A hub air tightness detection data real-time analysis and leakage point positioning method and system

Abstract

The application discloses a kind of wheel hub airtightness detection data real-time analysis and leakage point positioning method and system, specifically relates to acoustic leak detection technical field, for solving the problem that wheel hub airtightness detection is caused by the virtual source of the impact of the inner wall of the enclosure and leads to leakage point positioning deviation;By setting reference sound sensor on the inner wall of the enclosure and setting multiple measurement sound sensors in the ring cavity, synchronous acquisition is carried out, mutual power spectrum analysis is carried out on reference sound signal and measurement sound signal, and instantaneous phase difference plane is constructed, phase step mutation and impulse wave packet arrival time alignment structure sound component are identified, direct impact structure sound component is separated based on the geometric distance relationship between the inner wall of the enclosure and the outer wall of the wheel hub and filtered out, and residual sound signal is used as leakage hole jet direct wave to carry out time difference positioning, so as to eliminate the interference of virtual source on positioning result, and direct the position of real leakage hole to leakage point positioning.

Description

Technical Field

[0001] This invention relates to the field of acoustic leak detection technology, and more specifically, to a method and system for real-time analysis of wheel hub airtightness detection data and location of leak points. Background Technology

[0002] In the final inspection stage of automated production lines for automotive aluminum alloy wheels, airtightness testing is a necessary step to ensure driving safety. Existing automated testing equipment typically places the wheel hub between upper and lower pressure plates for sealing, fills the inner cavity with compressed air and maintains pressure, and determines the presence of leaks by monitoring the pressure decay within the cavity. For wheel hubs identified as having leaks, further determination of the leak's exact location is necessary for subsequent repairs. Currently, integrating acoustic sensor arrays into the equipment and employing beamforming or time-of-flight algorithms for online location of the leak's sound source is the main alternative to manual soap bubble inspection. This approach typically assumes that the sound waves generated by the leak radiate directly outward from the leak hole to the sensor, and calculates the sound source location based on an open-space sound propagation model.

[0003] However, in actual wheel hub airtightness testing equipment, in order to shorten the cycle time and protect the sensors, a shroud or protective shell is often set around the wheel hub, thus forming a relatively narrow confined cavity between the outer wall of the wheel hub and the equipment housing. Under this closed or semi-closed structure, the high-speed airflow generated by the leak is ejected from the micro-hole and directly impacts the inner wall of the shroud or the outer wall of the rim within a very short distance. The resulting sound field is no longer composed solely of the jet noise from the leak hole itself, but is mixed with additional noise generated by the airflow impacting the solid wall and structural sound radiation caused by wall vibration. Existing sound source localization methods based on the open space propagation assumption, after receiving these mixed sound wave signals generated by different physical mechanisms and originating from different spatial locations, will concentrate the energy on the virtual source formed by the impact effect, rather than pointing to the actual leak hole location, causing a systematic deviation in the leak point location. This results in subsequent repair processes being unable to align with the true defect location, affecting the wheel hub's rework pass rate and production efficiency. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a method and system for real-time analysis of wheel hub airtightness detection data and leakage point location to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for real-time analysis of wheel hub airtightness test data and location of leak points includes the following steps: S1: A reference acoustic sensor is installed on the inner wall of the enclosure of the wheel hub airtightness testing equipment, and multiple measuring acoustic sensors are installed in the annular cavity between the wheel hub and the enclosure to simultaneously collect the reference acoustic signal and the measuring acoustic signal during the pressure holding stage. S2: Perform cross-power spectrum analysis on the reference acoustic signal and each measured acoustic signal to obtain the cross-power spectral density of each measurement channel; S3: Construct the instantaneous phase difference plane of each measurement channel based on the phase spectrum of the cross power spectral density, identify the time-frequency component on the instantaneous phase difference plane where the phase changes in a step-like manner and the time of the change is aligned with the arrival time of the shock wave packet in the reference acoustic signal, and identify it as the structural acoustic component caused by the impact of the leaked airflow on the inner wall of the enclosure. S4: Based on the geometric distance relationship between the inner wall of the enclosure and the outer wall of the hub, the component in the structural acoustic component whose phase difference matches the geometric distance relationship is separated into the direct impact structural acoustic component. S5: Filter out the direct impact structural acoustic component from each measured acoustic signal to obtain the residual acoustic signal; S6: The residual acoustic signal is used as the direct wave of the leak hole jet. The spatial coordinates of each measuring acoustic sensor are used for time difference positioning to calculate the location of the leak hole.

