In-vehicle noise reduction method, device and vehicle

Abstract

The embodiment of the application provides a kind of in-vehicle noise reduction method, device and vehicle, the method comprises: obtaining the original noise signal of vehicle and the scene mode where current vehicle is located;According to scene mode, in the mapping relationship of pre-setting, the noise reduction priority of corresponding each seat is determined;Mapping relationship is the corresponding relationship between scene mode and the noise reduction priority of each seat;The original noise signal is processed by secondary channel estimation model, to obtain reference signal;According to the noise reduction priority, the reference signal and the residual noise signal collected in real time are iteratively processed by adaptive filtering algorithm, to obtain anti-noise signal;The noise of inhibition corresponding to anti-noise signal is output by the sound playing device in vehicle.This can meet the individualized noise reduction demand of user, improve user experience.

Description

Technical Field

[0001] This application relates to the field of active noise reduction control for intelligent vehicles, and more particularly to an in-vehicle noise reduction method, device, and vehicle. Background Technology

[0002] Active noise cancellation technology is mainly used to optimize cabin quietness in multi-seat vehicles such as sedans and SUVs.

[0003] Currently, by combining the original noise signal of the vehicle with the residual noise signal collected from various locations inside the vehicle, an adaptive filtering algorithm can be used to continuously adjust the control parameters to generate an anti-noise signal that cancels out the noise. This anti-noise signal is then output through in-vehicle speakers and other sound playback devices to reduce the residual noise level near each seat.

[0004] However, existing methods typically optimize control points for multiple seats simultaneously to achieve balanced noise reduction across all seats. In actual vehicle use, different scenarios may present personalized needs, such as primarily reducing noise at the driver's position or primarily at the rear seats. Existing methods cannot meet these personalized noise reduction requirements. Summary of the Invention

[0005] This application provides an in-vehicle noise reduction method, device, and vehicle, which can provide different noise reduction effects according to the scenario, meet users' personalized noise reduction needs, and improve user experience.

[0006] In a first aspect, embodiments of this application provide an in-vehicle noise reduction method, including:

[0007] Acquire the vehicle's raw noise signal and the current scene mode in which the vehicle is located;

[0008] Based on the scene mode, the noise reduction priority of each seat is determined in a pre-set mapping relationship; the mapping relationship is the correspondence between scene modes and the noise reduction priority of each seat.

[0009] The original noise signal is processed by a secondary channel estimation model to obtain a reference signal;

[0010] An anti-noise signal is obtained by iteratively processing the reference signal and the real-time acquired residual noise signal according to the noise reduction priority using an adaptive filtering algorithm.

[0011] The noise suppression signal corresponding to the noise reduction signal is output through the sound playback device inside the vehicle.

[0012] In one possible implementation, the noise reduction priority is a target weighting coefficient;

[0013] The step of iteratively processing the reference signal and the acquired real-time residual noise signal according to the noise reduction priority using an adaptive filtering algorithm to obtain an anti-noise signal includes:

[0014] Based on the target weight coefficients, generate a secondary channel weighting coefficient vector;

[0015] The weighted cost function in the adaptive filtering algorithm is determined based on the secondary channel weighted coefficient vector and the residual noise signal.

[0016] The anti-noise signal is obtained by iteratively updating the weight vector of the adaptive weighted filter bank based on the weighted cost function and the reference signal using an adaptive filtering algorithm according to the weighted filter update formula.

[0017] In one possible implementation, determining the weighted cost function in the adaptive filtering algorithm based on the secondary channel weighting coefficient vector and the residual noise signal includes:

[0018] The product of the secondary channel weighting coefficient vector and the residual noise signal is determined as the weighted cost function.

[0019] In one possible implementation, the scene mode is selected by the user from a plurality of pre-stored scene modes, or an automatic mode, or a default scene mode obtained when the vehicle is powered on, or a scene mode used before the vehicle was last powered off.

[0020] In one possible implementation, if the scene mode is automatic mode, determining the noise reduction priority of each seat according to the scene mode and a pre-set mapping relationship includes:

[0021] In a pre-set mapping relationship, the initial weight coefficient corresponding to the automatic mode is determined; the initial weight coefficient is the same for each seat.

[0022] Obtain the occupancy status of each seat in the vehicle;

[0023] The initial weight coefficients are adjusted according to the occupancy status of each seat to obtain the target weight coefficients; wherein the weight coefficients corresponding to seats in an occupied state are greater than the weight coefficients corresponding to seats in an unoccupied state.

[0024] In one possible implementation, the occupancy status of each seat in the vehicle is determined based on the seat pressure signal detected by the pressure detection device installed at each seat, and / or the images of each seat acquired by the image acquisition device installed in the vehicle.

[0025] In one possible implementation, the method further includes:

[0026] For the first seat with a target weight coefficient of 0, when the residual noise signal collected in real time is greater than the preset value, its residual noise signal is determined as the preset value.

[0027] In one possible implementation, after outputting the noise suppression signal corresponding to the noise reduction signal through the in-vehicle audio playback device, the method further includes:

[0028] In response to a user's adjustment operation on the noise reduction priority corresponding to at least one second seat, a new noise reduction priority for the at least one second seat is obtained;

[0029] The noise reduction priority is adjusted according to the new weighting coefficient, and a new anti-noise signal is determined based on the adjusted noise reduction priority.

[0030] The new noise suppression signal is output through the sound playback device inside the vehicle, corresponding to the noise suppression.

[0031] In one possible implementation, adjusting the noise reduction priority based on the new weighting coefficient includes:

[0032] Based on the new weighting coefficients, the noise reduction priority is adjusted using linear interpolation.

[0033] In one possible implementation, the method further includes:

[0034] Receive user input of new scene modes and new weight coefficients for each seat under the new scene modes;

[0035] The correspondence between the newly added scene mode and the newly added weight coefficient is added to the mapping relationship.

[0036] Secondly, embodiments of this application provide an in-vehicle noise reduction device, comprising:

[0037] The acquisition module is used to acquire the vehicle's original noise signal and the current scene mode of the vehicle;

[0038] The determination module is used to determine the noise reduction priority of each seat according to the scene mode and in a pre-set mapping relationship; the mapping relationship is the correspondence between the scene mode and the weight coefficient of each seat.

[0039] The processing module is used to process the original noise signal through a secondary channel estimation model to obtain a reference signal; and to iteratively process the reference signal and the residual noise signal acquired in real time according to the noise reduction priority using an adaptive filtering algorithm to obtain an anti-noise signal.

[0040] The output module is used to output the noise suppression signal corresponding to the noise reduction signal through the sound playback device in the vehicle.

[0041] Thirdly, embodiments of this application provide a vehicle, including: a vehicle body, a memory, and a processor;

[0042] The memory stores computer-executed instructions;

[0043] The processor executes computer execution instructions stored in the memory, causing the processor to perform the method described in the first aspect and / or various possible embodiments of the first aspect.

