signal. Indicates predicted energy. This represents a preset constant.

[0087] Step C4: Use the power of λ of the correlation coefficient as the correlation weight; where λ is used to adjust the sensitivity; λ (sensitivity adjustment) is a tuning knob. If λ > 1, the system penalizes signals with low correlation more severely (weights are drastically reduced), and the screening criteria are stricter. If λ < 1, the system is less sensitive to changes in correlation, and the screening is more lenient. For each sub-band signal, its energy is analyzed within a very short time window. This represents the actual measured noise intensity.

[0088] Step C5: Calculate the energy difference between the short-time energy of the current sub-band signal and the adjacent sub-band signal; Step C6: When the energy difference exceeds the energy threshold, set the transient weight of the current sub-band signal to 0.1; Step C7: When the energy difference does not exceed the energy threshold, set the transient weight of the current sub-band signal to 1; A large difference (exceeding the threshold) indicates the presence of an isolated, sharp transient event (such as a siren or horn). In this case, the system will set the transient weight of that subband to a very small value (such as 0.1), thereby strongly suppressing this signal component from entering the final target noise audio.

[0089] A small difference (not exceeding the threshold) means that the energy change is wide-bandwidth and gradual (which is characteristic of the target driving noise). In this case, the transient weight is 1, indicating that no additional suppression is performed.

[0090] This is a signal processing-based safety strategy that can effectively distinguish between continuous mechanical noise and sudden warning sounds.

[0091] Step C8: Match the noise type based on the current vehicle speed and current location type; wherein the noise type includes at least one of wind noise, tire noise, engine noise, and bump noise; For example, the mapping relationship between the current vehicle speed and the current location type and noise type is shown in Table 1: Table 1: Location type Vehicle speed range (km / h) Dominant noise types (ranked by importance) Triggering logic description City roads 0 - 20 (Starting speed, congestion) 1. Engine noise, 2. Tire noise At low speeds, wind noise is negligible. When the engine is running at low gears and high RPMs, noise becomes the dominant hazard. City roads 20 - 60 (Normal driving) 1. Tire noise, 2. Engine noise The engine runs relatively smoothly, and the noise from the friction between the tires and the common asphalt road surface is the main noise. City roads > 60 (Speeding / Expressway) 1. Wind noise, 2. Tire noise As vehicle speed increases, wind noise becomes noticeable and mixes with tire noise. highway 60 - 80 (Acceleration / Transition) 1. Tire noise, 2. Wind noise, 3. Engine noise Various noise sources are present, with tire noise being particularly prominent due to the typically rougher road surface. highway 80 - 120+ (cruising) 1. Wind noise, 2. Tire noise Wind noise increases cubically with vehicle speed, becoming the dominant noise. Tire noise is persistent. tunnel Any speed 1. Resonance noise, 2. Tire noise, 3. Engine noise (at low speeds) Resonance noise is a defining characteristic. Tunnels amplify tire and engine noise at specific frequencies (such as 200-300Hz). Non-road areas 0 - 30 (rough road surface) 1. Impact / bump noise Rough or unpaved roads cause frequent impact noises between the tires and the suspension system. This is the only scenario where the location type weight exceeds the vehicle speed. Step C9: Match the frequency range corresponding to the noise type and extract the sub-band signal corresponding to the frequency range; Step C10: Based on the noise type corresponding to the sub-band signal, substitute the current vehicle speed into the weighting function corresponding to the noise type to obtain the speed adjustment weight of the sub-band signal; Matching noise type: The system determines the main noise source based on speed / position.

[0092] Each noise type has a preset weighting function. The input is the current velocity, and the output is a weight value between 0 and 1.

[0093] Specifically, the weighting functions corresponding to the noise types include the wind noise baseline weighting function, the tire noise baseline weighting function, and the engine noise baseline weighting function.

[0094] ① ; This function describes the typical nonlinear growth relationship between wind noise and speed, and divides it into three policy intervals: Low-speed range (speed < 50 km / h): weight value is fixed at 0.2.

[0095] At low speeds, airflow is relatively gentle, and wind noise is not a major concern. Therefore, it is assigned a very low baseline weight to avoid the system overemphasizing it.