[0006] Furthermore, S1 includes: The reference acoustic sensor is rigidly mounted on the inner wall of the enclosure in the area directly opposite the outer wall of the wheel hub. Multiple acoustic sensors are arranged at intervals along the circumferential direction of the outer wall of the wheel hub within the annular cavity; After the wheel hub is fully inflated and enters the pressure holding stage, the sound signals generated by the leakage are collected synchronously by a reference sound sensor and multiple measurement sound sensors to obtain the reference sound signal and the measurement sound signal respectively.

[0007] Furthermore, S2 includes: Segmented windowing and Fourier transform are applied to the reference acoustic signal and the individual measured acoustic signal; Calculate the cross-power spectrum of the reference acoustic signal and the measured acoustic signal in the frequency domain; The cross power spectrum of multiple segments is averaged to obtain the cross power spectral density of the measurement channel, and the cross power spectral density of all measurement channels is obtained by traversing each measurement acoustic signal.

[0008] Furthermore, S3 includes: The instantaneous phase difference of each measurement channel is extracted from the phase spectrum of the cross power spectral density; A plane of instantaneous phase difference is constructed on a two-dimensional plane consisting of time and frequency. Continuous wavelet transform is performed on the phase spectrum of the cross-power spectral density to obtain the instantaneous phase at different center frequencies. Arranged according to center frequency and time to form an instantaneous phase difference plane; On the instantaneous phase difference plane, the time-frequency components whose phase jumps between adjacent time-frequency points exceed a preset angle and whose jump times coincide with the arrival time of the impulse wave packet in the reference acoustic signal are identified as structure acoustic components.

[0009] Furthermore, on the instantaneous phase difference plane, the time-frequency components whose phase jumps between adjacent time-frequency points exceed a preset angle and whose jump times coincide with the arrival time of the shock wave packet in the reference acoustic signal are identified as structure acoustic components. This includes: extracting the amplitude envelope of the reference acoustic signal through envelope detection, taking the peak time of the amplitude envelope that exceeds a set amplitude threshold as the arrival time of the shock wave packet, searching for time-frequency regions on the instantaneous phase difference plane whose time dimension coincides with the arrival time of the shock wave packet, detecting the phase jump variables of adjacent time-frequency points within the time-frequency region, and classifying the time-frequency points whose phase jump variables exceed a preset angle as structure acoustic components.

[0010] Furthermore, S4 includes: Obtain the normal clearance distance between the inner wall of the enclosure and the outer wall of the hub at the location of the measuring acoustic sensor; The normal clearance distance between the inner wall of the enclosure and the outer wall of the hub at the location of the sound sensor was calibrated multiple times, and the average value of the calibration was taken as the normal clearance distance. The theoretical phase difference is obtained by dividing the normal gap distance by the wavelength of the sound wave propagating in air at the ambient temperature of the detection environment; The measured phase difference and theoretical phase difference of each time-frequency component in the structural acoustic component are subtracted, and the components whose absolute value of the difference is less than the preset phase deviation threshold are separated into direct impact structural acoustic components.

[0011] Furthermore, S5 includes: Perform time-frequency transformation on the measured acoustic signal; The amplitude of the separated direct impact structural acoustic components is set to zero or weighted attenuation on the time-frequency plane; The signal is then restored to the time domain by inverse time-frequency transformation, yielding the residual acoustic signal.

[0012] Furthermore, the amplitude of the separated direct impact structural acoustic components is zeroed or weighted attenuated on the time-frequency plane, including: on the time-frequency plane, the amplitude of the time-frequency points belonging to the direct impact structural acoustic components is zeroed, the amplitude of the time-frequency points adjacent to the direct impact structural acoustic components and whose coherence values ​​exceed a set threshold is weighted attenuated according to the coherence value ratio, and the amplitude of the remaining time-frequency points remains unchanged.

[0013] Furthermore, S6 includes: The residual acoustic signal is used as the direct wave of the leak hole jet, and the time difference of the direct wave received by each measuring acoustic sensor is calculated. A set of positioning equations is constructed based on the spatial coordinates of each acoustic sensor and the time difference between the direct waves received by each acoustic sensor. The location of the leak hole is obtained by solving the system of positioning equations.