[0044] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the methods described in the first aspect and / or various possible implementations of the first aspect.

[0045] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the methods described in the first aspect and / or various possible implementations of the first aspect.

[0046] This application provides an in-vehicle noise reduction method, device, and vehicle. The method includes: acquiring the vehicle's original noise signal and the current scene mode of the vehicle; determining the noise reduction priority of each seat according to the scene mode in a pre-set mapping relationship; the mapping relationship is the correspondence between the scene mode and the noise reduction priority of each seat; processing the original noise signal through a secondary channel estimation model to obtain a reference signal; iteratively processing the reference signal and the real-time acquired residual noise signal according to the noise reduction priority using an adaptive filtering algorithm to obtain an anti-noise signal; and outputting the suppressed noise corresponding to the anti-noise signal through an in-vehicle sound playback device. Thus, by acquiring the vehicle's original noise signal and the current scene mode of the vehicle, and determining the noise reduction priority of each seat based on the scene mode, the in-vehicle active noise reduction process can be transformed from balanced control for all seats to differentiated control for the needs of different seats. By processing the original noise signal using a secondary channel estimation model to obtain a reference signal, and combining it with the residual noise signal using an adaptive filtering algorithm iteratively processing according to noise reduction priority, the generation of the anti-noise signal can more effectively allocate noise reduction capabilities to priority seats. This improves the noise reduction depth of key seats, enhances the resource utilization efficiency and scene adaptability of active noise reduction in multi-seat vehicles. Therefore, this application can provide different noise reduction effects according to the scene, meet users' personalized noise reduction needs, and improve user experience. Attached Figure Description

[0047] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0048] Figure 1 A flowchart illustrating an in-vehicle noise reduction method provided in this application;

[0049] Figure 2 A schematic diagram of an interface for users to select a scene mode, provided in this application;

[0050] Figure 3 A schematic diagram illustrating the weighting coefficients of each seat in different scenario modes provided in this application;

[0051] Figure 4 A flowchart illustrating another in-vehicle noise reduction method provided in this application;

[0052] Figure 5 A schematic diagram of a data interaction provided for this application;

[0053] Figure 6 This application provides a structural schematic diagram of an in-vehicle noise reduction device;

[0054] Figure 7 This is a structural schematic diagram of a vehicle provided in this application.

[0055] The accompanying drawings have illustrated specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to specific embodiments. Detailed Implementation

[0056] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application.

[0057] Active noise cancellation technology falls under the field of intelligent automotive acoustic control and is primarily applied to optimize cabin quietness in multi-seat vehicles such as sedans and SUVs. Specifically, it generates and outputs noise suppression signals through speakers installed within the vehicle to create noise reduction zones around each seat, thus reducing noise at each seating position. As users increasingly demand greater driving comfort and personalized experiences, the quietness requirements of occupants in different positions within the vehicle are becoming more distinct. For example, the driver prioritizes deep noise reduction for focused driving, while other seats may cater to different comfort needs.

[0058] Currently, by combining the original noise signal of the vehicle with the residual noise signal collected from various locations inside the vehicle, an adaptive filtering algorithm can be used to continuously adjust the control parameters to generate an anti-noise signal that cancels out the noise. This anti-noise signal is then output through in-vehicle speakers and other sound playback devices to reduce the residual noise level near each seat.

[0059] Current in-vehicle noise reduction solutions typically rely on simultaneous optimization of multiple control points, unifying noise reduction targets across multiple seats to achieve an overall reduction in residual noise throughout the vehicle and a roughly balanced noise reduction effect across different locations (i.e., each seat). However, as users increasingly demand greater driving comfort and personalized experiences, their noise reduction needs vary, and existing methods cannot meet these individualized requirements. For example, when some seats are vacant, allocating noise reduction resources to the corresponding control points forces adaptive filtering algorithms to consider areas without actual demand, unnecessarily consuming algorithmic freedom and hindering the achievement of deeper noise reduction effects on key seats. Therefore, existing technologies lack sufficient adaptability to different scenarios.

[0060] In view of this, this application proposes an in-vehicle noise reduction method. During vehicle operation, the original noise signal and the current scene mode of the vehicle are first acquired. Then, based on the scene mode, the noise reduction priority corresponding to each seat is determined in a pre-set mapping relationship. Subsequently, the original noise signal is processed by a secondary channel estimation model to obtain a reference signal. An adaptive filtering algorithm then iteratively processes the reference signal, noise reduction priority, and real-time acquired residual noise signal to output an anti-noise signal. Finally, the corresponding suppressed noise is output through the in-vehicle audio playback device. This method follows the basic architecture of in-vehicle active noise cancellation systems, which includes reference noise acquisition, controller processing, and audio playback device output. It introduces the seat priority requirements under different scenarios into the adaptive noise reduction process, allowing the anti-noise signal generation process to be adjusted around key seats. This ensures that the generated noise reduction effect meets the user's personalized noise reduction needs, thereby improving the scene adaptability of in-vehicle noise reduction and enhancing the user experience.

[0061] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0062] Figure 1 A flowchart illustrating an in-vehicle noise reduction method provided in this application is shown below. Figure 1 As shown, the method includes:

[0063] S101. Obtain the vehicle's original noise signal and the current scene mode of the vehicle.

[0064] The original noise signal is the noise source to be suppressed during vehicle operation and serves as the basic input for generating reference signals and driving noise reduction control. This original noise signal can be acquired by reference microphones deployed near the front bulkhead, door panels, headliner, floor passage, dashboard, or engine compartment sound insulation area. Alternatively, it can be obtained indirectly from acceleration sensors, vibration sensors, or wheel speed-related sensors, acquiring characteristic quantities related to road noise, wind noise, and powertrain noise, and then preprocessing and converting them. This application does not limit this approach.

[0065] The scene mode represents the current noise reduction configuration of the vehicle, and its function is to provide a control basis for the allocation of noise reduction resources for each seat. The scene mode can be set by the user remotely through the vehicle's infotainment interface, steering wheel buttons, voice interaction module, or mobile terminal, or it can be automatically determined by the vehicle based on the current operating status, such as whether there is someone in the driver's seat, whether there are passengers in the back seats, vehicle speed range, driving mode, historical usage records, or default configuration. This application embodiment does not limit this.

[0066] In practice, after the vehicle is powered on or the active noise cancellation function is enabled, the controller first initializes the noise acquisition link, including microphone channel self-test, analog-to-digital converter synchronization, sampling rate setting, and buffer establishment. The reference microphone continuously acquires relevant noise signals inside and outside the vehicle at a preset sampling frequency, which are then amplified at the front end, filtered for anti-aliasing, and digitally processed to form the original noise signal.

