A method for auxiliary adjustment of parameters of a car audio

By acquiring the equivalent volume of the occupant and the spatial coordinates of the target ear, dynamic time alignment delay parameters and target gain compensation are generated, solving the problems of sound image positioning offset and frequency band energy imbalance in vehicle audio adjustment, and realizing the accuracy of sound field imaging and the continuity of auditory experience.

CN123075338APending Publication Date: 2026-11-10SUZHOU JINGLI ELECTRONICS
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611544603.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-08
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing in-vehicle audio adjustment technology fails to effectively combine the coupling effects of occupant body shape, dynamic sitting posture, and window opening on the in-vehicle sound field, resulting in sound image positioning deviation and frequency band energy imbalance, and easily producing digital steps and pops during audio playback.

Method used

By acquiring perception state data, calculating the occupant's equivalent volume and the target ear's spatial coordinates, generating dynamic time alignment delay parameters and target gain compensation, and smoothly adjusting sound field parameters when the underlying audio data stream meets the zero-crossing detection condition, thus avoiding transient noise caused by direct parameter overwriting.

Benefits of technology

It effectively suppresses in-car bass overshoot and interior panel resonance, improves the accuracy of sound field imaging, ensures the continuity of the listening experience, and eliminates transient noise during the dynamic adjustment of audio parameters.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123075338A_ABST
    Figure CN123075338A_ABST
Patent Text Reader

Abstract

The application relates to the technical field of vehicle-mounted audio processing, and discloses a vehicle-mounted audio parameter auxiliary adjustment method, which comprises the following steps: acquiring sensing data containing seat states, pull-out lengths, masses and window opening degrees; decoupling the pull-out lengths according to the seat states to obtain effective belt wrapping lengths, and calculating the equivalent volumes of passengers in combination with the mass data; solving target human ear spatial coordinates based on the seat states and the equivalent volumes to generate dynamic time alignment delay parameters; calculating acoustic shielding coupling parameters in combination with the window opening degrees and the equivalent volumes to generate target gain compensation amounts; and monitoring underlying audio streams, and when a zero-crossing detection condition is met, the current sound field parameters are smoothly adjusted at a preset step length. The application solves the problems of low-frequency overcompensation and transient noise caused by parameter mutation under the working condition of a window being opened, and improves the accuracy of sound field imaging and the smoothness of auditory transition in a vehicle.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of vehicle audio processing technology, specifically to a method for auxiliary adjustment of vehicle audio parameters. Background Technology

[0002] With the development of automotive intelligence, in-vehicle audio systems typically dynamically compensate for sound field parameters based on changes in the internal and external environment. In existing sound field adjustment solutions, the system often relies on single-dimensional vehicle status data to trigger fixed acoustic compensation logic. For example, when the windows are open, conventional systems directly call a preset gain curve for low-frequency compensation based solely on the physical proportion of the window's descent, without considering the damping variables of the actual physical space inside the vehicle. Since occupants of different sizes exert varying degrees of physical shielding on sound waves due to reflection and leakage paths, this fixed compensation method, ignoring the physical volume of physical obstacles, can easily lead to overcompensation in specific frequency bands, resulting in bass overshoot and resonance in the interior trim panels.

[0003] In terms of sound field time alignment calculation, traditional in-vehicle audio processing solutions generally use the factory-set standard driver's seat coordinates as a fixed spatial reference point. In practical applications, different drivers will significantly adjust the seat rail displacement and backrest tilt angle according to their own body shape, and the fixed spatial reference point cannot accurately map the actual three-dimensional position of the current occupant's ears. This will result in an unavoidable phase difference when the audio signals from each channel speaker reach the ear, causing the final sound field imaging to be offset.

[0004] At the execution level of low-level control parameters, existing audio systems typically overwrite the hardware registers of the audio processing chip directly upon receiving new sound field compensation parameters. Because this process is not time-domain aligned with the real-time audio data stream, when parameter jumps occur instantaneously when the upper-level audio signal is in a high-amplitude state, it causes a vertical voltage step discontinuity in the time domain of the digital audio waveform. This nonlinear abrupt change ultimately translates into an uncontrolled transient electrical signal impacting the speaker unit, generating noticeable transient noise during dynamic parameter switching and reducing the continuity of the listening experience. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an auxiliary method for adjusting in-vehicle audio parameters. This method solves the problems of existing in-vehicle audio adjustment technologies, which mostly use static parameters or single-dimensional state triggers without considering the coupling effects of occupant body shape, dynamic sitting posture, and window opening on the in-vehicle sound field, resulting in sound image positioning deviation and frequency band energy imbalance. Furthermore, directly switching sound effect parameters during audio playback can easily produce digital steps and popping sounds.

[0006] To address the above problems, the present invention provides the following technical solution: The first aspect of this invention provides a method for auxiliary adjustment of vehicle audio parameters, comprising: Acquire perception status data, which includes seat status data, extension length data, mass data, and window opening data; Spatial kinematic decoupling is performed on the pull-out length data based on the seat status data to calculate the effective webbing wrapping length, and the equivalent volume of the occupant is calculated by combining the mass data. The target ear spatial coordinates are calculated based on seat status data and occupant equivalent volume, and dynamic time alignment delay parameters are generated based on the target ear spatial coordinates. Based on the window opening data and the equivalent volume of the occupants, the basic low-frequency leakage compensation amount and the occupant shielding correction coefficient are calculated respectively to generate the target gain compensation amount. Monitor the underlying audio data stream. When the underlying audio data stream meets the zero-crossing detection condition, smoothly adjust the current sound field parameters to the dynamic time alignment delay parameter and the target gain compensation amount according to the preset update step size.

[0007] Furthermore, the seat status data includes slide rail displacement data and backrest rotation angle data; based on the seat status data, the pull-out length data is spatially kinematically decoupled to calculate the effective webbing wrapping length, including: based on the preset seat mechanical dimension relationship, slide rail displacement data, and backrest rotation angle data, calculating the change in the three-dimensional straight-line distance between the upper fixing point and the lower fixing point of the seat belt relative to the reference zero position; subtracting the change from the pull-out length data, the effective webbing wrapping length is obtained.

[0008] Furthermore, the equivalent volume of the occupants is calculated by combining the mass data, including: obtaining the preset multivariate polynomial model parameters, which include mass term coefficients, length square term coefficients, cross term coefficients, and constant terms; substituting the mass data and the effective webbing wrapping length into the multivariate polynomial model parameters, and outputting the equivalent volume of the occupants after numerical multiplication and addition operations.

[0009] Furthermore, the target ear spatial coordinates are calculated based on seat status data and occupant equivalent volume, including: using the factory-set baseline ear coordinates as the starting point and combining the seat status data to calculate the baseline ear spatial coordinates; comparing the occupant equivalent volume with the preset standard dummy reference volume to obtain the volume difference; and calculating the coordinate offset of the target ear spatial coordinates relative to the baseline ear spatial coordinates based on the volume difference and the preset ear position mapping coefficient, thereby updating the target ear spatial coordinates.

[0010] Furthermore, dynamic time alignment delay parameters are generated based on the target ear spatial coordinates, including: using the target ear spatial coordinates as a reference, calculating the spatial distance from each speaker to the target ear spatial coordinates; using a preset reference sound speed and combining the corresponding spatial distance of each speaker, calculating the dynamic time alignment delay parameters that each channel speaker needs to be set.

[0011] Furthermore, the target gain compensation amount includes high-frequency target gain compensation amount and low-frequency target gain compensation amount; based on the window opening data and the occupant equivalent volume, the basic low-frequency leakage compensation amount and the occupant shielding correction coefficient are calculated respectively to generate the target gain compensation amount, including: subtracting the occupant equivalent volume from the preset standard dummy reference volume to obtain the volume difference value, and using the preset high-frequency gain compensation coefficient in combination with the volume difference value to calculate the high-frequency target gain compensation amount; obtaining the current window opening area corresponding to the window opening data, using the ratio of the current window opening area to the maximum physical opening area as the window opening ratio, and calculating the basic low-frequency leakage compensation amount based on the window opening ratio; using the preset volume shielding coefficient in combination with the occupant equivalent volume to calculate the occupant shielding correction coefficient; and multiplying the basic low-frequency leakage compensation amount and the occupant shielding correction coefficient to obtain the low-frequency target gain compensation amount.

[0012] Furthermore, the underlying audio data stream includes a continuous PCM audio data stream. The criteria for determining the zero-crossing detection condition include: extracting the instantaneous amplitude values ​​of two consecutive sampling points in the PCM audio data stream; when the product of the instantaneous amplitude values ​​of the two consecutive sampling points is less than or equal to zero, the current waveform is determined to meet the zero-crossing detection condition.