[0014] On the other hand, the present invention provides a system for real-time analysis of wheel hub airtightness detection data and leakage point location, comprising the following modules: The signal acquisition module is used to set a reference acoustic sensor on the inner wall of the enclosure of the wheel hub airtightness testing equipment, and to set multiple measuring acoustic sensors in the annular cavity between the wheel hub and the enclosure to simultaneously acquire the reference acoustic signal and the measuring acoustic signal during the pressure holding stage. The cross-spectral analysis module is used to perform cross-power spectrum analysis on the reference acoustic signal and each measured acoustic signal to obtain the cross-power spectral density of each measurement channel; The structure recognition module is used to construct the instantaneous phase difference plane of each measurement channel based on the phase spectrum of the cross power spectral density, identify the time-frequency components on the instantaneous phase difference plane where the phase undergoes a step-like abrupt change and the abrupt change time is aligned with the arrival time of the shock wave packet in the reference acoustic signal, and identify them as the structure acoustic components caused by the leakage airflow impacting the inner wall of the enclosure. The component separation module is used to separate the components in the structural acoustic component whose phase difference matches the geometric distance relationship, based on the geometric distance relationship between the inner wall of the enclosure and the outer wall of the hub, into direct impact structural acoustic components. The filtering processing module is used to filter out the direct impact structural acoustic components from each measured acoustic signal to obtain the residual acoustic signal; The positioning calculation module is used to use the residual acoustic signal as the direct wave of the leak hole jet, and to calculate the location of the leak hole by using the spatial coordinates of each measuring acoustic sensor for time difference positioning.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. By identifying and separating the impact structural acoustic components within the confined cavity, the signal source relied upon for time-difference positioning is restored from the aliased sound field to the direct wave of the leak hole jet. This eliminates the problem of the positioning point deviating from the actual leak hole due to the virtual impact source. The structural acoustic is identified by utilizing the phase step-like abrupt change characteristics on the instantaneous phase difference plane between the reference acoustic sensor and the measurement acoustic sensor. Then, the direct impact component is separated by using the geometric distance relationship between the inner wall of the enclosure and the outer wall of the wheel hub as a constraint. After filtering, the residual signal is used for time-difference positioning, which significantly improves the accuracy of the correspondence between the finally calculated leak hole location and the actual defect point, reducing the effective wheel hub rework and scrapping caused by mislabeling.

[0016] 2. It can be directly embedded into the existing structure of the wheel hub airtightness testing equipment. Only a reference acoustic sensor is added and virtual source suppression is completed at the signal analysis level. There is no need to change the main structure of the equipment or the open cavity design. While maintaining the original testing cycle, the positioning reliability is improved. Moreover, the combination of acoustic processing and geometric constraint separation is adaptable to changes in wheel hub model and cover size. It provides an online technical means for the accurate repair of leaking wheel hubs in automated production lines. Attached Figure Description

[0017] Figure 1 This is a flowchart of a method for real-time analysis of wheel hub airtightness test data and leakage point location according to the present invention; Figure 2 This is a schematic diagram of the structure of a wheel hub airtightness detection data real-time analysis and leak point location system according to the present invention. Detailed Implementation

[0018] 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1: Figure 1 This invention provides a method for real-time analysis of wheel hub airtightness test data and leakage point location, which includes the following steps: S1: A reference acoustic sensor is installed on the inner wall of the enclosure of the wheel hub airtightness testing equipment, and multiple measuring acoustic sensors are installed in the annular cavity between the wheel hub and the enclosure to simultaneously collect the reference acoustic signal and the measuring acoustic signal during the pressure holding stage. S2: Perform cross-power spectrum analysis on the reference acoustic signal and each measured acoustic signal to obtain the cross-power spectral density of each measurement channel; S3: Construct the instantaneous phase difference plane of each measurement channel based on the phase spectrum of the cross power spectral density, identify the time-frequency component on the instantaneous phase difference plane where the phase changes in a step-like manner and the time of the change is aligned with the arrival time of the shock wave packet in the reference acoustic signal, and identify it as the structural acoustic component caused by the impact of the leaked airflow on the inner wall of the enclosure. S4: Based on the geometric distance relationship between the inner wall of the enclosure and the outer wall of the hub, the component in the structural acoustic component whose phase difference matches the geometric distance relationship is separated into the direct impact structural acoustic component. S5: Filter out the direct impact structural acoustic component from each measured acoustic signal to obtain the residual acoustic signal; S6: The residual acoustic signal is used as the direct wave of the leak hole jet. The spatial coordinates of each measuring acoustic sensor are used for time difference positioning to calculate the location of the leak hole.