[0067] The current scene mode of the vehicle can be obtained through the following two possible implementations:

[0068] One possible implementation is that after the user selects the enhanced quiet mode for the driver's seat, the balanced mode for the whole vehicle, the front row priority mode, the rear row rest mode, or the automatic mode on the central control interface, the selected mode identifier is written into the scene storage area, thereby obtaining the current scene mode.

[0069] Another possible implementation is that the controller extracts state parameters from seat occupancy sensors, door status, vehicle speed, navigation trip length, and driving behavior characteristics, and automatically generates scene modes through preset rules, such as determining the driver's seat enhanced quiet mode when only the driver's seat is occupied and the vehicle is in high-speed cruising mode.

[0070] Regarding the synchronization relationship between the original noise signal and the scene mode, this embodiment can add a timestamp or control cycle number to each data frame so that the original noise signal collected in that cycle corresponds to the scene mode that is effective in the same cycle.

[0071] S102. Based on the scene mode, determine the noise reduction priority of each seat in the pre-set mapping relationship.

[0072] The mapping relationship is the correspondence between scene modes and noise reduction priorities of each seat.

[0073] The mapping relationship is used to establish the correspondence between scene modes and the noise reduction priorities of each seat. For example, it can be a lookup table in the vehicle's infotainment system memory, a calibration parameter table, a matrix parameter set, or a rule engine. The noise reduction priority of each seat is used to characterize the importance of each seat in the active noise reduction optimization process.

[0074] For example, the noise reduction priority can be directly represented by the weight coefficient corresponding to each seat, or it can be further converted into a normalized priority value from the weight coefficient. This application embodiment does not limit this.

[0075] Seats may include the driver's seat, front passenger seat, second-row left seat, second-row middle seat, second-row right seat, or, depending on the vehicle model, extend to the third row of seats.

[0076] Specifically, multiple scene modes and their corresponding noise reduction priorities are pre-saved. For example, in the driver's seat enhanced quietness mode, the noise reduction priority for the driver's seat is set to a higher priority, while the front passenger seat and rear seats are set to a lower priority. In the overall vehicle balanced mode, the noise reduction priorities for all seats are set to the same or nearly the same priority. In the front seat priority mode, the front seats have a higher priority than the rear seats. In the rear seat rest mode, the rear seats have a higher priority than the front seats.

[0077] It should be noted that each weighting coefficient can be limited to a preset value range.

[0078] When the system detects a user actively switching modes, passengers getting on or off the vehicle, or changes in the automatic scene determination results, the controller re-queries the mapping relationship and updates the noise reduction priority. During the update process, a smooth transition coefficient can be introduced to gradually interpolate and switch between the old and new weights over several control cycles, reducing abrupt changes in the sound field.

[0079] S103. The original noise signal is processed by the secondary channel estimation model to obtain the reference signal.

[0080] The secondary channel estimation model is used to characterize the acoustic transmission path from the in-vehicle sound playback device to the error detection location. This path typically includes the speaker's electroacoustic conversion characteristics, cabin space propagation characteristics, seat and interior reflection characteristics, and error microphone reception characteristics. Since the active noise cancellation control signal does not directly act on the target location after being output by the speaker, but propagates through the secondary acoustic channel, the original noise signal needs to be processed by the secondary channel estimation model to form a reference signal suitable for the convergence of the adaptive algorithm.

[0081] The reference signal is a key input for subsequent filter updates, used to describe the equivalent excitation upon which noise control is based after taking into account the effects of secondary channels.

[0082] In practice, a secondary channel estimation model is pre-acquired or maintained online. This model can be obtained through vehicle factory calibration, i.e., injecting test signals into each speaker under static testing conditions, measuring the response of the corresponding error microphone, and establishing a discrete impulse response model using a system identification method; alternatively, it can be updated during vehicle operation using auxiliary identification signals, historical estimates, or a self-calibration mechanism. After acquiring the raw noise signal, the data can be organized according to the number of speaker channels and the number of error channels, and then filtered using the secondary channel estimation model.

[0083] In one possible embodiment, if a finite-time impulse response model is used, the original noise signal can be convolved with the estimated secondary channel impulse response to obtain a filtered reference signal.

[0084] In another exemplary implementation, if the system is implemented in the frequency domain, the controller first performs frame windowing and frequency domain transformation on the original noise signal, then multiplies it by the secondary channel estimated response in the frequency domain, and then performs inverse transformation or directly retains the frequency domain result as a reference signal for subsequent iterations.

[0085] It should be noted that, in discrete time domain representation, the original noise signal can be denoted as: The secondary channel estimation model can be denoted as: Then the reference signal can be expressed as ,in This represents the convolution operation. The expression reflects that parameter updates are not performed directly using the original noise signal, but rather using a filtered reference that considers the propagation characteristics from the loudspeaker to the error point. This makes the gradient estimation of the subsequent adaptive filter closer to the real physical system, improving the algorithm's convergence speed and stability, and avoiding update direction deviations due to uncompensated secondary channels.

[0086] S104. An adaptive filtering algorithm is used to iteratively process the reference signal and the residual noise signal acquired in real time according to the noise reduction priority to obtain the noise-resistant signal.

[0087] The adaptive filtering algorithm is used to continuously update the control filter parameters based on error feedback, so that the generated anti-noise signal cancels out the original noise at the target location.

[0088] The residual noise signal is used to characterize the residual noise level after noise reduction. It is usually collected in real time by error microphones deployed near the head of each seat, near the backrest, inside the headrest, or in the area adjacent to the seat.

[0089] In this step, noise reduction priority is not merely a static label, but directly participates in the iterative update process, used to change the contribution of different seating errors to the cost function and filter coefficient correction. The noise reduction signal is the control output calculated by the adaptive filter based on the current reference signal, historical filter coefficients, and weighted error feedback.

[0090] In practice, the controller synchronously acquires the reference signal and the residual noise signals of each error channel within each control cycle. The residual noise signals can first undergo sampling synchronization, gain correction, bandwidth limiting, and outlier removal processing consistent with the reference signal to improve the consistency of the algorithm input. Subsequently, each error channel is weighted according to the noise reduction priority of each seat obtained in S102. If a seat-level weight-to-error-channel correspondence is adopted, the controller first determines the weight of each error microphone based on the seat to which it belongs, and then forms a weighted error vector. Taking a weighted FxLMS (Filtered-x Least MeanSquares) type algorithm as an example, the controller can construct a weighted cost function. , where wi represents the weight coefficient of the i-th seat or the corresponding error channel. This represents the residual noise signal of the i-th error channel at time n. This represents the total residual noise energy under the current control cycle.

[0091] By calculating the gradient of the cost function and updating the filter coefficients in conjunction with the reference signal, the seat errors with larger weights can account for a higher proportion in the update. If the weight of a certain seat is increased, the corresponding error term will have a greater impact on the direction and magnitude of the coefficient correction, thus preferentially reducing the residual noise near that seat.