[0013] Furthermore, the current sound field parameters are smoothly adjusted to the dynamic time alignment delay parameter and the target gain compensation amount according to the preset update step size, including: triggering a parameter stepping once each time the zero-crossing detection condition is met, and updating the current sound field parameters in the underlying register once; after multiple step updates after meeting the zero-crossing detection condition, until the current sound field parameters reach the corresponding dynamic time alignment delay parameter and the target gain compensation amount respectively, the closed-loop adjustment of the sound field parameters is completed.

[0014] Furthermore, after calculating the effective webbing wrapping length, the method also includes: retrieving pre-stored human body measurement feature boundary data, obtaining the lower and upper boundaries of the effective webbing wrapping length corresponding to the current quality data interval; determining whether the calculated effective webbing wrapping length is greater than or equal to the lower boundary and less than or equal to the upper boundary; if it is determined to be greater than or equal to the lower boundary and less than or equal to the upper boundary, then the sensor data is confirmed to be valid and the subsequent equivalent volume calculation is triggered.

[0015] Furthermore, when acquiring the perceived state data, the method also includes: acquiring the original pull-out length data and the original mass data; applying a first-order inertial low-pass filter to the original pull-out length data and the original mass data to eliminate interference caused by vehicle vibration, resulting in smoothed pull-out length data and mass data; using the smoothed pull-out length data and mass data as the basic parameters for subsequent spatial kinematic decoupling and volume calculation; and, after calculating the occupant equivalent volume, determining whether the calculated occupant equivalent volume is within a preset legal constraint range; if it is within the legal constraint range, the result is output as the effective volume equivalent.

[0016] A second aspect of the present invention provides a vehicle audio parameter auxiliary adjustment system, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the vehicle audio parameter auxiliary adjustment method as described in the first aspect of the present invention.

[0017] This invention provides an auxiliary method for adjusting vehicle audio parameters. It has the following beneficial effects: 1. This invention solves the problem of low-frequency overcompensation under open window conditions by coupling window opening data with the equivalent volume of the occupants to generate a target gain compensation amount. Traditional compensation strategies typically apply a fixed low-frequency gain directly based on the window opening ratio, which easily leads to bass overshoot. The present invention incorporates the equivalent volume of the occupants as an acoustic shielding parameter into the calculation, enabling the identification of differences in physical damping of sound wave leakage paths caused by occupants of different body sizes, and then adaptively correcting the basic low-frequency leakage compensation amount proportionally. This mechanism effectively suppresses standing wave overshoot and interior panel resonance phenomena occurring inside the vehicle under specific window opening conditions, maintaining the flatness of the low-frequency response.

[0018] 2. This invention utilizes seat state data to decouple the seatbelt extension length from spatial kinematics and combines this with mass data to calculate the target ear's spatial coordinates, thus improving the effectiveness of time alignment parameters. Compared to traditional in-vehicle audio processing solutions that use fixed spatial reference points, this invention quantifies the actual auditory spatial position differences of occupants of different heights and body types under different seat adjustment postures through effective webbing wrapping length and a multivariate polynomial model. The delay parameters generated based on these dynamic coordinates ensure that the audio signals of each speaker channel arrive at the current occupant's actual ear position in strict synchronization, improving the accuracy of sound field imaging under complex seating conditions.

[0019] 3. This invention employs a parameter smoothing update mechanism based on zero-crossing detection, eliminating transient noise generated during dynamic adjustment of audio parameters. During the parameter execution phase, this invention monitors the underlying audio data stream and only adds or subtracts the sound field parameters with a preset small step size when the instantaneous amplitude product of adjacent sampling points meets the zero-crossing condition. This constraint mechanism avoids the vertical voltage step discontinuity caused by traditional direct parameter overwrite logic at large amplitude waveforms, ensuring that gain or delay jumps occur near the zero potential where signal energy is weakest. This blocks the excitation path of nonlinear high-frequency harmonics, guaranteeing the continuity of auditory output during closed-loop parameter adjustment. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the architecture of the vehicle audio parameter auxiliary adjustment system of the present invention; Figure 2 This is a flowchart illustrating the auxiliary adjustment method for vehicle audio parameters according to the present invention. Figure 3 This is a timing control diagram of the data acquisition and time-domain filtering logic of the present invention; Figure 4 This is a schematic diagram of the spatial kinematics decoupling and feature verification process of the present invention; Figure 5 This is a schematic diagram illustrating the principle of the occupant physical volume equivalent calculation model of the present invention; Figure 6 This is a schematic diagram of the dynamic time alignment parameter generation logic of the present invention; Figure 7 This is a schematic diagram illustrating the coupling logic principle of frequency band gain compensation and acoustic masking in this invention. Figure 8 This is a timing diagram of the smooth update mechanism for control parameters of the present invention; Figure 9 This is a schematic diagram showing the comparison curves of the low-frequency response effect of the in-vehicle sound field according to the present invention; Figure 10 This is a schematic diagram comparing the time-domain waveforms before and after the smooth update of the control parameters in this invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] See attached document Figure 1 The present invention provides an auxiliary adjustment system for vehicle audio parameters, which may include: a sensing module, a domain control module and an audio processing module.

[0023] The sensing module is deployed in the vehicle's passenger compartment. The sensing module includes a gravity sensor located at the bottom of the seat pan, a Hall encoder integrated into the seat belt retractor, a slide rail position encoder and a backrest adjuster position encoder integrated into the seat adjustment mechanism, and a window Hall sensor located in the window lifting mechanism.

[0024] The gravity sensor outputs an electrical signal corresponding to the mass of the occupant in the seat. The sensing module obtains the occupant's mass data based on the pre-calibrated correspondence between the sensor output value and the mass value.

[0025] Hall encoders are used to detect the rotation of the seat belt retractor and output coded data that characterizes the length of the seat belt webbing pulled out; slide rail position encoders are used to output the slide rail displacement data of the seat along the vehicle's fore-and-aft direction; backrest adjuster position encoders are used to output the rotation angle data of the seat back relative to a preset reference position; and window Hall sensors are used to output coded data of the window lifting position corresponding to the window lifting position.

[0026] The sensing module converts the window lifting position coding data into window opening area data based on the pre-calibrated correspondence between the window lifting position coding value and the actual window opening area. The window opening area data is used as the window opening degree data in this embodiment.

[0027] The correspondence is pre-calibrated as follows: the car window is sequentially adjusted to multiple known lifting positions, and the position code value of the window Hall sensor corresponding to each lifting position is obtained. The actual opening area of ​​the window at the corresponding position is measured, and the position code value and the corresponding actual opening area of ​​the window are recorded in pairs. During operation, when the current window lifting position code value is between the code values ​​of two adjacent calibration positions, linear interpolation is performed based on the records corresponding to the two adjacent calibration positions to obtain the window opening area data corresponding to the current window lifting position.

[0028] The sensing module and the domain control module communicate via an onboard communication bus. The domain control module receives occupant mass data, seatbelt extension length data, slide rail displacement data, backrest rotation angle data, and window opening area data output by the sensing module.

[0029] The domain control module performs time-domain filtering, seat pose influence correction, human body measurement feature boundary determination, occupant equivalent volume calculation, human ear spatial position calculation, and speaker channel delay and frequency band gain calculation on sensor data to obtain a set of target control parameters for adjusting the in-vehicle audio output.

[0030] The domain control module and the audio processing module communicate via an audio control bus. The target control parameter set includes at least the dynamic time alignment delay parameter for each speaker channel, the target gain compensation amount for the high-frequency band, and the target gain compensation amount for the low-frequency band.

[0031] The audio processing module receives the target control parameter set sent by the domain control module, adjusts the audio delay of the corresponding speaker channel according to the dynamic time alignment delay parameter, and adjusts the gain of the corresponding frequency band according to the target gain compensation amount of the high frequency band and the target gain compensation amount of the low frequency band, so as to achieve auxiliary adjustment of the output parameters of each speaker channel.

[0032] See attached document Figure 2 This invention provides a method for auxiliary adjustment of vehicle audio parameters, comprising the following steps: S10, the domain control module synchronously acquires the raw data of occupant mass, raw data of seat belt extension length, slide rail displacement, backrest rotation angle and window opening area sent by the sensing module according to the preset sampling period, and performs first-order inertial low-pass filtering on the raw data of occupant mass and raw data of seat belt extension length to obtain smooth mass data and smooth extension length data.