[0096] Mid-to-high speed cruising range (50 km / h ≤ speed ≤ 120 km / h): The weight starts at 0.2 and increases linearly with speed. Function logic: 0.2 + 1.3 * (speed - 50) / 70. In this range, wind noise is proportional to the square of the speed or even higher powers, and energy increases sharply. The function approximates this accelerating trend with linear growth. When the speed reaches 120 km / h, the weight increases to 0.2 + 1.3 = 1.5. This is a climbing range where the system's focus on wind noise continues to increase.

[0097] Ultra-high speed range (speed > 120 km / h): weight value is fixed at 1.5.

[0098] Once the speed exceeds a certain threshold, wind noise energy becomes enormous and tends to reach a stable dominant position. Therefore, the weight reaches its peak and remains stable.

[0099] The wind noise weighting function embodies an intelligent control strategy that ignores noise at low speeds, enhances it at medium speeds, and saturates it at high speeds.

[0100] ② ; This function describes the approximately linear relationship between tire noise and speed. It is a linear function with speed / 130 as the variable. Tire noise is mainly generated by the friction between the tire and the road surface, and its growth is generally more gradual and closer to a linear relationship than wind noise. The function maps speed to a range from 0.5 (at speed=0) to 1.2 (at speed=130). The baseline value of 0.5 ensures that tire noise has a basic weight even at speed 0, because tire noise is present as long as the vehicle is on the road.

[0101] The tire noise weighting function reflects a characteristic of steadily increasing with speed.

[0102] ③ .

[0103] This function describes the characteristics of engine noise: it is significant at low speeds and masked at high speeds.

[0104] Low speed / idle range (speed ≤ 40 km / h): The weight value is fixed at 1.4. In urban areas or during initial acceleration, engine speed fluctuates greatly, making this the main area contributing to noise. Therefore, a high baseline weight is assigned (1.4 is the maximum value among the three functions) so that the system can concentrate its efforts on suppressing engine noise and vibration at low speeds.

[0105] Medium to high speed range (speed > 40 km / h): Weight values ​​start at 1.4 and decrease exponentially with increasing speed. Function logic. As the vehicle speed increases, the engine enters a stable operating range, and its noise is masked by stronger wind and tire noise. (Using an exponential decay model) This is a classic and efficient method for describing this masking effect. The attenuation constant of 0.02 controls the rate of weight reduction.

[0106] The engine noise weighting function reflects a reverse thinking: low-speed noise is dominant and high-speed noise is attenuated.

[0107] These weighting functions are designed to dynamically adjust the influence weights of different noise sources based on changes in vehicle speed, thereby optimizing the noise suppression effect of the speakers. Through reasonable weight allocation, better noise control and ride comfort can be achieved under various driving conditions.

[0108] Step C11: Multiply the correlation weight, the transient weight, and the velocity adjustment weight to obtain the adaptive weight.

[0109] Adaptive weight = correlation weight × transient weight × velocity adjustment weight.

[0110] For a sub-band signal to receive high weight, it must simultaneously meet three conditions: its spectral characteristics are highly correlated with the prediction model; its time-domain characteristics are consistently stable rather than abruptly sharp; and its frequency band is precisely the band most likely to generate noise under the current vehicle speed and road conditions.

[0111] In the embodiments corresponding to steps C1 to C11, the entire adaptive weight calculation process is a multi-dimensional comprehensive consideration process, aiming to flexibly adjust the processing priority of each sub-band signal based on the short-time characteristics of the signal, the frequency domain characteristics of the noise, and the dynamic conditions of the vehicle. This process not only increases the effectiveness of noise suppression but also provides a theoretical and practical basis for achieving high-quality audio output.

[0112] Step 1024: Based on the adaptive weights, the sub-band signals are weighted and merged to obtain the target noise audio.

[0113] The system multiplies all sub-band signals by their respective calculated adaptive weights, and then sums all the weighted signals together. Sub-band signals identified as highly likely to belong to the target driving noise are preserved almost entirely after merging. Sub-band signals identified as not belonging to the target noise (such as speech or sirens) have very low weights, resulting in significant energy suppression in the merged result. Therefore, the output target noise audio is a relatively pure signal that contains almost only the driving noise we want to cancel.