[0092] In one possible embodiment, a time-domain multi-channel weighted FxLMS algorithm can be used. First, candidate noise-resistant signals for each speaker channel are generated based on the reference signal and the current filter coefficients. Then, the filter is updated according to a control step size via error feedback. The update formula includes an error term modulated by the weight coefficients. The step size can be set based on noise intensity, vehicle speed, or system stability margin to balance convergence speed and stability.

[0093] In another exemplary implementation, block processing or frequency domain adaptive algorithms can be used to reduce the computational burden of multi-channel processing, and a weighting strategy consistent with seat priority can be introduced in each frequency band so that control resources in different frequency bands can be allocated according to scenario requirements.

[0094] In this application, taking noise reduction priority as the target weight coefficient as an example, the adaptive filtering algorithm iteratively processes the reference signal and the acquired real-time residual noise signal according to the noise reduction priority to obtain the anti-noise signal. This process may include: generating a secondary channel weighting coefficient vector based on the target weight coefficient; determining the weighting cost function in the adaptive filtering algorithm based on the secondary channel weighting coefficient vector and the residual noise signal; and iteratively updating the weight vector of the adaptive weighted filter bank according to the weighted filter update formula based on the weighting cost function and the reference signal to obtain the anti-noise signal.

[0095] The target weight coefficient is used to characterize the target priority of each seat in noise reduction optimization.

[0096] The secondary channel weighted coefficient vector is used to organize the target weight coefficients corresponding to multiple seats into a unified vector form, so as to serve as the weighted input for each channel in the multi-channel adaptive noise reduction calculation.

[0097] The weighted cost function measures the overall error of residual noise under the current weight configuration. Its value changes with the weights corresponding to different seats, thus driving the filter to converge toward a direction with higher target priority. The adaptive weighted filter bank generates a control output corresponding to the reference signal. Its weight vector represents the current parameter state of the filter bank and is continuously adjusted with iterative updates to achieve continuous suppression of residual noise inside the vehicle.

[0098] In its implementation, after receiving the reference signal and the real-time residual noise signal, the controller first arranges the target weight coefficients according to the correspondence between seats or control points to form a secondary channel weighted coefficient vector. This vector can then be normalized to ensure that each channel participates in subsequent calculations under the same dimensions. Subsequently, a weighted cost function is constructed based on the secondary channel weighted coefficient vector and the residual noise signal. The weighted cost function can be composed of the mean square error of the residual noise signal and a weight term, so that the error corresponding to the high-weight seats has a higher proportion in the cost function.

[0099] Furthermore, the weight vector of the adaptive weighted filter bank is recursively updated according to the weighted filter update formula. During the update process, a step size parameter is introduced to balance the convergence speed and stability. Finally, an anti-noise signal matching the reference signal is output and played by the in-vehicle speaker to suppress noise.

[0100] For example, the weighted filter update formula is:

[0101]

[0102] in, This is the updated filter weight vector. This is the filter weight vector before the update. Step size factor For reference signal, This is the weighted cost function.

[0103] In this way, by constructing a secondary channel weighted coefficient vector and a weighted cost function, the adaptive filtering update process can allocate resources and converge errors around key seats. Since the iterative update of the weight vector is directly constrained by the weighted cost function, higher priority areas can achieve stronger noise suppression capabilities, while lower priority areas can still maintain basic noise reduction effects. This improves the scenario adaptability of active noise cancellation in multi-seat vehicles, enhances the noise reduction depth of key seats, and improves the utilization efficiency of noise reduction resources.

[0104] In this embodiment of the application, determining the weighted cost function in the adaptive filtering algorithm based on the secondary channel weighted coefficient vector and the residual noise signal may include: determining the weighted cost function by multiplying the secondary channel weighted coefficient vector and the residual noise signal.

[0105] The secondary channel weighting coefficient vector is used to characterize the noise suppression weights corresponding to different seats. It is generated by the target weighting coefficient and participates in the cost calculation so that the residual noise in different locations inside the vehicle is weighted according to a preset priority in the adaptive filtering update.

[0106] The residual noise signal is an error signal collected in real time inside the vehicle, reflecting the residual noise level after noise reduction, and is used as the input for multiplying with the weighting coefficient vector of the secondary channel.

[0107] The weighted cost function is used to carry the optimization objective of the adaptive filtering algorithm. Its numerical change directly drives the subsequent filter weight update, thereby affecting the generation of the noise-resistant signal.

[0108] In practical implementation, after obtaining the secondary channel weighting coefficient vector and the residual noise signal at the corresponding time, the controller can multiply the two according to their component correspondence. The product result is then used as the weighted cost function input to the adaptive filtering algorithm. For multi-channel in-vehicle active noise cancellation scenarios, the secondary channel weighting coefficient vector can be represented as a set of weights corresponding to each speaker output channel. For example, the secondary channel weighting coefficient vector can be... This directly determines the contribution weight of each seat in the adaptive iteration, thereby controlling the noise reduction depth of each seat at the final convergence.

[0109] The residual noise signal can be represented as the set of error signals collected by each error sensor. The weighted cost obtained by multiplying the two can apply a higher optimization intensity to the channels corresponding to key seats. This product operation can be completed by the numerical calculation unit inside the controller, and the result can be directly used as the filter cost term in the iterative update.

[0110] For example, the weighted cost function is:

[0111]

[0112] in, The weighted cost function value, Let be the target weighting coefficient for the i-th seat, i.e., the secondary channel weighting coefficient. This is a residual noise signal.

[0113] In this way, by determining the weighted cost function by multiplying the secondary channel weighting coefficient vector with the residual noise signal, a direct mapping relationship can be established between the cost function and the priority of the seats in the vehicle. This allows the adaptive filtering algorithm to more effectively suppress residual noise near key seats during the iteration process and reduce ineffective optimization of low-priority areas.

[0114] S105. The noise suppression signal is output through the in-vehicle sound playback device to suppress noise.

[0115] The in-vehicle sound playback device converts the noise-canceling signal output by the controller into actual sound waves and transmits them to the in-vehicle sound field to cancel out the original noise in the target area. The sound playback device can be a door panel speaker, dashboard speaker, headrest speaker, ceiling speaker, subwoofer, or a speaker array specifically designed for active noise cancellation in a vehicle audio system; this application does not specifically limit this. The noise-canceling signal, after digital-to-analog conversion, power amplification, and speaker electroacoustic conversion, forms suppressed noise. This suppressed noise is superimposed on the original noise in the corresponding seat area, thereby reducing the residual noise signal level and forming part of the closed-loop control.

[0116] In practice, after completing the filter output calculation in S104, the noise suppression signals of each speaker channel are written into the audio output buffer and synchronously sent to the vehicle audio amplifier or independent amplification unit according to the system clock. To ensure the suppression effect, the output delay of the noise suppression signal needs to be coordinated with the reference signal processing and error acquisition delay to ensure that the output suppressed noise and the target noise maintain an appropriate phase relationship in space and time.