[0020] In step S1, in the wheel hub airtightness testing equipment, the enclosure is a protective structure surrounding the outer perimeter of the wheel hub, forming an annular cavity between the enclosure and the wheel hub. The inner wall of the enclosure is the wall surface facing the wheel hub. The reference acoustic sensor is a sensor used to pick up vibrations of the inner wall of the enclosure and the sound pressure of the air near the inner wall of the enclosure. A piezoelectric accelerometer or electret condenser microphone is selected. The reference acoustic sensor is rigidly connected to the inner wall of the enclosure by means of threaded fastening, adhesive bonding, or clamping. There is no relative displacement between the sensitive element of the reference acoustic sensor and the inner wall of the enclosure. The area of ​​the inner wall of the enclosure facing the outer wall of the wheel hub is the area where the inner wall of the enclosure and the outer circumferential surface of the wheel rim are projected in the radial direction. The leaking airflow directly impacts this area after being ejected from the wheel hub leak hole. The installation point of the reference acoustic sensor is determined according to the structural dimensions of the enclosure and the wheel hub. The relative positional relationship between the inner wall of the enclosure and the reference acoustic sensor is measured by a laser rangefinder to obtain the spatial coordinates of the reference acoustic sensor, which are then recorded and stored.

[0021] The acoustic sensor is used to pick up leakage sound waves propagating through the air medium within the annular cavity. It employs a pre-polarized capacitive microphone or a MEMS acoustic sensor, with a frequency response range covering, for example, 20Hz to 20kHz. Its sensitivity meets the dynamic range requirements for sound pressure levels of leakage jet noise and impact structural noise within the annular cavity. The annular cavity is the space between the inner wall of the enclosure and the outer wall of the hub. The outer wall of the hub includes the outer circumferential surface of the rim, the outer surface of the spokes, and the area surrounding the valve holes. Multiple acoustic sensors are arranged circumferentially within the annular cavity along the outer wall of the hub. Using the rim axis as a reference center line, the installation positions of the acoustic sensors are set at preset intervals on the circumference of the annular cavity corresponding to the rim cross-section. The preset interval angle is determined based on the number of acoustic sensors and the rim diameter. For example, when there are 8 acoustic sensors and the rim diameter is 400mm, the preset interval angle is 45 degrees, ensuring that the acoustic aperture of the acoustic sensors covers the entire circumferential range of the hub. The measuring acoustic sensors are fixed to the inner wall of the enclosure by a bracket, with the acoustic sensitive end of the sensor facing the inside of the annular cavity. The spatial coordinates of each measuring acoustic sensor are calibrated and recorded using a laser rangefinder. The spatial coordinates of the reference acoustic sensor and each measuring acoustic sensor are unified in the same device coordinate system, which is established with the geometric center of the detection device base as the origin. The spatial coordinates of the reference acoustic sensor and each measuring acoustic sensor are used as known parameters for subsequent time-difference positioning and stored in the storage unit of the detection device.

[0022] After the wheel hub airtightness testing equipment completes the inflation of the wheel hub and enters the pressure holding stage, the reference acoustic sensor and multiple measurement acoustic sensors begin to synchronously collect the acoustic signals generated by the leakage. The wheel hub is determined to be fully inflated when the air pressure inside the wheel hub reaches the set detection pressure and the inflation valve is closed. The pressure holding stage is the period during which the pressure inside the wheel hub remains stable at the set detection pressure after the inflation valve is closed and no further active inflation is performed. Synchronous acquisition is performed by the same sampling clock driving the reference acoustic sensor and multiple measurement acoustic sensors to sample the acoustic signals at equal intervals. The sampling clock frequency is not less than twice the highest analysis frequency of the leakage acoustic signal. The sampling time starts at the beginning of the pressure holding stage, and the sampling duration covers the complete continuous process of the leakage acoustic signal during the pressure holding stage. The collected data includes the reference acoustic signal output by the reference acoustic sensor and the measurement acoustic signals output by each measurement acoustic sensor. The reference acoustic signal reflects the vibration response of the inner wall of the enclosure under the impact of the leaking airflow and the time-domain change of the air sound pressure near the inner wall of the enclosure. The measurement acoustic signals reflect the time-domain changes of the leakage jet noise and the impact structure sound in the ring cavity as they propagate through the air to each measurement acoustic sensor. During the acquisition process, the reference acoustic signal and the measured acoustic signal are synchronized between channels by sharing a sampling clock. The acquired reference acoustic signal and measured acoustic signal are stored in a time sequence corresponding to the channel number. The stored data record contains the sampling time point and the sensor identification code.