[0114] In the embodiments corresponding to steps 1021 to 1024, the process of acquiring audio data of the external environment using a microphone module, and how to perform frequency domain analysis using a gamma-pass filter bank, calculate adaptive weights, and finally synthesize the target noise audio, comprehensively considers the complexity of driving noise and achieves effective noise suppression through an adaptive approach, ensuring that the final audio signal is clearer and more comfortable.

[0115] Step 103: Generate the inverse audio corresponding to the target noise audio and play the inverse audio through a speaker; the inverse audio is used for noise suppression.

[0116] Generating inverse audio is a classic physical application of active noise cancellation technology. According to the principle of sound wave interference, when two sound waves with the same frequency but opposite phases are superimposed, they will cancel each other out.

[0117] The system processes the target noise audio obtained in the second step in real time to generate a sound wave with the same amplitude but completely opposite phase (180 degrees out of phase), i.e., reverse audio.

[0118] The generated reverse audio is played through the car's speakers. When this reverse sound wave meets the original target noise from outside the car at the passenger's ears, they undergo destructive interference, thus significantly reducing or even eliminating the driving noise perceived by the passenger, achieving a noise reduction effect and improving driving comfort.

[0119] In the embodiments corresponding to steps 101 to 103, traditional techniques rely solely on vehicle speed, resulting in a single model and poor adaptability. This solution, however, obtains the current location type and combines it with vehicle speed to match corresponding driving noise characteristics, achieving contextualized and refined noise reduction strategies. The noise spectrum, sound pressure level, and main sources of vehicles vary significantly in different environments such as highways, urban roads, and tunnels. This solution pre-establishes a noise feature library for these specific scenarios, enabling the system to select the most suitable noise reduction target based on local conditions. For example, in tunnels, the system prioritizes suppressing strong reverberation and low-frequency resonance; on urban roads, it may focus more on mid-to-high frequency tire noise and surrounding vehicle noise. Compared to traditional methods, this solution can more accurately pinpoint the noise components most in need of suppression in the current environment, thereby significantly improving the depth and breadth of noise reduction, providing users with a more stable and pure acoustic environment, and greatly enhancing driving comfort. Through technical means, harmful noise and safety warning sounds are isolated at the audio signal level, ensuring that the noise reduction process does not endanger driving safety. This fundamentally eliminates the risk of important external warning sounds being masked by the active noise cancellation system. Drivers can clearly hear danger signals from outside the vehicle, allowing them to react promptly.

[0120] like Figure 2 This invention provides a noise suppression device for a loudspeaker; please refer to [link / reference]. Figure 2 , Figure 2 A schematic diagram of a noise suppression device for a loudspeaker provided by the present invention is shown, as follows: Figure 2 The noise suppression device for a loudspeaker shown includes: The acquisition unit 21 is used to acquire the current vehicle speed and the current location type, and match the driving noise features corresponding to the current vehicle speed and the current location type; wherein, the location type includes at least one of non-road area, highway, urban road and tunnel; the driving noise features do not include prompt type audio features, and the prompt type audio features include at least one of the features corresponding to siren sound, honk sound and pedestrian prompt sound; The identification unit 22 is used to collect audio data of the external environment through the microphone module, identify and suppress the audio data corresponding to the driving noise feature from the audio data of the external environment, and obtain the target noise audio. The generation unit 23 is used to generate the inverse audio corresponding to the target noise audio and play the inverse audio through a speaker; the inverse audio is used for noise suppression.

[0121] This invention provides a noise suppression device for loudspeakers. Traditional technologies rely solely on vehicle speed, resulting in a single model and poor adaptability. This solution, however, obtains the current location type and combines it with vehicle speed to match corresponding driving noise characteristics, achieving contextualized and refined noise reduction strategies. The noise spectrum, sound pressure level, and main sources of vehicles vary significantly in different environments such as highways, urban roads, and tunnels. This solution pre-establishes a noise feature library for these specific scenarios, enabling the system to select the most suitable noise reduction target based on the environment. For example, in tunnels, the system prioritizes suppressing strong reverberation and low-frequency resonance; on urban roads, it may focus more on mid-to-high frequency tire noise and surrounding vehicle noise. Compared to traditional methods, this solution can more accurately pinpoint the noise components most in need of suppression in the current environment, thereby significantly improving the depth and breadth of noise reduction, providing users with a more stable and pure acoustic environment, and greatly enhancing driving comfort. Through technical means, harmful noise and safety warning sounds are isolated at the audio signal level, ensuring that the noise reduction process does not jeopardize driving safety. This fundamentally eliminates the risk of important external warning sounds being masked by the active noise cancellation system. Drivers can clearly hear danger signals from outside the vehicle, allowing them to react promptly.