[0117] This application allows for amplitude limiting, smoothing, bandwidth constraint, and fault monitoring of the noise-canceling signal to prevent excessive speaker output or perceptible distortion due to abnormal calculations. When a microphone malfunction, speaker open circuit, excessive model mismatch, or unstable control is detected, the system can degrade the output or temporarily shut down the active noise cancellation channel to maintain safe system operation.

[0118] In one possible embodiment, the output sound field effect continues to be collected by the error microphone to form a new residual noise signal, which is then fed back to S104 to enter the next control cycle, forming a continuous closed-loop iteration.

[0119] In another exemplary implementation, the total gain of the noise suppression signal can be dynamically adjusted according to changes in vehicle speed or scene mode, so that the noise suppression output intensity matches the actual noise conditions.

[0120] Therefore, this application provides an in-vehicle noise reduction method. By establishing a mapping relationship between scene modes and the noise reduction priorities of each seat, and directly incorporating these priorities into the adaptive iterative process of active noise cancellation, the contribution of each seat in the optimization process can be dynamically adjusted based on the current occupant distribution and quietness requirements. This reduces the impact of vacant or low-demand seats on the algorithm's degrees of freedom, while enabling stronger noise reduction control in key areas such as the driver's seat, front seats, and rear seats, meeting users' personalized noise reduction needs. This improves the scene adaptability of in-vehicle active noise cancellation and enhances the user experience.

[0121] In this application, the above-mentioned in-vehicle noise reduction method can be triggered in different scenarios, and the specific triggering scenario is related to the source of the scenario mode.

[0122] The scene mode is selected by the user from multiple pre-stored scene modes, or automatic mode, or the default scene mode obtained when the vehicle is powered on, or the scene mode used before the vehicle was last powered off.

[0123] Different scene modes correspond to different scenes. For example, if the scene mode is selected by the user from multiple pre-stored scene modes, it could be selected by the user from multiple scene modes displayed on a display screen installed in the vehicle.

[0124] Figure 2 This is a schematic diagram of an interface for users to select a scene mode, as provided in this application.

[0125] like Figure 2 As shown, the display screen can display scene mode 1, scene mode 2, scene mode 3 and scene mode 4. When the user wants to select scene mode 2, they can input a touch operation for scene mode 2 to select it.

[0126] In this embodiment, the scene mode is used to characterize the noise reduction application environment currently corresponding to the vehicle. It can be read from multiple pre-stored scene modes by the onboard controller and selected by the user. These scene modes can be pre-stored in the onboard storage unit or cloud configuration files, and can correspond to different usage scenarios such as focused driving, quiet commuting, and family travel. The scene mode can also be set to automatic mode, where the vehicle automatically determines the current scene and switches to the corresponding mode based on vehicle speed, gear position, power output status, door status, or occupant distribution information, thus completing scene adaptation without manual intervention. The scene mode can also be a default scene mode read when the vehicle is powered on. This default scene mode can be pre-written into the controller or storage unit by the vehicle's factory configuration parameters, and can be directly called as the current mode when the vehicle is powered on for the first time or when user preferences are not saved. The scene mode can also be a scene mode used before the vehicle was last powered off. The controller writes the current mode to non-volatile memory before the vehicle is powered off so that the previous usage state can be restored upon power-on, thereby maintaining the continuity of user settings.

[0127] For example, the scene mode selected by the user can be received through the human-machine interface, or an automatic mode identifier can be generated when preset automatic switching conditions are met, and this identifier can be used as the current scene mode in subsequent weight mapping. The default scene mode can be read from the configuration area and loaded into the running memory by the controller during the power-on initialization process, and the scene mode used by the vehicle before the last power failure can be written to the memory through the power failure saving mechanism and reloaded after power is restored.

[0128] This allows for flexible switching between manual settings, automatic recognition, default activation, and power outage recovery, ensuring that the current scene mode is consistent with the vehicle's usage status. This provides a stable input for determining the priority of seat noise reduction and improves the adaptability of in-vehicle noise reduction control to different driving habits and different startup states, thus meeting users' personalized noise reduction needs.

[0129] Figure 3 This is a schematic diagram illustrating the weighting coefficients of each seat in different scenario modes provided in this application.

[0130] like Figure 3 As shown, taking a vehicle with 4 seats as an example, in the equal seating mode (i.e., automatic mode), the weighting coefficient for each seat is: .

[0131] In the equal access mode, if a user selects the priority mode, the priority mode ( The driver's seat (the most important position) is assigned the highest weight, while the other three seats are assigned a weight of 0.2. The adaptive filter focuses on minimizing residual noise in the driver's seat, providing the deepest noise reduction for that position.

[0132] In the all-passenger equal mode, if the user enables automatic sensing, that is, selects the seat occupancy sensing mode, it is detected whether there are occupants on 4 seats. If it is detected that there are occupants on the driver seat and the front passenger seat, the device enters the driver seat and front passenger seat mode, and the weight coefficients corresponding to each seat are determined as . If it is detected that only the driver seat is occupied, and the left and right rear seats are occupied, the weight coefficients corresponding to each seat are determined as . If it is detected that only the driver seat is occupied, the device enters the driver exclusive mode, and the weight coefficients corresponding to each seat are determined as .

[0133] It should be noted that when only K seats out of M seats are occupied (K<M), the optimization problem is reduced from M-dimensional constraints to K-dimensional constraints, the algorithm converges under the constraints of fewer control points, and a single control point can obtain a deeper noise reduction effect. Taking the situation that only the driver seat is occupied as an example: the four-channel error constraint is reduced to one channel, all resources of the speaker array are used to eliminate noise in the driver's head area, and the noise reduction depth is significantly improved compared with that in the equal optimization of four points.

[0134] In the all-passenger equal mode, if the user selects rear quietness, the device enters the rear quiet mode, and the weight coefficients corresponding to each seat are determined as .

[0135] In the present application, when the scenario mode is the automatic mode, determining the corresponding noise reduction priority of each seat in a preset mapping relationship according to the scenario mode may include: determining an initial weight coefficient corresponding to the automatic mode in the preset mapping relationship; the weight coefficients corresponding to each seat in the initial weight coefficients are the same; acquiring the occupancy status of each seat in the vehicle; adjusting the initial weight coefficients according to the occupancy status of each seat to obtain target weight coefficients; wherein, the weight coefficient corresponding to an occupied seat is greater than the weight coefficient corresponding to an unoccupied seat.

[0136] Wherein, the automatic mode is used to automatically determine the noise reduction priority of each seat according to the seat occupancy situation when the distribution of vehicle occupants changes. The initial weight coefficients may be pre-stored in the mapping relationship, or may be generated with the same numerical value when entering the automatic mode, so as to facilitate subsequent adjustment.

[0137] The target weight coefficients are obtained by dynamically correcting the initial weight coefficients in combination with the occupancy status. After correction, the weight of occupied seats is higher than that of unoccupied seats, so that the occupied area can obtain higher priority in subsequent noise reduction calculation.