[0033] S20, the domain control module determines the spatial relationship between the seat belt fixing points based on the slide rail displacement data and the backrest rotation angle data, calculates the change in distance between the seat belt fixing points corresponding to the current seat posture relative to the reference seat posture, calculates the effective webbing wrapping length data based on the smooth pull-out length data and the change in distance between the seat belt fixing points, and performs human body measurement feature boundary determination on the smooth quality data and the effective webbing wrapping length data based on the quality interval to which the smooth quality data belongs and the upper and lower boundaries of the effective webbing wrapping length corresponding to the quality interval.

[0034] S30, for the smooth quality data and effective webbing wrapping length data determined by the boundary of human body measurement features, the domain control module calculates the occupant equivalent volume of the corresponding seat according to the pre-calibrated multivariate polynomial model; the occupant equivalent volume is used to characterize the equivalent physical volume of the target occupant in the cabin space related to sound wave reflection, absorption and shielding.

[0035] S40, the domain control module calculates the reference human ear spatial coordinates based on the slide rail displacement data and backrest rotation angle data, and calculates the human ear coordinate offset based on the occupant equivalent volume and the pre-calibrated human ear position mapping coefficient; determines the target human ear spatial coordinates based on the reference human ear spatial coordinates and the human ear coordinate offset, and calculates the dynamic time alignment delay parameters corresponding to each speaker channel based on the target human ear spatial coordinates, forming a dynamic time alignment parameter set.

[0036] S50, the domain control module calculates the target gain compensation amount in the high-frequency band based on the occupant equivalent volume and the pre-calibrated high-frequency gain mapping relationship; calculates the basic low-frequency leakage compensation amount based on the window opening area data, and calculates the occupant shielding correction coefficient based on the occupant equivalent volume; and calculates the target gain compensation amount in the low-frequency band based on the basic low-frequency leakage compensation amount and the occupant shielding correction coefficient.

[0037] S60, the domain control module assembles the dynamic time alignment parameter set, the high-frequency target gain compensation amount, and the low-frequency target gain compensation amount into a target control parameter set, and sends the target control parameter set to the audio processing module; the audio processing module monitors the instantaneous amplitude values ​​of adjacent sampling points in the digital audio data stream, and when it detects that the digital audio data stream has reached or crossed zero potential, it gradually adjusts the current execution parameters according to the preset update step size until the current execution parameters reach the corresponding target control parameters.

[0038] See attached document Figure 3 The following specific implementation steps are provided for step S10.

[0039] S11, the domain control module generates a global synchronization trigger frame according to a preset sampling period and sends the global synchronization trigger frame to the perception module through the vehicle communication bus.

[0040] The preset sampling period is determined based on the data update period of each sensor and the communication scheduling period of the vehicle communication bus. Specifically, it is the shortest configurable period that can complete one data update of each sensor and complete transmission of the corresponding data within one sampling period, and is stored in advance as the configuration parameter of the corresponding vehicle model.

[0041] Each sensor node in the sensing module responds to the global synchronization trigger frame at the same discrete time step. Each detection data is synchronously latched. The domain control module obtains the current discrete time step. Corresponding raw data of occupant mass Original data on seatbelt pull-out length Slide rail displacement data Backrest rotation angle data and data on the opening area of ​​car windows .in, This represents the discrete time step determined according to a preset sampling period; Represents discrete time steps The corresponding raw data on occupant mass; Represents discrete time steps The corresponding raw data for the seatbelt pull-out length; Represents discrete time steps The corresponding slide rail displacement data; Represents discrete time steps The corresponding backrest rotation angle data; Represents discrete time steps The corresponding window opening area data.

[0042] Data transmission between the perception module and the domain control module can be encapsulated, verified, and parsed according to the communication protocol corresponding to the on-board communication bus used.

[0043] S12, the domain control module classifies the data obtained in step S11 according to the subsequent processing method.

[0044] For slide rail displacement data Backrest rotation angle data and data on the opening area of ​​car windows The domain control module stores it in the state data cache for subsequent calculations of seat pose, ear spatial position, and low-frequency gain compensation.

[0045] For raw data on occupant mass Raw data on seatbelt pull-out length Considering the vehicle's driving vibration and the data fluctuations caused by sensor sampling quantization, the domain control module treats it as occupant status data to be filtered and performs low-pass filtering in step S13.

[0046] S13, the domain control module processes the raw occupant mass data respectively. Raw data on seatbelt pull-out length Perform a first-order inertial low-pass filter. The domain control module reads the pre-stored length low-pass filter coefficients. and quality low-pass filter coefficients ,in, and All are dimensionless coefficients in the range of 0 to 1, used to determine the update weights of the current sampled data in the seat belt pull-out length filtering and occupant quality filtering, respectively.

[0047] During the vehicle calibration phase, standard loads of known mass and seatbelt test positions of known pull-out lengths were used as calibration benchmarks. Continuous sensor data was collected under both stationary and preset vibration conditions. Multiple candidate length and candidate mass low-pass filter coefficients were set within the 0-1 range according to the parameter configuration accuracy allowed by the domain control module. Each candidate filter coefficient was then used to filter the corresponding continuous sensor data. The average deviation between the filtering results corresponding to different candidate filter coefficients and the corresponding calibration benchmark values ​​was compared. The candidate filter coefficient with the smallest average deviation was determined as the corresponding length low-pass filter coefficient. or quality low-pass filter coefficient And it is pre-stored in the non-volatile memory of the domain control module.

[0048] The domain control module reads the previous discrete time step. Corresponding smooth pull-out length data And based on the current discrete time step Original data of corresponding seat belt pull-out length The smoothed pull-out length data corresponding to the current discrete time step is calculated according to the following formula. : ; In the formula, Indicates the current discrete time step The corresponding smooth pull-out length data; Indicates the previous discrete time step The corresponding smooth pull-out length data; Indicates the current discrete time step The corresponding raw data for the seatbelt pull-out length; This represents the length of the low-pass filter coefficient.

[0049] The domain control module reads the previous discrete time step. Corresponding smoothing quality data And based on the current discrete time step Corresponding raw data of occupant mass The smoothing quality data corresponding to the current discrete time step is calculated according to the following formula. : ; In the formula, Indicates the current discrete time step The corresponding smoothed quality data; Indicates the previous discrete time step The corresponding smoothed quality data; Indicates the current discrete time step The corresponding raw data on occupant mass; This represents the low-pass filter coefficient.

[0050] If the filtering result of the previous discrete time step is not available in the first discrete time step after system startup, the original data of seat belt pull-out length obtained in the first sampling period is used as the initial value of smooth pull-out length data, and the original data of occupant mass obtained in the first sampling period is used as the initial value of smooth mass data; starting from the next discrete time step, the recursive calculation is performed according to the above first-order inertial low-pass filtering method.

[0051] The domain control module extracts the smoothed length data obtained from the filtering process. and smoothing quality data The data is stored in the state data cache area for step S20 to correct the change in distance between seat belt anchor points and to determine the boundary of human body measurement features.

[0052] See attached document Figure 4 The following specific implementation steps are provided for step S20.

[0053] S21, the domain control module establishes a three-dimensional spatial coordinate system within the vehicle's passenger compartment and determines the positional relationship between the upper and lower fixing points of the seat belt in the three-dimensional spatial coordinate system.

[0054] Specifically, the seatbelt anchor point at the B-pillar of the vehicle body is set as a fixed coordinate point, and its coordinates are as follows: The lower anchor point of the seat belt, which moves with the seat position, is set as the follow-up coordinate point, and its coordinates are: The domain control module reads the slide rail displacement data from the status data buffer. and backrest rotation angle data And obtain the current discrete time step based on the predetermined mechanical dimensional relationship of the seat. The coordinates of the corresponding anchor point under the seat belt.

[0055] In this embodiment, the positional change of the lower fixing point of the seat belt along the Y-axis is relatively small compared to the X-axis and Z-axis directions, therefore... As predetermined fixed coordinate values, the coordinates of the anchor point under the seat belt in the X and Z axes are determined using the following position transformation relationship: ; ; In the formula, This indicates the X-axis coordinate of the lower anchor point of the seatbelt in the current seat position; This represents the Z-axis coordinate of the lower anchor point of the seatbelt in the current seat position; Indicates the current discrete time step The corresponding slide rail displacement data; Indicates the current discrete time step The corresponding backrest rotation angle data; This indicates the positional transformation relationship between the slide rail displacement data and the X-axis coordinate of the fixed point under the seat belt; This indicates the positional transformation relationship between the backrest rotation angle data and the Z-axis coordinate of the lower fixed point of the seat belt.