[0023] Channel consistency between the reference acoustic sensor and the measurement acoustic sensor is ensured through a calibration process. During calibration, a standard sound source within the ring cavity emits a calibration signal with a known spectrum. The responses of the reference acoustic sensor and each measurement acoustic sensor to the calibration signal are simultaneously recorded. The amplitude and phase correction coefficients for each measurement channel relative to the reference channel are calculated, stored, and applied to correct the measurement and reference acoustic signals during subsequent acquisition. The frequency range of the calibration signal covers the analysis frequency band of the leakage acoustic signal. The calibration signal is emitted by the standard sound source at a preset sound pressure level, selected within the linear response range of both the measurement and reference acoustic sensors. Through calibration correction, inter-channel deviations introduced by sensitivity differences and phase response inconsistencies between the reference and measurement acoustic sensors are eliminated, ensuring that the reference and measurement acoustic signals have consistent amplitude and phase references in subsequent cross-power spectral analysis.

[0024] In step S2, the reference acoustic signal and the individual measured acoustic signal are segmented, windowed, and subjected to Fourier transform. Segmentation involves dividing the reference and measured acoustic signals into multiple equal-length time-domain data segments in chronological order. The segmentation is based on a preset number of sampling points per segment and the overlap ratio between adjacent segments. The preset number of sampling points per segment is determined according to the analysis frequency resolution and response time requirements of the leaked acoustic signal; for example, 4096 sampling points per segment. The overlap ratio between adjacent segments is 75%. A higher overlap ratio results in stronger time continuity between segments and lower variance in the spectral estimation. Windowing involves multiplying each time-domain data segment by a window function. A Hanning window is used, and its length is equal to the number of sampling points per segment. Windowing is used to suppress spectral leakage caused by segmentation truncation. The Fourier transform is performed on each windowed time-domain data segment to obtain the complex spectrum of the reference acoustic signal and the complex spectrum of the measured acoustic signal. The complex spectrum contains the amplitude and phase information at each frequency point. After segmented windowing and Fourier transform, each time-domain data segment of the reference acoustic signal corresponds to a reference complex spectrum, and each time-domain data segment of the measured acoustic signal corresponds to a measured complex spectrum. The reference complex spectrum and the measured complex spectrum have the same resolution on the frequency axis, which is determined by the number of sampling points per segment and the sampling clock frequency. The sampling clock frequency and the number of sampling points per segment are recorded in the storage unit of the detection device, and the length of the Fourier transform is the same as the number of sampling points per segment.

[0025] The cross-power spectrum (CPS) of the reference and measured acoustic signals in the frequency domain is calculated. The CPS is the result of multiplying the reference complex spectrum and the measured complex spectrum (RCS) corresponding to the same time-domain data segment by their conjugates. Conjugate multiplication involves multiplying the conjugate of the measured complex spectrum with the reference complex spectrum frequency-by-frequency. The resulting CPS is a complex sequence with the same length as the complex spectrum. The real and imaginary parts at each frequency point represent the in-phase and quadrature power components of the reference and measured acoustic signals, respectively, at that frequency. The CPS is calculated independently for each time-domain data segment, with each segment corresponding to a specific CPS. The CPS of different time-domain data segments are arranged by segment number along the time dimension. The frequency range of the CPS is determined by the sampling clock frequency, with an upper limit of half the sampling clock frequency. The intervals on the frequency axis equal the frequency resolution. The amplitude of the CPS reflects the common energy distribution of the reference and measured acoustic signals at different frequencies, and the phase of the CPS reflects the phase difference between the reference and measured acoustic signals at different frequencies.