[0138] For example, upon recognizing the automatic mode, the basic weight configuration corresponding to that mode is first read to ensure all seats are at the same initial priority level. Then, the occupancy information of each seat is collected and updated, and the weight coefficients are redistributed according to preset gain rules. For seats that are occupied, their weight coefficients can be increased to enhance the control strength of the corresponding noise reduction area. For seats that are not occupied, the controller can maintain the original value or decrease their weight coefficients, thereby reducing the control resource usage in empty seat areas.

[0139] In this way, the weight configuration in automatic mode can be dynamically adjusted according to the occupancy status of the vehicle, preserving the unified foundation of the initial mapping relationship while giving higher noise reduction priority to seats with occupants. This reduces the ineffective allocation of resources to vacant seats, improves the utilization efficiency of noise reduction control resources, and enhances the adaptability of in-vehicle active noise cancellation to actual riding scenarios.

[0140] In this embodiment of the application, the occupancy status of each seat in the vehicle is determined based on the seat pressure signal detected by the pressure detection device installed at each seat, and / or the image of each seat acquired by the image acquisition device installed in the vehicle.

[0141] The pressure detection device is a sensor device used to detect the occupancy status of the seat. It can be installed inside the seat cushion, under the seat frame, or between the seat cover and the load-bearing layer. It is used to sense the pressure changes generated after the occupant sits in the seat and output the corresponding seat pressure signal.

[0142] Seat pressure signal is an electrical or digital signal obtained by a pressure detection device. It can characterize the force on the seat surface and thus serve as a basis for determining whether the seat is occupied.

[0143] The image acquisition device can be installed in the vehicle's ceiling, the area of ​​the front reading lights, above the dashboard, or other locations that can cover the areas of each seat, to acquire images of each seat and identify whether a human body is present in the seat. This image acquisition device can be a camera, but this application embodiment does not limit it to this.

[0144] In practical implementation, the system can receive seat pressure signals output from the pressure detection devices at each seat and compare these signals with a preset threshold. When the seat pressure signal is greater than the threshold, the corresponding seat is determined to be occupied; when the seat pressure signal is less than or equal to the threshold, the corresponding seat is determined to be unoccupied. The threshold setting can be calibrated based on the seat structure, sensor sensitivity, and vehicle usage scenarios to avoid misjudgments caused by slight vibrations or short-term disturbances.

[0145] When using an image acquisition device, human detection, human contour recognition, or human feature region extraction can be performed on the acquired seat images. Based on the recognition results, it can be determined whether the corresponding seat is occupied. When a human torso, head, or occupant posture feature is identified in the image, the seat is determined to be occupied. Pressure detection results can also be cross-validated with image recognition results. When the two are consistent, the final occupancy status is output; when they are inconsistent, the result with higher confidence can be used as the determination criterion, thereby improving the accuracy of occupancy status identification.

[0146] This allows occupied seats to be assigned a higher noise reduction priority, while unoccupied seats receive a lower priority. By identifying whether a seat is occupied in real time, noise reduction resources can be avoided by allocating them to vacant seats, thus concentrating noise suppression more effectively on the area where the actual occupant is located. Using pressure detection devices or image acquisition devices to determine occupancy status improves the real-time performance and reliability of seat occupancy assessment, thereby enhancing the adaptability of in-vehicle active noise cancellation to different riding scenarios and improving the noise reduction effect on key seats.

[0147] In this application, for the first seat with a corresponding target weight coefficient of 0, when the residual noise signal collected in real time is greater than a preset value, its residual noise signal is determined as the preset value.

[0148] The first seat corresponds to the seat with a target weight coefficient of 0. This seat does not participate in or is basically not involved in the noise reduction priority allocation in the current scenario. Therefore, its residual noise signal is only used as a limited reference quantity in subsequent control processing.

[0149] The preset value is used to constrain the upper limit of the residual noise signal of the first seat. It can be pre-written or set based on the statistical results of historical residual noise in the vehicle. For example, the preset value can be +3dB. This application embodiment does not limit this.

[0150] Specifically, after receiving the real-time residual noise signals of each seat, it can first determine whether the target weight coefficient corresponding to that seat is 0. If it is determined to be 0, the residual noise signal of that seat is compared with a preset value; when the detection result is greater than the preset value, the controller directly replaces the current collected value with the preset value and sends the replaced signal into the subsequent weighted filtering or error feedback calculation.

[0151] In this way, by limiting the residual noise signal of zero-weighted seats, it is possible to avoid abnormal noise peaks from vacant or low-priority seats occupying algorithm iteration resources, thereby reducing their interference with the update of adaptive filter weights. This allows more computational resources to be allocated to seat areas with actual noise reduction needs. This processing can also suppress error amplification caused by local strong noise fluctuations, improve the stability of adapting to complex noise environments inside the vehicle, and make the noise reduction effect of key seats more concentrated and consistent.

[0152] Based on the aforementioned embodiments, after outputting the noise suppression signal corresponding to the noise reduction signal through the in-vehicle audio playback device, the method further includes: in response to the user's adjustment operation on the noise reduction priority corresponding to at least one second seat, obtaining a new noise reduction priority for at least one second seat; adjusting the noise reduction priority according to the new weighting coefficient, determining a new noise reduction signal based on the adjusted noise reduction priority; and outputting the noise suppression signal corresponding to the new noise reduction signal through the in-vehicle audio playback device.

[0153] The second seat refers to the seat whose noise reduction priority can be adjusted by the user. The new noise reduction priority is the priority configuration formed after the user completes the adjustment, which is used to redistribute the noise reduction resources in the vehicle. The new weight coefficient is the parameter obtained from the new noise reduction priority and participates in the update of the anti-noise signal. The anti-noise signal is the output signal regenerated according to the adjusted noise reduction priority.

[0154] For example, the user's adjustment of the noise reduction priority for at least one second seat can be an operation that adjusts the noise reduction priority or weight coefficients for each seat displayed on the screen. For instance, this can be achieved by sliding a progress bar. The value range of the seat's weight coefficient is... .

[0155] In a practical implementation, the in-vehicle audio playback device may include speaker units arranged in the cabin ceiling, door panels, or seat backs. After the controller detects that a user has adjusted the priority of at least one second seat via the central control screen, voice commands, or seat interaction interface, it first converts the adjustment operation into a new priority value for the corresponding seat. Then, based on a pre-established priority-weighting coefficient mapping relationship, it generates new weighting coefficients and updates the input parameters used for adaptive filtering accordingly. After correcting the original noise reduction priority using the new weighting coefficients, the target control quantity is recalculated and a new anti-noise signal is generated. Subsequently, the audio playback device outputs noise suppression corresponding to the new anti-noise signal, adapting the residual noise distribution in the corresponding seat area to the user's new quiet requirements.