[0056] Position transformation relationship and The coordinates are obtained through vehicle seat mechanical dimension calibration and can be represented using a fixed proportional relationship or a one-dimensional lookup table. When using a one-dimensional lookup table, the seat is adjusted to multiple known slide rail positions and multiple known backrest angle positions. The X-axis and Z-axis coordinates corresponding to the seatbelt anchor point at each calibration position are measured, and the correspondence between slide rail displacement and X-axis coordinate, and between backrest angle and Z-axis coordinate are recorded. When the input value during operation lies between two adjacent calibration points, linear interpolation is performed based on the corresponding adjacent calibration data to obtain the corresponding coordinates.

[0057] S22, the domain control module uses the fixed coordinates of the fixed points on the seat belt. and the moving coordinates of the fixed point under the seat belt Calculate the spatial distance between two fixed points under the current seat posture, and determine the change in distance between the seat belt fixing points accordingly.

[0058] The domain control module reads the pre-stored zero-point reference distance. Zero-point reference distance This indicates the three-dimensional straight-line distance between the upper and lower anchor points of the seatbelt when the seat is in the factory-set reference position. It is obtained by adjusting the seat to the reference position and measuring the actual straight-line distance between the two anchor points.

[0059] Current discrete time step The corresponding change in distance between seat belt anchor points Calculate according to the following formula: ; In the formula, This indicates the change in distance between seatbelt anchor points relative to the current seat position compared to the baseline seat position. This indicates the fixed coordinates of the anchor points on the seatbelt; This indicates the coordinates of the lower anchor point of the seatbelt in the current seat position; This indicates the baseline straight-line distance between the upper and lower anchor points of the seatbelt in the reference seat position.

[0060] when When, it indicates that the distance between the upper and lower anchor points of the seat belt has increased relative to the reference seat posture in the current seat position; when When the time is specified, it indicates that the distance between the upper and lower anchor points of the seat belt is reduced relative to the reference seat posture under the current seat posture.

[0061] To ensure that the change in distance between seatbelt anchor points and the seatbelt pull-out length data use the same length change benchmark, in this embodiment, the webbing position when the seat is in a benchmark seat posture and the seatbelt is in a benchmark unwrapped state is used as the measurement benchmark for the seatbelt pull-out length. The seatbelt pull-out length data output by the Hall encoder represents the change in webbing pull-out length relative to the measurement benchmark.

[0062] The domain control module reads the smooth pull-out length data obtained in step S10. Calculate the effective webbing lining length data according to the following formula. : ; In the formula, This indicates the effective webbing length extending from the occupant's body after deducting changes in the distance between seatbelt anchor points caused by changes in seat position; Indicates the current discrete time step The corresponding smooth pull-out length data; This indicates the change in distance between seatbelt anchor points relative to the current seat position compared to the baseline seat position.

[0063] Based on the above calculations, corrections are made for the changes in seatbelt pull-out length caused by seat forward / backward movement and backrest angle changes, resulting in effective webbing wrapping length data for subsequent anthropometry feature boundary determination. .

[0064] S23, the domain control module based on smoothing quality data And effective webbing wrap length data Determine the boundaries of human anthropometric features.

[0065] The domain control module pre-stores the human body measurement feature boundary table. Human Measurement Feature Boundary Table Multiple mass ranges and the lower and upper boundaries of the effective webbing wrapping length corresponding to each mass range were recorded. The mass ranges were divided based on anthropometric data covering the 5th percentile of females to the 95th percentile of males, and were calibrated using standard dummies of corresponding body types on the target vehicle's seats.

[0066] Specifically, for each mass range, the mass data of a standard dummy falling within that mass range and the corresponding measured effective webbing wrapping length are obtained. The minimum effective webbing wrapping length in the calibration samples corresponding to that mass range is determined as the lower boundary of that mass range, and the maximum effective webbing wrapping length is determined as the upper boundary of that mass range. The mass range and its corresponding upper and lower boundaries of effective webbing wrapping length are recorded in the anthropometric feature boundary table. middle.

[0067] During operation, the domain control module uses the current smoothing quality data... Determine its quality range and judge the effective webbing wrapping length data. Whether it is located between the lower and upper boundaries of the effective webbing wrap length corresponding to this quality range.

[0068] When the effective webbing wrap length data When the data is located between the corresponding upper and lower boundaries, determine the current smoothing quality data. Data on effective webbing wrapping length Anthropometry feature boundary determination and smoothing quality data And effective webbing wrap length data The data is then passed to step S30 for the calculation of the occupant equivalent volume.

[0069] When the effective webbing wrap length data When the current quality range exceeds the upper and lower boundaries, it is determined that the current sensor combination data does not meet the pre-calibrated correspondence of human body measurement characteristics, and enters the failure degradation mode.

[0070] In failure degradation mode, the domain control module will use the current valid webbing wrap length data. Marked as invalid, it will no longer be used as input to the multivariate polynomial model in step S30, and will be adjusted based on the smoothing quality data. Query the pre-stored one-dimensional standby mapping table to obtain the estimated equivalent volume of degraded occupants under failure degradation mode. .

[0071] The one-dimensional backup mapping table is established using the standard acoustic dummy data used for multivariate polynomial model calibration in step S31, recording the correspondence between the mass of each standard acoustic dummy and the corresponding occupant equivalent volume. During operation, when the current smoothed mass data... When located between two adjacent mass calibration points, linear interpolation is performed based on the two adjacent mass calibration points and their corresponding occupant equivalent volumes to obtain the current smoothed mass data. Corresponding occupant equivalent volume .

[0072] in, Indicates the current discrete time step The corresponding smoothed quality data; Indicates the current discrete time step Corresponding effective webbing wrap length data; A human body measurement feature boundary table representing the upper and lower boundaries of the recorded quality range and the corresponding effective webbing wrapping length; This represents the occupant equivalent volume determined based on smoothed quality data under failure degradation mode.

[0073] The estimated value of the degraded occupant equivalent volume obtained in the failure degradation mode is used as a backup input to replace the normal occupant equivalent volume under abnormal conditions, and is used to continue to calculate the ear position and frequency band gain parameters in steps S40 and S50; in the normal mode, the occupant equivalent volume is still calculated in step S30 based on the smoothing quality data and the effective webbing wrapping length data.

[0074] See attached document Figure 5 The following specific implementation steps are provided for step S30.

[0075] S31, The domain control module obtains the valid webbing wrap length data transmitted in step S20. and smoothing quality data For data determined by the boundaries of anthropometric features, the domain control module reads pre-calibrated polynomial model parameters from non-volatile memory. The polynomial model parameters include mass term coefficients. Length squared term coefficient Cross term coefficient and constant terms .

[0076] The polynomial model parameters are obtained through vehicle model calibration. During calibration, multiple standard acoustic dummies that conform to the preset anthropometric data range are placed on the seats of the target vehicle model. The mass, effective webbing wrapping length, and known solid volume of each standard acoustic dummy are obtained, and the known solid volume of each standard acoustic dummy is used as the corresponding occupant equivalent volume calibration value.

[0077] Under the same vehicle acoustic test conditions, in-vehicle sound field sweep tests were conducted on standard acoustic dummies of different volumes to obtain the in-vehicle sound field response under different standard acoustic dummy settings. This was used to confirm that the standard acoustic dummy sample used covered the volume range in the target vehicle model that required occupant equivalent volume compensation.

[0078] Based on the mass, effective webbing wrapping length, and corresponding occupant equivalent volume calibration values ​​of each standard acoustic dummy, the following multivariate polynomial model is fitted using the least squares method to determine the mass term coefficients. Length squared term coefficient Cross term coefficient and constant terms The determined parameters are pre-stored in the non-volatile memory of the domain control module.

[0079] S32, the domain control module reads the polynomial model parameters and effective webbing wrapping length data in step S31. and smoothing quality data Calculate the current discrete time step Corresponding occupant equivalent volume Passenger equivalent volume Used to characterize the equivalent physical volume of a target occupant within the cabin space in relation to sound wave reflection, absorption, and shielding.

[0080] Passenger equivalent volume Calculate according to the following formula: ; In the formula, Indicates the current discrete time step The corresponding occupant equivalent volume; Indicates the current discrete time step The corresponding smoothed quality data; Indicates the current discrete time step Corresponding effective webbing wrap length data; Indicates quality item The corresponding mass term coefficient; Represents the term with length squared The corresponding coefficient of the squared length term; The term representing the intersection of mass and effective webbing wrap length. The corresponding cross term coefficients; This represents the constant term in a multivariate polynomial model.