[0026] Averaging multiple cross-power spectra involves summing the cross-power spectra calculated from different time-domain data segments in the frequency domain and dividing by the number of segments to obtain the cross-power spectral density of the measurement channel. Unweighted cumulative averaging is used. When weighted cumulative averaging is used, the weights of each cross-power spectrum segment are set based on its signal-to-noise ratio (SNR). The SNR of a specific cross-power spectrum segment is obtained as follows: A main signal frequency band and a noise reference frequency band are defined within the frequency sequence of the cross-power spectrum segment. The main signal frequency band is determined based on the typical spectral range of the leakage sound signal. The noise reference frequency band is selected from frequency bands within the analysis frequency range that do not overlap with the main signal frequency band. The ratio of the sum of amplitudes at each frequency point within the main signal frequency band to the sum of amplitudes at each frequency point within the noise reference frequency band is calculated, and this ratio is used as the SNR of that cross-power spectrum segment. Segments with SNR higher than a preset SNR segment threshold are assigned weight one, and segments with SNR lower than or equal to the preset SNR segment threshold are assigned weight two, where weight one is greater than weight two. The preset signal-to-noise ratio (SNR) segment threshold is determined as follows: After the testing equipment performs a complete pressure holding phase on a known leak-free standard wheel hub, the acquired reference acoustic signal and individual measured acoustic signals are processed using the same segmented windowing, Fourier transform, and cross-power spectrum calculation methods as in normal testing to obtain the cross-power spectrum for each segment. The SNR of each cross-power spectrum is calculated, and the statistical average of the SNR of each segment is taken as the preset SNR segment threshold. The cumulative averaged cross-power spectral density is a complex sequence representing the average phase difference and average correlation energy distribution of the reference and measured acoustic signals in the frequency domain. Traversing each measured acoustic signal involves replacing one measured acoustic signal with another, repeatedly performing segmented windowing and Fourier transform, calculating the cross-power spectrum, and averaging multiple cross-power spectra until the cross-power spectral density calculation of all measured acoustic signals with the reference acoustic signal is completed, thus obtaining the cross-power spectral density of all measurement channels. The cross-power spectral density of all measurement channels is stored in the order of the measurement acoustic sensor numbers. The cross-power spectral density of each measurement channel is associated with the spatial coordinates of the corresponding measurement acoustic sensor and the spatial coordinates of the reference acoustic sensor, providing frequency domain input data for subsequent steps.

[0027] The phase spectrum of the cross-power spectral density (CPSD) of the measurement channel is extracted from the complex sequence of CPSDs. The extraction method involves calculating the arctangent of the ratio of the imaginary to the real part of the CPSD at each frequency point to obtain the phase angle. The phase angle ranges from -π to π, and the unit is radians. When the ratio of the imaginary to the real part exceeds the domain of the arctangent function, the phase angle is made continuous through phase dewinding to eliminate discontinuous abrupt changes at the -π to π boundary, resulting in a continuous phase spectrum. Each frequency point in the continuous phase spectrum corresponds to a phase value, and the frequency resolution of the continuous phase spectrum is consistent with that of the CPSD. Phase spectrum extraction is performed after obtaining the CPSD, and the phase spectrum, as part of the CPSD information, participates in the subsequent construction of the instantaneous phase difference plane. The sampling synchronization maintained between the reference and measured acoustic signals during acquisition ensures that the phase spectrum of the CPSD accurately reflects the phase difference between the reference and measured acoustic signals at the same frequency. The accuracy of the phase difference depends on the stability of the sampling clock frequency and the time synchronization accuracy between the channels. The time synchronization accuracy is ensured during the calibration process of S1 by sharing a sampling clock. The stability of the sampling clock frequency is determined by the frequency drift characteristics of the sampling clock source in the detection equipment.