[0122] Figure 3 This is a schematic diagram of a terminal device provided in an embodiment of the present invention. Figure 3As shown, a terminal device 3 in this embodiment includes: a processor 30, a memory 31, and a computer program 32 stored in the memory 31 and executable on the processor 30, such as a noise suppression program for a speaker. When the processor 30 executes the computer program 32, it implements the steps in the various embodiments of the noise suppression method for a speaker described above, for example... Figure 1 Steps 101 to 103 are shown. Alternatively, when the processor 30 executes the computer program 32, it implements the functions of each unit in the above-described device embodiments, for example... Figure 2 The function of the unit shown.

[0123] For example, the computer program 32 can be divided into one or more units, which are stored in the memory 31 and executed by the processor 30 to complete the present invention. The one or more units can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program 32 in the terminal device 3. For example, the specific functions of each unit of the computer program 32 can be divided as follows: The acquisition unit is used to acquire the current vehicle speed and the current location type, and match the driving noise features corresponding to the current vehicle speed and the current location type; wherein, the location type includes at least one of non-road area, highway, urban road and tunnel; the driving noise features do not include prompt type audio features, and the prompt type audio features include at least one of the features corresponding to siren sound, horn sound and pedestrian prompt sound; The identification unit is used to collect audio data of the external environment through the microphone module, identify and suppress the audio data corresponding to the driving noise feature from the audio data of the external environment, and obtain the target noise audio. The generation unit is used to generate the inverse audio corresponding to the target noise audio and play the inverse audio through a speaker; the inverse audio is used for noise suppression.

[0124] The terminal device includes, but is not limited to, a processor 30 and a memory 31. Those skilled in the art will understand that... Figure 3 This is merely an example of a terminal device 3 and does not constitute a limitation on a terminal device 3. It may include more or fewer components than shown, or combine certain components, or different components. For example, the terminal device may also include input / output devices, network access devices, buses, etc.

[0125] The processor 30 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0126] The memory 31 can be an internal storage unit of the terminal device 3, such as a hard disk or memory of the terminal device 3. The memory 31 can also be an external storage device of the terminal device 3, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the terminal device 3. Furthermore, the memory 31 can include both internal and external storage units of the terminal device 3. The memory 31 is used to store the computer program and other programs and data required by the roaming control device. The memory 31 can also be used to temporarily store data that has been output or will be output.

[0127] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0128] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of the present invention. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0129] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this invention. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0130] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0131] This invention provides a computer program product that, when run on a mobile terminal, enables the mobile terminal to implement the steps described in the above-described method embodiments.

[0132] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying the computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0133] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0134] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0135] In the embodiments provided by this invention, it should be understood that the disclosed apparatus / network devices and methods can be implemented in other ways. For example, the apparatus / network device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0136] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units.

[0137] It should be understood that, when used in this specification and the appended claims, terms include indicating the presence of the described feature, integral, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0138] It should also be understood that the terms used in this specification and the appended claims refer to any combination of one or more of the associated listed items and all possible combinations, and include such combinations.

[0139] As used in this specification and the appended claims, the term "if" can be interpreted, depending on the context, as when, once, or in response to determination or in response to detection. Similarly, the phrase "if determined" or "if detected [the described condition or event]" can be interpreted, depending on the context, as once determined or in response to determination or detection [the described condition or event] or in response to detection [the described condition or event].

[0140] Furthermore, in the description of this invention and the appended claims, the terms first, second, third, etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0141] References to one or more embodiments described in this specification mean that a particular feature, structure, or characteristic described in connection with that embodiment is included in one or more embodiments of the invention. Therefore, phrases appearing in different parts of this specification as referring to one embodiment, some embodiments, some other embodiments, and others do not necessarily refer to the same embodiment, but rather mean one or more, but not all, embodiments, unless otherwise specifically emphasized. The terms include, comprise, have, and variations thereof mean including but not limited to, unless otherwise specifically emphasized.