[0081] in, , , γ and The values ​​are determined by least-squares fitting of the standard acoustic dummy calibration data from step S31. The dimensions of each coefficient are based on... , and The units of measurement used are determined so that each term in the formula corresponds to the equivalent volume of the occupants. They have consistent dimensions.

[0082] In a multivariate polynomial model, the mass term Used to incorporate the influence of occupant mass on the equivalent volume calculation results; length square term Used to incorporate the influence of occupant body size, as reflected by the effective webbing wrap length, on the calculation results; cross term This is used to incorporate the impact of variations in both occupant mass and effective webbing wrap length on the calculation results.

[0083] S33, The domain control module calculates the occupant equivalent volume in step S32. A valid constraint range determination is performed to confirm whether the calculated occupant equivalent volume is within the volume range covered by the calibration sample.

[0084] The domain control module reads the pre-stored upper limit of the occupant equivalent volume. Lower limit of equivalent volume for occupants . and Based on the establishment of the human body measurement feature boundary table in step S23 The volume data corresponding to the calibration sample used was determined, among which, Take the largest volume calibration value from the calibration sample. Take the smallest volume calibration value from the calibration sample.

[0085] in, This indicates the upper limit of the allowed output occupant equivalent volume. This represents the lower limit of the allowed output occupant equivalent volume, and both are related to the occupant equivalent volume. Use the same unit of volume measurement.

[0086] The domain control module will use the occupant equivalent volume calculated in step S32. and and Compare. When the calculation is obtained... Greater than At that time, Set as When calculated Less than At that time, Set as When calculated Greater than or equal to and less than or equal to At that time, the calculation result remains unchanged.

[0087] The effective volume output after the above-mentioned valid constraint range judgment The equivalent data is stored in the state data cache of the domain control module. It is used as input for steps S40 and S50, and is used for subsequent calculation of ear position and frequency band gain parameters.

[0088] See attached document Figure 6 The following specific implementation steps are provided for step S40.

[0089] S41, The domain control module acquires the slide rail displacement data from the status data buffer. and backrest rotation angle data Based on the pre-calibrated factory-referenced ear coordinates and the distance from the center of rotation of the seat back to the ear reference point, the reference ear spatial coordinates under the current seat posture are calculated.

[0090] The domain control module reads the pre-stored factory-referenced human ear coordinates. The distance from the center of rotation of the seat back to the reference point of the human ear. and the initial tilt angle of the seat back when it is in the factory zero position. Factory-standard human ear coordinates and The results were obtained based on the three-dimensional measurements of the 50th percentile standard dummy when the seat was in the factory zero position. Take the angle calibration value corresponding to the backrest adjuster position encoder when the seat is in the factory zero position.

[0091] Current discrete time step Corresponding reference human ear spatial coordinates Calculate according to the following formula: ; ; ; In the formula, , and These represent the current discrete time steps. The corresponding reference human ear X-axis, Y-axis, and Z-axis coordinates; , and These represent the X, Y, and Z axis reference coordinates of the ear of the 50th percentile standard dummy when the seat is in the factory zero position. Indicates the current discrete time step The corresponding slide rail displacement data; Indicates the current discrete time step The corresponding backrest rotation angle data; This indicates the initial tilt angle of the seat back when it is in the factory zero position. This indicates the distance from the center of rotation of the seat back to the 50th percentile reference point of the standard dummy's ear.

[0092] S42, the domain control module obtains the occupant equivalent volume output in step S30. The human ear coordinate offset is calculated based on the difference between the occupant's equivalent volume and the standard dummy's reference volume.

[0093] The domain control module reads the pre-stored standard dummy reference volume. X-axis ear position mapping coefficient Mapping coefficients of human ear position along the Z-axis .in, Take the volume calibration value of the 50th percentile standard dummy used in step S41 to ensure that the volume reference object is consistent with the calibration object of the zero-position human ear reference coordinate.

[0094] During the vehicle calibration phase, the actual X-axis and Z-axis coordinates of the human ear are measured for multiple standard acoustic dummies with known volumes used in step S31, under the same seat posture. Using the volume difference between each standard acoustic dummy and the 50th percentile standard dummy as input, and the corresponding differences in the X-axis and Z-axis coordinates of the human ear as output, linear fitting is performed, and the slope of the X-axis fitting is determined as follows. The slope of the Z-axis fitting direction is determined as .

[0095] in, and The unit can be mm / L, used to characterize the correction amount of the X-axis and Z-axis positions of the human ear for each unit change in the occupant's equivalent volume relative to the standard dummy reference volume. Current discrete time step Corresponding X-axis coordinate offset of human ear and Z-axis coordinate offset Calculate according to the following formula: ; ; In the formula, This represents the correction amount of the current occupant's ear X-axis coordinate relative to the 50th percentile standard dummy; This represents the correction amount of the current occupant's ear Z-axis coordinate relative to the 50th percentile standard dummy; Indicates the current discrete time step The corresponding occupant equivalent volume; The reference volume represents the 50th percentile standard dummy; and These represent the ear position mapping coefficients on the X and Z axes, respectively.

[0096] The domain control module calculates the target ear's spatial coordinates based on the reference ear spatial coordinates and the ear coordinate offset. : ; ; ; In the formula, , and These represent the current discrete time steps. The corresponding spatial coordinates of the target human ear along the X, Y, and Z axes.

[0097] Based on the above calculations, the X-axis and Z-axis positions of the human ear are corrected according to the occupant's equivalent volume, based on the reference ear position determined by the seat rail position and backrest angle, to obtain the target ear spatial coordinates for speaker channel delay calculation.

[0098] S43, the domain control module calculates the spatial distance between the target ear and each speaker based on the target ear's spatial coordinates.

[0099] The domain control module reads the pre-stored fixed spatial coordinates of each speaker. , where subscript Indicates the first Each speaker channel has a fixed spatial coordinate system. The fixed spatial coordinates of each speaker are predetermined based on the speaker installation position of the target vehicle model, which can be obtained from vehicle design data or actual vehicle 3D coordinate measurements.

[0100] No. Spatial distance between each speaker and the target ear Calculate according to the following formula: ; In the formula, Indicates the current discrete time step Next The spatial distance between each speaker and the target ear; Indicates the current spatial coordinates of the target human ear; Indicates the first Fixed spatial coordinates of a speaker.

[0101] S44, the domain control module determines the spatial distances corresponding to each speaker channel. Construct a distance set and determine the maximum spatial distance from the distance set. Maximum spatial distance The speaker channel that corresponds to the current target ear is furthest away.

[0102] Domain control modules with maximum spatial distance As a time alignment benchmark, the dynamic time alignment delay parameter of the corresponding speaker channel is calculated based on the difference between the spatial distance of each speaker channel and the maximum spatial distance. .

[0103] The domain control module reads the pre-stored reference speed of sound. Reference speed of sound The air velocity corresponding to the ambient temperature during acoustic parameter calibration of the target vehicle model is taken and stored as the acoustic configuration parameter for that model. When the spatial distance is expressed in millimeters, the reference sound velocity is used. The corresponding unit is millimeter per millisecond.

[0104] No. Dynamic time alignment delay parameters for each speaker channel Calculate according to the following formula: ; In the formula, Indicates the current discrete time step Next The required time delay for each speaker channel relative to the furthest speaker channel; This represents the maximum spatial distance from each speaker to the target ear; Indicates the first The spatial distance from the speaker to the target ear; This represents a pre-set reference speed of sound.

[0105] Among them, when the first Each speaker corresponds to At that time, its dynamic time alignment delay parameter The delay is set to 0; for speaker channels that are closer to the target ear, the corresponding delay is set according to the distance difference between them and the maximum spatial distance, so as to reduce the propagation time difference of the sound from each speaker channel to the target ear position.

[0106] The domain control module adjusts the dynamic time alignment delay parameters for each speaker channel. Composition of the current discrete time step The corresponding set of dynamic time alignment parameters is stored in the transmission buffer so that it can be sent to the audio processing module together with the high-frequency target gain compensation amount and the low-frequency target gain compensation amount in step S60.

[0107] See attached document Figure 7 The following specific implementation steps are provided for step S50.

[0108] S51, the domain control module obtains the occupant equivalent volume generated in step S30. And calculate the current discrete time step based on the pre-calibrated high-frequency gain mapping relationship. Corresponding high-frequency target gain compensation amount .