[0028] In step S3, the instantaneous phase difference of each measurement channel is extracted from the phase spectrum of the cross-power spectral density. The instantaneous phase difference refers to the rate of change of the phase difference between the reference acoustic signal and the measured acoustic signal at a certain frequency at a given time point. The extraction process is as follows: For the continuous phase spectrum of the cross-power spectral density of each measurement channel obtained in S2, the phase value sequence of that frequency point is retrieved along the frequency axis point by point in the time series. Each element of the phase value sequence corresponds to the phase value at the center of a time-domain data segment. A difference operation is performed on the phase value sequence along the time direction. The difference operation subtracts the phase values ​​corresponding to adjacent time-domain data segments. The difference result is divided by the time interval between the center times of adjacent time-domain data segments to obtain the instantaneous phase difference sequence at that frequency point. The time interval is determined by the number of sampling points in each segment, the overlap ratio, and the sampling clock frequency. The length of the instantaneous phase difference sequence is 1 less than the length of the phase value sequence. Each value in the instantaneous phase difference sequence represents the rate of phase change between adjacent segments at that frequency point. The unit of the instantaneous phase difference is radians per second. Repeat the differential operation for each frequency point of the continuous phase spectrum to obtain the instantaneous phase difference sequence corresponding to each frequency point. Combine the instantaneous phase difference sequences of all frequency points to form the instantaneous phase difference matrix of the measurement channel. The rows of the instantaneous phase difference matrix correspond to the frequency and the columns correspond to the time. The instantaneous phase difference matrix is ​​the data basis for the instantaneous phase difference plane to be constructed.

[0029] A continuous wavelet transform is performed on the phase spectrum of the cross-power spectral density by constructing an instantaneous phase difference plane on a two-dimensional plane composed of time and frequency, to obtain the instantaneous phase at different center frequencies. The specific method of the continuous wavelet transform is as follows: a complex Morlet wavelet is selected as the mother wavelet function. The center frequency and bandwidth parameters of the complex Morlet wavelet are set according to the analysis frequency range of the leakage sound signal. The center frequency is chosen so that the main lobe of the Fourier spectrum of the complex Morlet wavelet function covers the main energy frequency band of the leakage sound signal. The bandwidth parameter determines the frequency resolution of the complex Morlet wavelet function and takes a value between 1 and 3. A continuous wavelet transform is performed on each frequency slice of the continuous phase spectrum of the cross-power spectral density along the frequency axis. The scale factor sequence of the continuous wavelet transform is determined according to the sampling clock frequency and the desired center frequency sequence, with each scale factor in the scale factor sequence corresponding to a center frequency. During continuous wavelet transform, the complex Morlet wavelet is scaled at a certain scale factor and then convolved with a frequency slice of the continuous phase spectrum. The convolution result corresponds to a complex sequence. The modulus of the complex sequence represents the degree of matching between the continuous phase spectrum and the complex Morlet wavelet function at that scale factor, and the argument of the complex sequence represents the instantaneous phase of the continuous phase spectrum at that scale factor. A series of instantaneous phases at center frequencies are obtained by traversing all scale factors. These instantaneous phases at each center frequency are arranged in ascending order of center frequency and chronological order to form an instantaneous phase difference plane. The horizontal axis of the instantaneous phase difference plane represents time, and the time scale corresponds to the sampling time axis of the holding phase. The vertical axis represents the center frequency, and the center frequency scale corresponds to the frequency value of the scale factor sequence of the continuous wavelet transform. The value at each time-frequency point in the instantaneous phase difference plane is the instantaneous phase value of that center frequency and time, and the unit of the instantaneous phase value is radians.

[0030] After arranging the instantaneous phase difference plane according to the center frequency and time, the time-frequency components whose phase jumps between adjacent time-frequency points exceed a preset jump angle threshold and whose jump time coincides with the arrival time of the impulse wave packet in the reference acoustic signal are identified as structure acoustic components. Adjacent time-frequency points refer to two time-frequency points that are directly adjacent in both the time and frequency dimensions on the instantaneous phase difference plane. A phase jump is the absolute value of the difference between the instantaneous phase values ​​of adjacent time-frequency points; the phase jump occurs on the time coordinate corresponding to the later of the adjacent time-frequency points. The preset jump angle threshold is set based on the typical phase jump caused by the sudden change in acoustic impedance resulting from the impact of the leaking airflow on the inner wall of the enclosure. The preset jump angle threshold is determined as follows: when the detection equipment tests the calibration hub with a known leak hole location, the diameter and location of the leak hole on the calibration hub are known. The collected reference sound signal and measured sound signal are processed by continuous wavelet transforms of S1, S2 and S3 to obtain the instantaneous phase difference plane under the calibration data. The phase fluctuation range of the direct wave region of the leak hole jet is statistically analyzed on the instantaneous phase difference plane under the calibration data. The upper limit of the phase fluctuation range plus the safety margin is taken as the preset jump angle threshold. The upper limit of the phase fluctuation range is the difference between the maximum value and the mean value of the phase fluctuation. The safety margin is 20% to 50% of the phase fluctuation range.