[0142] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method of noise suppression for a loudspeaker, characterized by, The noise suppression method of the loudspeaker comprises: Obtaining the current vehicle speed and the current location type, and matching the driving noise characteristics corresponding to the current vehicle speed and the current location type; wherein the location type comprises at least one of the non-road area, the highway, the urban road and the tunnel; the driving noise characteristics do not contain prompt type audio characteristics, and the prompt type audio characteristics comprise at least one of the corresponding characteristics of the siren sound, the whistle sound and the pedestrian prompt sound; Collecting the external environment audio data of the vehicle through the microphone module, identifying and suppressing the audio data corresponding to the driving noise characteristics from the external environment audio data of the vehicle, and obtaining the target noise audio; Generating the reverse audio corresponding to the target noise audio, and playing the reverse audio through the loudspeaker; the reverse audio is used for noise suppression.

2. The method of claim 1, wherein the speaker is a speakerphone. The step of obtaining the current vehicle speed and the current location type, and matching the driving noise characteristics corresponding to the current vehicle speed and the current location type comprises: Obtaining the standard noise characteristics corresponding to a plurality of preset parameter combinations; wherein each preset parameter combination comprises a preset vehicle speed and a preset location type; the standard noise characteristics comprise amplitude-frequency characteristic parameters and time-domain envelope parameters; the amplitude-frequency characteristic parameters comprise the peak amplitude corresponding to each of a plurality of resonance peaks, the center frequency corresponding to each of a plurality of resonance peaks, the bandwidth corresponding to each of a plurality of resonance peaks and the global spectral slope; the time-domain envelope parameters comprise the constant RMS value corresponding to the steady-state component, the vibration time of the transient component, the decay time of the transient component and the pulse repetition frequency of the transient component; Matching a plurality of target standard noise characteristics corresponding to the current location type; Sorting the preset vehicle speeds corresponding to a plurality of target standard noise characteristics and the current vehicle speed in descending order of value, and extracting the first preset vehicle speed and the second preset vehicle speed adjacent to the current vehicle speed; Based on the current vehicle speed, performing interpolation processing on the first standard noise characteristics corresponding to the first preset vehicle speed and the second standard noise characteristics corresponding to the second preset vehicle speed, to obtain the current noise characteristics corresponding to the current vehicle speed; According to the current noise characteristics, constructing the target amplitude-frequency response of the frequency domain part and the target envelope function of the time domain part; The target amplitude-frequency response and the target envelope function are used as the driving noise characteristics.

3. The method of claim 2, wherein the speaker is a speakerphone. The step of performing interpolation processing on the first standard noise characteristics corresponding to the first preset vehicle speed and the second standard noise characteristics corresponding to the second preset vehicle speed based on the current vehicle speed, to obtain the current noise characteristics corresponding to the current vehicle speed comprises: Subtracting the first preset speed corresponding to the first standard noise characteristics from the second preset speed corresponding to the second standard noise characteristics to obtain a first speed difference; wherein the order corresponding to the first preset speed is before the order corresponding to the second preset speed; Subtracting the current vehicle speed from the second preset speed to obtain a second speed difference; Dividing the first speed difference by the second speed difference to obtain a speed weight factor; extracting two parameters corresponding to the same parameter type in the first standard noise feature and the second standard noise feature; subtracting the parameter corresponding to the first standard noise feature from the parameter corresponding to the second standard noise feature to obtain a parameter difference value; multiplying the parameter difference value by the speed weight factor to obtain a current value; adding the parameter corresponding to the second standard noise feature to the current value to obtain an interpolation result; taking the interpolation result corresponding to each parameter type as a current noise feature corresponding to the current vehicle speed.