[0109] Considering that changes in occupant volume will alter the absorption and scattering of high-frequency sound waves inside the vehicle, in this embodiment, the difference between the current occupant's equivalent volume and the reference volume of a standard dummy is used as input, and the corresponding high-frequency target gain compensation amount is determined through a one-dimensional linear mapping.

[0110] High-frequency target gain compensation Calculate according to the following formula: ; In the formula, Indicates the current discrete time step The corresponding high-frequency target gain compensation amount; Indicates the current equivalent volume of the occupants; The reference volume of the standard dummy is taken as the volume calibration value of the 50th percentile standard dummy used in step S41. This represents the high-frequency gain compensation coefficient.

[0111] When the equivalent volume of the occupants is expressed in liters and the target gain compensation in the high-frequency band is expressed in decibels. The unit is decibels per liter (dB / L).

[0112] High-frequency gain compensation coefficient Obtained through acoustic calibration. During calibration, multiple standard acoustic dummies with known volumes (as described in step S31) are used to conduct high-frequency sound field tests under the same vehicle conditions, audio output conditions, and measurement positions. The volume difference between each standard acoustic dummy and the 50th percentile standard dummy is used as input, and the corresponding high-frequency sound pressure level attenuation difference is used as output for linear fitting. The resulting fitting slope is determined as the high-frequency gain compensation coefficient. And it is pre-stored in the domain control module.

[0113] S52, the domain control module obtains the window opening area data from step S10. The basic low-frequency leakage compensation amount is determined based on the window opening area, considering only the window opening state.

[0114] The domain control module reads the maximum physical opening area of ​​the window corresponding to the target vehicle model. The current window opening area data With the maximum physical opening area The ratio of the two values ​​is used as the window opening ratio. .in, The value range is from 0 to 1; the window opening ratio is 0 when the window is completely closed, and 1 when the window reaches its maximum physical opening area.

[0115] in, Indicates the current discrete time step The corresponding window opening area; This indicates the maximum physical opening area of ​​the corresponding car window; This indicates the ratio of the current window opening area to the maximum physical opening area.

[0116] The domain control module reads the pre-stored low-frequency compensation amount when the windows are fully open. . During the vehicle acoustic calibration phase, the following was determined: Under the baseline condition that there are no occupants obstructing the corresponding seats, while keeping the seat position and audio output state unchanged, the in-vehicle sound pressure level of the target low-frequency band was measured when the windows were fully closed and when the windows reached their maximum physical opening. Based on the difference in low-frequency sound pressure level between the two states, the low-frequency compensation amount corresponding to the fully open window state was determined.

[0117] This indicates the low-frequency gain compensation amount corresponding to the maximum physical opening area of ​​the car window, expressed in decibels.

[0118] The domain control module adjusts according to the window opening ratio. Low-frequency compensation amount when car windows are fully open The basic low-frequency leakage compensation amount is calculated according to the following formula. : ; In the formula, Indicates the current discrete time step The following is a basic low-frequency leakage compensation amount determined solely based on the window opening. Indicates the window opening ratio; This indicates the low-frequency compensation amount corresponding to when the car window reaches its maximum physical opening area.

[0119] S53, the domain control module generates the occupant equivalent volume based on step S30. The equivalent volume of the occupants is dimensionless, and based on a pre-defined volume masking mapping relationship, the occupant masking correction coefficient used to correct the basic low-frequency leakage compensation is calculated. .

[0120] During the vehicle calibration phase, multiple standard acoustic dummies with known volumes, as described in step S31, were used. While maintaining consistent window opening, seat position, and audio output status, the low-frequency acoustic energy leakage changes under each standard acoustic dummy's configuration were measured. Using the volume of the standard acoustic dummy as input and the low-frequency acoustic energy leakage attenuation ratio relative to the unoccupied reference state as output, a linear fit was performed. The slope of the fitted ratio change relative to the volume change was determined as the volume shielding coefficient. .

[0121] Current discrete time step Corresponding occupant shielding correction factor Calculate according to the following formula: ; In the formula, This represents the occupant shielding correction factor determined based on the current occupant equivalent volume; Indicates the volumetric occlusion factor; This represents the current equivalent volume of the occupants. The occupant shielding correction factor is used to adjust the basic low-frequency leakage compensation amount determined by the window opening.

[0122] The domain control module also pre-stores the lower limit of the occlusion correction coefficient. . Based on the low-frequency leakage calibration results of the aforementioned standard acoustic dummies of different volumes, the minimum positive occupant shielding correction coefficient corresponding to the condition where low-frequency acoustic energy leakage still exists in the calibration samples is taken as... .

[0123] When obtained from the above formula Less than At that time, Set as ;when Not less than At that time, the calculation result remains unchanged.

[0124] When the calibration results for the corresponding vehicle model in this embodiment meet the relevant conditions, For example, 0.1 can be taken.

[0125] S54, the domain control module is based on the basic low-frequency leakage compensation amount. And occupant shielding correction factor Calculate the current discrete time step Corresponding low-frequency target gain compensation amount : ; In the formula, Indicates the current discrete time step The corresponding target gain compensation amount in the low-frequency band; This indicates the basic low-frequency leakage compensation amount determined solely based on the current window opening. This represents the occupant shielding correction factor determined based on the current occupant equivalent volume.

[0126] Based on the above calculation relationship, when other conditions remain unchanged and the occupant equivalent volume increases, the occupant shading correction factor... The corresponding reduction will decrease the contribution of the basic low-frequency leakage compensation to the final low-frequency target gain compensation.

[0127] After completing the above calculations, the domain control module will compensate the high-frequency target gain. and low-frequency target gain compensation As a frequency band gain parameter, it is stored together with the dynamic time alignment parameter set generated in step S40 in the transmission buffer, so that step S60 can form a target control parameter set and send it to the audio processing module.

[0128] See attached document Figure 8 The following specific implementation steps are provided for step S60.

[0129] S61, the domain control module uses the dynamic time alignment delay parameters of each speaker channel generated in step S40. The high-frequency target gain compensation amount generated in step S50 and low-frequency target gain compensation The target control parameter set is formed and sent to the audio processing module via the audio control bus.

[0130] The audio processing module receives the target control parameter set and writes the parsed target parameters into the target parameter buffer. Among these parameters is the dynamic time alignment delay parameter. These are respectively used as the target delay values ​​for the corresponding speaker channels; and the target gain compensation for the high-frequency band. and low-frequency target gain compensation These are respectively used as the target gain adjustment values ​​for the corresponding high-frequency EQ band and low-frequency EQ band.

[0131] The high-frequency EQ band and low-frequency EQ band adopt the fixed EQ bands pre-configured by the audio processing module of the target vehicle model, and are pre-stored as audio configuration parameters of the corresponding vehicle model.

[0132] After the target control parameter set is written into the target parameter buffer, the currently executing delay and gain parameters are not immediately replaced. The currently executing delay and gain parameters are collectively referred to as the current sound field parameters. Instead, the audio processing module updates the current sound field parameters step by step according to the zero-crossing detection and step size update method in steps S62 to S64.

[0133] S62, the digital signal processor in the audio processing module reads the digital audio data stream point by point according to the preset audio sampling rate and obtains the current discrete audio sampling point. Corresponding instantaneous amplitude value .

[0134] in, Indicates the sequence number of discrete sample points in a digital audio data stream; Indicates the first The instantaneous amplitude value corresponding to each sampling point. In this embodiment, the underlying audio data stream adopts a continuous PCM audio data stream, and its sampling bit depth is determined according to the configuration of the target vehicle's audio system.

[0135] To reduce transient noise caused by sudden changes in delay or gain parameters when the audio signal amplitude is large, the audio processing module performs zero-crossing detection on the instantaneous amplitude values ​​of adjacent sampling points and determines whether to perform the parameter update based on the detection results.

[0136] S63, the audio processing module reads the instantaneous amplitude value of the previous sampling point. and the instantaneous amplitude value of the current sampling point The zero-crossing detection flag is determined based on the sign relationship between the two. .

[0137] When the following equation is satisfied: ; The audio processing module determines whether the digital audio signal reaches or crosses zero potential between two adjacent sampling points and sets the zero-crossing detection flag accordingly. Set to 1; when At that time, the zero-crossing detection flag will be... Set to 0.

[0138] in, Indicates the first The instantaneous amplitude value corresponding to each sampling point; Indicates the first The instantaneous amplitude value corresponding to each sampling point; Indicates the first The zero-crossing detection flag corresponding to each sampling point This indicates that the zero-crossing condition for this parameter update is met. This indicates that the zero-crossing condition for this parameter update is not met.