[0156] This method can promptly adjust the weight allocation and regenerate the anti-noise signal after the user changes the noise reduction requirements of the second seat, so that the key seats can obtain the noise reduction effect that matches the current preference. At the same time, it avoids the original control parameters from continuing to act on the changed usage scenario, thereby improving the response speed, scenario adaptability and resource utilization efficiency of in-vehicle active noise cancellation.

[0157] In this embodiment of the application, adjusting the noise reduction priority according to the new weighting coefficient includes: adjusting the noise reduction priority using a linear interpolation method according to the new weighting coefficient.

[0158] In this embodiment, the new weighting coefficient is used to characterize the user's reset result for the noise reduction intensity corresponding to at least one seat.

[0159] Linear interpolation refers to establishing a continuous transition relationship between the old and new noise reduction priorities, and determining the transition ratio based on the new weight coefficients, so that the adjusted priority converges smoothly with the weight changes, avoiding abrupt changes in control parameters.

[0160] Specifically, after obtaining the new weight coefficients, they can be mapped to the target priority value, and interpolated using the original priority value at the current moment to obtain the intermediate transition priority. If the change in the new weight coefficients is large, they can be updated sequentially in an incremental manner within a preset control period, ensuring that the priority output in each period is between the previous and subsequent states.

[0161] For example, the interpolation step size can be pre-configured, for example, a value within 100ms to 500ms, which is not limited in this embodiment of the application.

[0162] In this way, by employing linear interpolation, the noise reduction priority can smoothly transition from the original state to the new state and simultaneously affect subsequent adaptive filtering calculations. This makes the amplitude change of the noise reduction signal more continuous, and the noise suppression output of the in-vehicle speakers is less prone to abrupt changes. This allows for timely response to new weighting coefficients when a user adjusts the noise reduction needs of a particular seat, while maintaining stable control, reducing auditory fluctuations, and improving noise reduction adaptability and comfort in multi-seat scenarios.

[0163] In this application, users can also add different scene modes and noise reduction priority or weight coefficients for each seat in the scene mode.

[0164] The process of adding a scene mode by the user can include: receiving the user's input of the new scene mode and the new weight coefficient of each seat under the new scene mode; and adding the correspondence between the new scene mode and the new weight coefficient to the mapping relationship.

[0165] The newly added scene modes are scene configuration items used to expand the set of recognizable scenes for the vehicle, and can correspond to noise reduction scenes customized by users under different riding needs. The newly added weight coefficients are seat priority parameters that accompany the newly added scene modes, representing the weight distribution relationship of each seat in the active noise reduction process under this new scene. The mapping relationship is used to store the corresponding data between scene modes and the weight coefficients of each seat; the controller can directly query and call the corresponding noise reduction priority based on this relationship.

[0166] For example, the in-vehicle interface can include a scene editing entry point. Users can input a new scene mode name and configure new weight coefficients for the driver's seat, front passenger seat, rear left seat, and rear right seat. These weight coefficients can be generated through numerical input, sliding adjustment, or selection from a preset template. Upon receiving this input, the system performs uniqueness and format checks on the new scene mode and writes it along with its corresponding weight coefficients into a mapping table in the storage unit. This allows for direct retrieval of the corresponding seat weights when the scene mode is detected later. The mapping table can be stored in the non-volatile memory of the vehicle controller to ensure that user-defined configurations are retained even after the vehicle is powered off.

[0167] By adding user-inputted new scene modes and weight coefficients to the mapping relationship, the scene recognition range of the vehicle can be expanded. Active noise cancellation control is no longer limited to preset scenes, but can generate corresponding seat priorities according to user-defined needs. When a new scene mode is subsequently recognized, the mapping relationship can be directly invoked to determine the noise reduction weight of each seat, thus ensuring that the noise reduction signal generation process is consistent with the user's personalized needs.

[0168] In this way, the mapping relationship is scalable and configurable, which can support users to supplement and define noise reduction strategies under different driving habits, passenger distribution and comfort preferences. This reduces the adaptation limitations caused by insufficient preset scenarios, improves the response capability of the in-vehicle active noise reduction system to complex usage environments, and enhances the noise reduction targeting of key seats and the overall vehicle quietness experience, thus meeting users' personalized noise reduction needs.

[0169] In conjunction with the content of the above embodiments, Figure 4 A flowchart illustrating another in-vehicle noise reduction method provided in this application.

[0170] like Figure 4As shown, the method includes: the user selects a scenario through the HMI scenario selection interface (central control screen / voice); based on the user-selected scenario, the scenario is configured in the scenario mode library to obtain the weighted coefficients of each seat in the vehicle; the weighted coefficients are input into the weighted coefficient vector manager; the weighted coefficient vector manager also receives the current seat occupancy status detected by the seat occupancy detection module (pressure sensor / cabin camera) to generate a weighted coefficient vector. ( The noise reduction weights for the i-th seat are fed into a weighted FxLMS adaptive filter bank. This bank also receives reference signals from error microphone arrays (microphones at each seat's headrest), reference signals from reference microphone arrays (microphones at the doors, chassis, etc.), and corrected reference signals (i.e., reference signals) output by the secondary channel estimation model. These input signals are processed to obtain an anti-noise signal, which is then used to drive the speakers in the output layer (N speakers throughout the vehicle) to output the anti-noise signal.

[0171] Figure 5 This is a schematic diagram of a data interaction provided for this application.

[0172] like Figure 5 As shown, the seat sensor / camera reports sensor data to the seat occupancy detection module every 100ms.

[0173] The seat occupancy detection module infers the seat occupancy vector based on the reported data. When the seat occupancy status changes, a seat occupancy status change event △O is sent to the weighted coefficient vector manager.

[0174] The weighted coefficient vector manager queries the scene mode library manager for the current automatic mode rules.

[0175] The scene pattern library manager returns the mapping rule (O→α vector) from the placeholder vector to the weighted coefficient vector to the weighted coefficient vector to the weighted coefficient vector.

[0176] The weighted coefficient vector manager calculates the new weighted coefficient vector α_new according to the mapping rules and sends a switching instruction to the weighted FxLMS controller. The instruction includes the old weighted coefficient vector α_old, the new weighted coefficient vector α_new, and a transition time of 200ms.

[0177] The weighted FxLMS controller performs linear interpolation smooth transition calculations based on the switching command. The output noise-resistant signal based on the new α is sent to the speaker array.

[0178] In summary, the in-vehicle noise reduction method provided in this application has the following technical effects: by directly controlling the weight of each seat in the optimization objective through a weighted coefficient vector, users can configure personalized modes such as "driver's seat is the quietest" or "rear children are the quietest" as needed. Compared with the traditional equal weight scheme, additional noise reduction depth can be obtained in priority seats.