4. The method of claim 2, wherein the speaker noise is suppressed by, The step of constructing the target amplitude-frequency response of the frequency domain part and the target envelope function of the time domain part according to the current noise feature comprises: Based on the global spectrum slope, the peak value amplitude corresponding to each of the plurality of resonance peaks, the center frequency corresponding to each of the plurality of resonance peaks and the bandwidth corresponding to each of the plurality of resonance peaks, an amplitude-frequency function is constructed; wherein the amplitude-frequency function is: H_target(f) = G* (f / f_ref) (S_slope) +Σ[ A_k * exp( -((f - f_k) / bw_k) 2 ) ]; H_target(f) represents the noise energy at the frequency point, f represents the frequency point, G represents the overall gain, f_ref represents the reference frequency, S_slope represents the global spectrum slope, A_k represents the peak value amplitude of the kth resonance peak, f_k represents the center frequency of the kth resonance peak, and bw_k represents the bandwidth of the kth resonance peak. taking the noise energy of the amplitude-frequency function at each frequency point between 20 Hz and 20000 Hz as the target amplitude-frequency response; setting the envelope value corresponding to the steady-state component as 1; setting the target envelope function corresponding to the transient component as: Env_target(t) =Σ[ u(t - n / RR) * exp( - (t - n / RR) / τ) * sin(2π* f_impact * (t - n / RR)) ]; wherein Env_target(t) represents the envelope value at time t, u(t - n / RR) represents the unit step function with respect to t - n / RR, RR represents the pulsation repetition frequency in the current noise feature, n represents the order of the filter, τ represents the decay time in the current noise feature, and f_impact represents the typical frequency of the impact noise.

5. The method of claim 1, wherein the speaker is a speakerphone. The step of collecting the off-road audio data through the microphone module, identifying and suppressing the audio data corresponding to the driving noise feature from the off-road audio data, and obtaining the target noise audio comprises: collecting the off-road audio data through the microphone module; inputting the off-road audio data into a gamma tunnel filter set to obtain sub-band signals output by each gamma tunnel filter; calculating an adaptive weight corresponding to each sub-band signal; based on the adaptive weight, weighting and merging the sub-band signals to obtain the target noise audio.

6. The method of claim 5, wherein the speaker noise is suppressed by, The step of calculating an adaptive weight corresponding to each sub-band signal comprises: calculating the short-time energy of each sub-band signal; obtaining the frequency range of the sub-band signal, and integrating the amplitude-frequency function in the frequency range to obtain the predicted energy of the sub-band signal; calculating the correlation coefficient between the short-time energy and the predicted energy; taking the λth power of the correlation coefficient as the correlation weight; wherein λ is used to adjust the sensitivity; calculating the energy difference between the short-time energy of the current sub-band signal and the short-time energy of the adjacent sub-band signal; when the energy difference exceeds the energy threshold, setting the transient weight of the current sub-band signal to 0.1; when the energy difference does not exceed the energy threshold, setting the transient weight of the current sub-band signal to 1; matching the noise type according to the current vehicle speed and the current location type; wherein the noise type comprises at least one of wind noise, tire noise, engine noise, and jolt noise. Match the frequency range corresponding to the noise type, and extract the sub-band signal corresponding to the frequency range; Based on the noise type corresponding to the sub-band signal, the current vehicle speed is substituted into the weighting function corresponding to the noise type to obtain the speed adjustment weight of the sub-band signal; The adaptive weight is obtained by multiplying the correlation weight, the transient weight, and the velocity adjustment weight.

7. The method of claim 6, wherein the speaker noise is suppressed by, The weighting function corresponding to the noise type is: ； ； 。 8. A noise suppression device for a loudspeaker, characterized by The noise suppression device for the loudspeaker includes: The acquisition unit is used to acquire the current vehicle speed and the current location type, and match the driving noise features corresponding to the current vehicle speed and the current location type; wherein, the location type includes at least one of non-road area, highway, urban road and tunnel; the driving noise features do not include prompt type audio features, and the prompt type audio features include at least one of the features corresponding to siren sound, horn sound and pedestrian prompt sound; The identification unit is used to collect audio data of the external environment through the microphone module, identify and suppress the audio data corresponding to the driving noise feature from the audio data of the external environment, and obtain the target noise audio. The generation unit is used to generate the inverse audio corresponding to the target noise audio and play the inverse audio through a speaker; the inverse audio is used for noise suppression.

9. A terminal device, comprising: The terminal device includes: a memory, a processor, and a noise suppression program for a speaker stored in the memory and executable on the processor, the noise suppression program for the speaker being configured to implement the steps in the noise suppression method for the speaker as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, the computer-readable storage medium comprising: When the computer program is executed by the processor, it implements the steps in the noise suppression method for the loudspeaker as described in any one of claims 1 to 7.

Images