[0139] S64, zero-crossing detection flag bit At that time, the audio processing module updates the current sound field parameters one step at a time towards the corresponding target parameter direction, according to a pre-set update step size; when At the same time, keep the current sound field parameters unchanged.

[0140] For any control parameter, let the corresponding target parameter in the target parameter buffer be... The current parameters executed at the previous sampling point are The execution parameters after this update are: The maximum single update step size for the corresponding control parameters is .

[0141] The execution parameters will be updated according to the following formula: In the formula, This indicates the target parameter corresponding to the current control parameter; This indicates the current execution parameters corresponding to the previous sampling point; Indicates the first Execution parameters updated for each sampling point; Indicates the zero-crossing detection flag; This indicates the maximum allowable single update step size for the corresponding control parameter; Represents a symbolic function; This indicates that the smaller value is selected during the operation.

[0142] when When the sign function outputs 1, the function outputs 1; when When the two are equal, the sign function outputs -1; when the two are equal, the sign function outputs 0.

[0143] When the absolute value of the difference between the target parameter and the current execution parameter is greater than or equal to At that time, the absolute value of the parameter update was... When the absolute value of the difference is less than When this parameter is updated, the absolute value of the difference is taken, so that the updated execution parameter reaches the corresponding target parameter.

[0144] Set corresponding update step sizes for the dynamic time alignment delay parameter, high-frequency band gain parameter, and low-frequency band gain parameter, respectively. The update step size unit and value for different control parameters are determined separately. The update step size for the delay parameter is in time unit, and the update step size for the gain parameter is in decibel unit.

[0145] Each update step size is determined through vehicle model calibration. With the digital signal processor's processing cycle and audio sampling conditions remaining constant, multiple candidate update step sizes are set for delay parameters, high-frequency gain parameters, and low-frequency gain parameters, respectively. Transient changes in the speaker output signal are measured when the corresponding parameters change. Using the upper limit of transient noise allowed in the target vehicle's audio system's established sound quality test specifications as a constraint, the candidate value with the largest value among the candidate update step sizes that satisfy the constraint is selected as the update step size for the corresponding control parameter and written into the audio processing module's configuration parameters.

[0146] The audio processing module will update the execution parameters. Write the data to the control register of the corresponding audio processing channel. As the subsequent digital audio data stream continues to meet the zero-crossing condition, the audio processing module repeatedly performs step size updates in the above manner until the current execution parameter reaches the corresponding target parameter.

[0147] To aid in understanding the present invention, a specific numerical calculation process is provided below, taking into account specific occupant body types and vehicle conditions.

[0148] The application scenario is set as follows: an adult male driver who is 180cm tall and weighs 75kg sits in the driver's seat. The vehicle is in motion, the driver adjusts the seat forward by 50mm, tilts the backrest back by 20°, and lowers the driver's side window by 30%.

[0149] The domain control module acquires the underlying data sent by the sensing module according to a preset sampling period. After first-order inertial low-pass filtering to eliminate interference from vehicle vibration, the smoothed mass data is 75kg, and the smoothed pull-out length data is 960mm. Simultaneously, the domain control module reads the status data buffer, obtaining the current slide rail displacement data as +50mm, the backrest rotation angle data as 20°, and the window opening area data as 0.06m². 2 .

[0150] During the spatial kinematics decoupling and feature verification phase, the domain control module calculates, based on pre-determined seat mechanical dimensions, that the three-dimensional straight-line distance between the upper anchor point and the lower anchor point of the seatbelt in the current seat posture is 15mm shorter than the baseline zero position. The domain control module subtracts this negative change of -15mm from the smoothed length data of 960mm, calculating the effective webbing wrapping length as 975mm. Based on human anthropometric feature boundary determination, the upper and lower boundaries of the effective webbing wrapping length corresponding to a 75kg weight range are 910mm to 1050mm. The current calculated value of 975mm falls within this valid range, and the system determines the sensor data to be valid.

[0151] During the occupant physical volume equivalent calculation phase, the domain control module retrieves polynomial model parameters from non-volatile memory. The mass term coefficient is set to 0.4, the length square term coefficient to 0.00002, the cross term coefficient to 0.00022547, and the constant term to 12. The domain control module substitutes 75kg and 975mm into the multivariate polynomial model, and after numerical multiplication and addition operations, calculates the current driver's occupant equivalent volume to be approximately 77.5L. This volume value falls within the preset legal constraint range of 40L to 120L and is directly output as the effective volume equivalent.

[0152] During the dynamic time alignment parameter generation phase, the domain control module uses the factory-set baseline ear coordinates as a starting point, combined with a 50mm forward movement of the slide rail and a 20° backrest tilt, to calculate the baseline ear spatial coordinates. Subsequently, the domain control module retrieves the standard dummy reference volume of 70L, determining that the current occupant's equivalent volume is 7.5L larger than the reference value. Based on the preset X-axis and Z-axis ear position mapping coefficients, the target ear spatial coordinates are calculated to be offset backward by 12mm and downward by 8mm from the baseline position. Using the updated target ear spatial coordinates as a reference, the domain control module calculates the spatial distance to the farthest right rear door speaker to be 1650mm, and the spatial distance to the nearest left front A-pillar tweeter to be 780mm. Using a reference sound velocity of 0.34mm / µs, the domain control module calculates that the required dynamic time alignment delay parameter for the left front A-pillar tweeter is approximately 2.56ms.

[0153] During the frequency band gain compensation and acoustic masking coupling stage, for the high-frequency band, the domain control module uses a high-frequency gain compensation coefficient of 0.12dB / L, combined with a volume difference of 7.5L, to calculate the target gain compensation amount for the high-frequency band as +0.9dB. For the low-frequency band, the maximum known physical opening area is 0.2m². 2The current window opening ratio is 30%. Based on the preset low-frequency compensation of 12dB when the window is fully open, the basic low-frequency leakage compensation is calculated to be 3.6dB. Subsequently, the domain control module uses 0.28 as the volumetric shielding coefficient, and combined with the occupant equivalent volume of 77.5L, calculates the occupant shielding correction coefficient to be 0.69. Multiplying the basic low-frequency leakage compensation of 3.6dB by the shielding correction coefficient of 0.69 yields the final low-frequency target gain compensation of +2.48dB.

[0154] During the smooth update phase of control parameters, the domain control module sends the target control parameter set, consisting of the aforementioned 2.56ms delay, +0.9dB high-frequency gain, and +2.48dB low-frequency gain, to the audio processing module. The audio processing module monitors the underlying PCM audio data stream. When the product of the instantaneous amplitude values ​​of two consecutive sampling points is less than or equal to zero, it triggers a parameter step. If the update step size of the low-frequency gain is set to 0.1dB, the audio processing module will, after detecting 25 consecutive waveform zero-crossings, perform 25 small accumulations. The first 24 updates are performed with an update step size of 0.1dB each, and the 25th update is performed with 0.08dB based on the remaining difference between the current low-frequency gain and the target low-frequency gain. This smoothly adjusts the low-frequency gain in the underlying register to +2.48dB, completing the closed-loop adjustment of the entire sound field parameters.

[0155] See attached document Figure 9 To verify the authenticity and effectiveness of the coupling logic between mid-frequency gain compensation and acoustic masking in this invention, a real vehicle acoustic measurement experiment was conducted in a semi-anechoic chamber environment.

[0156] The experiment used a target vehicle equipped with a standard multi-channel in-vehicle audio system. A standard acoustic dummy with simulated human acoustic impedance characteristics was placed in the driver's seat, with an equivalent volume of 85L. The driver's side window was 50% open, while the other windows were closed. The test excitation source was full-frequency pink noise. The in-vehicle sound field response signal was collected using microphones positioned at the dummy's ears, and the amplitude frequency response curves for the low-frequency band from 20Hz to 200Hz were extracted.

[0157] The figure contains two amplitude frequency response curves. The first curve represents the response using a traditional fixed-proportion compensation strategy, which directly compensates for the corresponding low-frequency gain based solely on a 50% window opening. Test data shows that in the VSWR-sensitive frequency band from 60Hz to 90Hz, the first curve exhibits a significant amplitude spike, with the peak value overshooting the target flat response by approximately 4.2dB, resulting in a low-frequency overload response.