[0179] In the automatic seat occupancy detection mode, when there are only 1 to 2 passengers in the vehicle, the error constraint dimension of the ANC algorithm is reduced (from 4 dimensions to 1 to 2 dimensions), and the increased degrees of freedom allow occupied seats to achieve deeper noise suppression than in the all-occupancy mode.

[0180] The preset scene mode library allows users to switch with one click or automatically activate based on seat occupancy status, with a smooth switching time of ≤500ms, achieving an adaptive noise reduction strategy that is imperceptible when the driving and riding scenarios change.

[0181] Figure 6 This is a structural schematic diagram of an in-vehicle noise reduction device provided in this application, as shown below. Figure 6 As shown, the in-vehicle noise reduction device 60 provided in this embodiment includes:

[0182] The acquisition module 601 is used to acquire the vehicle's original noise signal and the current scene mode of the vehicle.

[0183] The determination module 602 is used to determine the noise reduction priority of each seat according to the scene mode and in a pre-set mapping relationship; the mapping relationship is the correspondence between the scene mode and the weight coefficient of each seat.

[0184] The processing module 603 is used to process the original noise signal through the secondary channel estimation model to obtain the reference signal; and to iteratively process the reference signal and the residual noise signal acquired in real time according to the noise reduction priority through an adaptive filtering algorithm to obtain the noise-resistant signal.

[0185] Output module 604 is used to output noise suppression signals corresponding to noise reduction signals through the in-vehicle sound playback device.

[0186] The in-vehicle noise reduction device provided in this embodiment can perform the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0187] Figure 7 This is a structural schematic diagram of a vehicle provided for this application. For example... Figure 7 As shown, the vehicle 70 provided in this embodiment includes: a vehicle body, at least one processor 701, and a memory 702. Optionally, the vehicle 70 also includes a communication component 703. The processor 701, memory 702, and communication component 703 are connected via a bus 704.

[0188] In the specific implementation process, at least one processor 701 executes computer execution instructions stored in memory 702, causing at least one processor 701 to execute the above-mentioned in-vehicle noise reduction method.

[0189] The specific implementation process of processor 701 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0190] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0191] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0192] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0193] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described in-vehicle noise reduction method.

[0194] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the aforementioned in-vehicle noise reduction method.

[0195] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0196] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0197] The division of units is merely a logical functional division; 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 indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0198] 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; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0199] In addition, the functional units in the various embodiments of the present invention 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.

[0200] If a function 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, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0201] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0202] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A method for reducing noise inside a vehicle, characterized in that, include: Acquire the vehicle's raw noise signal and the current scene mode in which the vehicle is located; Based on the scene mode, the noise reduction priority of each seat is determined in a pre-set mapping relationship; the mapping relationship is the correspondence between scene modes and the noise reduction priority of each seat. The original noise signal is processed by a secondary channel estimation model to obtain a reference signal; An anti-noise signal is obtained by iteratively processing the reference signal and the real-time acquired residual noise signal according to the noise reduction priority using an adaptive filtering algorithm. The noise suppression signal corresponding to the noise reduction signal is output through the sound playback device inside the vehicle.

2. The method according to claim 1, characterized in that, The noise reduction priority is the target weight coefficient; The step of iteratively processing the reference signal and the acquired real-time residual noise signal according to the noise reduction priority using an adaptive filtering algorithm to obtain an anti-noise signal includes: Based on the target weight coefficients, generate a secondary channel weighting coefficient vector; The weighted cost function in the adaptive filtering algorithm is determined based on the secondary channel weighted coefficient vector and the residual noise signal. The anti-noise signal is obtained by iteratively updating the weight vector of the adaptive weighted filter bank based on the weighted cost function and the reference signal using an adaptive filtering algorithm according to the weighted filter update formula.

3. The method according to claim 2, characterized in that, The step of determining the weighted cost function in the adaptive filtering algorithm based on the secondary channel weighted coefficient vector and the residual noise signal includes: The product of the secondary channel weighting coefficient vector and the residual noise signal is determined as the weighted cost function.

4. The method according to claim 1, characterized in that, The scene mode is selected by the user from multiple pre-stored scene modes, or is an automatic mode, or is a default scene mode obtained when the vehicle is powered on, or is a scene mode used before the vehicle was last powered off.

5. The method according to claim 2, characterized in that, If the scene mode is automatic mode, determining the noise reduction priority of each seat according to the scene mode and a pre-set mapping relationship includes: In a pre-set mapping relationship, the initial weight coefficient corresponding to the automatic mode is determined; the initial weight coefficient is the same for each seat. Obtain the occupancy status of each seat in the vehicle; The initial weight coefficients are adjusted according to the occupancy status of each seat to obtain the target weight coefficients; wherein the weight coefficients corresponding to seats in an occupied state are greater than the weight coefficients corresponding to seats in an unoccupied state.

6. The method according to claim 5, characterized in that, The occupancy status of each seat in the vehicle is determined based on the seat pressure signal detected by the pressure detection device installed at each seat, and / or the images of each seat acquired by the image acquisition device installed in the vehicle.

7. The method according to claim 2, characterized in that, The method further includes: For the first seat with a target weight coefficient of 0, when the residual noise signal collected in real time is greater than the preset value, its residual noise signal is determined as the preset value.

8. The method according to any one of claims 1-7, characterized in that, After the noise suppression corresponding to the noise reduction signal is output through the sound playback device in the vehicle, the method further includes: In response to a user's adjustment operation on the noise reduction priority corresponding to at least one second seat, a new noise reduction priority for the at least one second seat is obtained; The noise reduction priority is adjusted according to the new weighting coefficient, and a new anti-noise signal is determined based on the adjusted noise reduction priority. The new noise suppression signal is output through the sound playback device inside the vehicle, corresponding to the noise suppression.

9. The method according to claim 8, characterized in that, The step of adjusting the noise reduction priority according to the new weighting coefficient includes: Based on the new weighting coefficients, the noise reduction priority is adjusted using linear interpolation.

10. The method according to any one of claims 1-7, characterized in that, The method further includes: Receive user input of new scene modes and new weight coefficients for each seat under the new scene modes; The correspondence between the newly added scene mode and the newly added weight coefficient is added to the mapping relationship.

11. A vehicle interior noise reduction device, characterized in that, include: The acquisition module is used to acquire the vehicle's original noise signal and the current scene mode of the vehicle; The determination module is used to determine the noise reduction priority of each seat according to the scene mode and in a pre-set mapping relationship; the mapping relationship is the correspondence between the scene mode and the weight coefficient of each seat. The processing module is used to process the original noise signal through a secondary channel estimation model to obtain a reference signal; and to iteratively process the reference signal and the residual noise signal acquired in real time according to the noise reduction priority using an adaptive filtering algorithm to obtain an anti-noise signal. The output module is used to output the noise suppression signal corresponding to the noise reduction signal through the sound playback device in the vehicle.

12. A vehicle, characterized in that, include: Vehicle body, memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-10.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-10.

14. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method described in any one of claims 1-10.

Images