[0158] The second curve represents the response result using the coupling logic compensation strategy of this invention. Because this invention incorporates the occupant equivalent volume into the calculation, the system identifies the 85L solid volume as physically damping the sound wave leakage path, automatically calculates and applies an occupant shielding correction coefficient less than 1, thereby proportionally reducing the basic low-frequency leakage compensation. Test data shows that the second curve exhibits a smoother amplitude across the entire test frequency band, effectively suppressing gain overshoot in the 60Hz to 90Hz range, with the maximum measured deviation reduced to within 0.8dB. The above experimental comparison demonstrates that the coupling strategy of this invention avoids the low-frequency overcompensation defect under specific operating conditions and improves the flatness of the low-frequency sound field response.

[0159] See attached document Figure 10 To verify the effect of the zero-crossing detection-based parameter smoothing update mechanism on suppressing step change noise, digital audio time-domain waveform data of the audio processing module's underlying pipeline at the moment of parameter switching were extracted.

[0160] In the experiment, a gain jump command with an increment of 3dB was manually sent to the audio processing module, and the voltage amplitude evolution at the speaker channel output was recorded within 50ms before and after the command was sent. In the figure, the horizontal axis represents discrete time, and the vertical axis represents the transient voltage amplitude of the audio signal.

[0161] The upper part of the diagram shows the waveform state using a traditional direct overwrite mechanism. At 20ms, the system directly writes the new gain parameter into the execution register. At this time, the audio waveform is at the peak of a large amplitude output, and the instantaneous jump in the parameter causes a vertical voltage step fault in the time domain. This fault excites a large number of high-frequency harmonic components in the frequency domain, causing uncontrolled transient impacts on the speaker cone and resulting in nonlinear noise in the speaker output.

[0162] The lower half of the figure shows the waveform state using the zero-crossing detection smooth update mechanism of this invention. After issuing the incremental command at the same 20ms point, the audio processing module does not immediately execute parameter overwriting, but continuously monitors the amplitude product of adjacent sampling points. When the time domain waveform crosses the zero potential reference line, the system performs a gradual update according to a preset 0.1dB step size. The waveform recording shows that the entire gain amplification process is distributed across multiple subsequent waveform zero-crossing points. Since each gain jump occurs near the zero potential where the signal energy is relatively weak, and the single step amount is extremely small, the output waveform maintains good continuity without voltage step discontinuity. The above waveform comparison shows that the step approximation execution logic disclosed in this embodiment of the invention eliminates step abrupt noise during the dynamic parameter switching process, achieving a smooth transition in acoustic parameter adjustment.

[0163] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for auxiliary adjustment of vehicle audio parameters, characterized in that, include: Acquire sensing status data, which includes seat status data, pull-out length data, mass data, and window opening data; Based on the seat status data, the pull-out length data is spatially kinematically decoupled to calculate the effective webbing wrapping length, and the occupant equivalent volume is calculated by combining the mass data. The target ear spatial coordinates are calculated based on the seat status data and the occupant equivalent volume, and a dynamic time alignment delay parameter is generated based on the target ear spatial coordinates. Based on the window opening data and the occupant equivalent volume, the basic low-frequency leakage compensation amount and the occupant shielding correction coefficient are calculated respectively to generate the target gain compensation amount. Monitor the underlying audio data stream, and when the underlying audio data stream is detected to meet the zero-crossing detection condition, smoothly adjust the current sound field parameters to the dynamic time alignment delay parameter and the target gain compensation amount according to the preset update step size.

2. The method for auxiliary adjustment of vehicle audio parameters according to claim 1, characterized in that, The seat status data includes slide rail displacement data and backrest rotation angle data; The step of decoupling the pulled-out length data spatially from the seat state data and calculating the effective webbing wrapping length includes: Based on the preset seat mechanical dimensions, the slide rail displacement data, and the backrest rotation angle data, calculate the change in the three-dimensional straight-line distance between the upper fixing point of the seat belt and the lower fixing point of the follower relative to the reference zero position. Subtract the change from the pulled-out length data to obtain the effective webbing wrapping length.

3. The method for auxiliary adjustment of vehicle audio parameters according to claim 1, characterized in that, The calculation of the occupant equivalent volume based on the mass data includes: Obtain preset multivariate polynomial model parameters, which include mass term coefficients, length square term coefficients, cross term coefficients, and constant terms; Substitute the mass data and the effective webbing wrapping length into the parameters of the multivariate polynomial model, and output the equivalent volume of the occupants after numerical multiplication and addition operations.

4. The method for auxiliary adjustment of vehicle audio parameters according to claim 1, characterized in that, The step of calculating the target ear's spatial coordinates based on the seat state data and the occupant's equivalent volume includes: Starting from the factory-standard human ear coordinates, the reference human ear spatial coordinates are calculated by combining the seat status data; The equivalent volume of the occupant is compared with the preset reference volume of a standard dummy to obtain the volume difference. Based on the volume difference and the preset ear position mapping coefficient, the coordinate offset of the target ear spatial coordinates relative to the reference ear spatial coordinates is calculated, thereby updating the target ear spatial coordinates.

5. The method for auxiliary adjustment of vehicle audio parameters according to claim 4, characterized in that, The generation of dynamic time alignment delay parameters based on the target ear spatial coordinates includes: Using the target ear's spatial coordinates as a reference, calculate the spatial distance from each speaker to the target ear's spatial coordinates; Using a preset reference sound velocity and the corresponding spatial distance of each speaker, the dynamic time alignment delay parameter that needs to be set for each channel speaker is calculated.

6. The method for auxiliary adjustment of vehicle audio parameters according to claim 1, characterized in that, The target gain compensation amount includes high-frequency target gain compensation amount and low-frequency target gain compensation amount; The step of calculating the fundamental frequency band gain compensation and the occupant shielding correction coefficient based on the window opening data and the occupant equivalent volume, and generating the target gain compensation, includes: The volume difference is obtained by subtracting the equivalent volume of the occupant from the preset standard dummy reference volume. The target gain compensation amount in the high-frequency band is calculated by combining the preset high-frequency gain compensation coefficient with the volume difference. Obtain the current window opening area corresponding to the window opening data, take the ratio of the current window opening area to the maximum physical opening area as the window opening ratio, and calculate the basic low-frequency leakage compensation amount based on the window opening ratio. The occupant ... The low-frequency leakage compensation amount is obtained by multiplying the basic low-frequency leakage compensation amount by the occupant shielding correction coefficient.

7. The method for auxiliary adjustment of vehicle audio parameters according to claim 1, characterized in that, The underlying audio data stream includes a continuous PCM audio data stream, and the criteria for determining the zero-crossing detection condition include: Extract the instantaneous amplitude values ​​of two consecutive sampling points in the PCM audio data stream; When the product of the instantaneous amplitude values ​​of two consecutive sampling points is less than or equal to zero, the current waveform is determined to meet the zero-crossing detection condition.

8. The method for auxiliary adjustment of vehicle audio parameters according to claim 7, characterized in that, The step of smoothly adjusting the current sound field parameters to the dynamic time alignment delay parameter and the target gain compensation amount according to the preset update step size includes: Each time the zero-crossing detection condition is met, a parameter step is triggered to update the current sound field parameters in the underlying register by a step size. After multiple zero-crossing detection condition step size updates, the current sound field parameters are adjusted until they reach the corresponding dynamic time alignment delay parameter and the target gain compensation amount, thus completing the closed-loop adjustment of the sound field parameters.

9. The method for auxiliary adjustment of vehicle audio parameters according to claim 1, characterized in that, After calculating the effective webbing wrapping length, the method further includes: Retrieve pre-stored human body measurement feature boundary data to obtain the lower and upper boundaries of the effective webbing wrapping length of the interval corresponding to the current quality data; Determine whether the calculated effective webbing wrapping length is greater than or equal to the lower boundary and less than or equal to the upper boundary; if it is determined to be greater than or equal to the lower boundary and less than or equal to the upper boundary, then confirm that the sensor data is valid and trigger subsequent equivalent volume calculation.

10. The method for auxiliary adjustment of vehicle audio parameters according to claim 1, characterized in that, When acquiring the perceived state data, the method further includes: Obtain the raw pull-out length data and raw mass data; The original pull-out length data and the original mass data are processed by a first-order inertial low-pass filter to eliminate the interference caused by vehicle vibration, resulting in smoothed pull-out length data and mass data; and the smoothed pull-out length data and mass data are used as the pull-out length data and mass data in claim 1 for subsequent calculations. Furthermore, after calculating the occupant equivalent volume, it is determined whether the calculated occupant equivalent volume is within a preset legal constraint range; if it is within the legal constraint range, the result is output as the effective volume equivalent.