[0031] The timing of the transition is identified as being consistent with the arrival time of the shock wave packet in the reference acoustic signal. Specifically, the amplitude envelope of the reference acoustic signal is extracted using envelope detection. Envelope detection involves performing a Hilbert transform on the reference acoustic signal and then taking the modulus value to obtain the amplitude envelope sequence. The peak time in the amplitude envelope sequence of the reference acoustic signal that exceeds a set amplitude envelope threshold is taken as the arrival time of the shock wave packet. The set amplitude envelope threshold is determined based on the amplitude envelope level of the background noise during the pressure holding phase. The determination method is as follows: When the detection equipment performs the pressure holding phase on a known leak-free standard wheel hub, a Hilbert transform is performed on the acquired reference acoustic signal to obtain the modulus value, resulting in a background amplitude envelope sequence. The statistical mean and statistical standard deviation of the background amplitude envelope sequence are calculated, and the statistical mean plus three times the statistical standard deviation is used as the set amplitude envelope threshold. On the instantaneous phase difference plane, a time-frequency region whose time dimension coincides with the arrival time of the shock wave packet is searched. Time dimension coincidence means that the difference between the time coordinate of the time-frequency point and the arrival time of the shock wave packet is less than the time coincidence tolerance threshold, which is half the time interval between the center times of adjacent time domain data segments. Within the searched time-frequency region, the phase jump variable of adjacent time-frequency points is detected. Time-frequency points whose phase jump variable exceeds a preset jump angle threshold are classified into the structure sound component. Classification into the structure sound component means that the time-frequency point is marked as a structure sound component caused by the impact of the leaked airflow on the inner wall of the enclosure. The marking information includes the time coordinate and center frequency coordinate of the time-frequency point. The marking information is stored together with the instantaneous phase difference plane for use in subsequent steps to separate the directly impacting structure sound component.

[0032] In step S4, the normal clearance distance between the inner wall of the enclosure and the outer wall of the hub at the location of the acoustic sensor is obtained. The normal clearance distance refers to the straight-line distance from the location of the acoustic sensor along the normal direction of the inner wall of the enclosure to the outer wall of the hub. The location of the acoustic sensor refers to the spatial coordinates of the geometric center of the acoustic sensing end of the acoustic sensor in the device coordinate system. These spatial coordinates have been calibrated and stored in S1 using a laser rangefinder. The normal direction of the inner wall of the enclosure at the location of the acoustic sensor is determined as follows: using the spatial coordinate point cloud of the inner wall of the enclosure calibrated in S1, a plane fitting is performed on the point cloud of the inner wall of the enclosure in the vicinity of the acoustic sensor. The plane fitting uses the least squares method to determine the tangent plane of the inner wall of the enclosure in this region. The unit normal vector of the tangent plane is the normal direction of the inner wall of the enclosure at this location, and the normal direction points towards the inside of the annular cavity. The normal clearance distance is measured as follows: Starting from the geometric center of the acoustic sensor's sensitive end, a ray is drawn along the normal direction. The intersection of this ray with the outer wall of the hub is obtained by calculating the intersection of the spatial coordinate point cloud of the hub's outer wall. The Euclidean distance between the starting point and the intersection point is then used to obtain the normal clearance distance. The spatial coordinate point cloud of the hub's outer wall is calibrated in S1 when calibrating the spatial coordinates of the acoustic sensor, or it can be obtained by extracting point cloud data from the outer circumference of the rim and the outer surface of the spokes based on the hub's CAD model.

[0033] The normal clearance distance between the inner wall of the enclosure and the outer wall of the wheel hub at the location of the acoustic sensor is calibrated multiple times, and the average value is taken as the normal clearance distance. The multiple calibrations are performed as follows: Under no-load conditions, the normal clearance distance at each location of the acoustic sensor is measured repeatedly, for example, 5 or 8 times. During each measurement, the circumferential angle of the wheel hub in the fixture is slightly rotated or the wheel hub is re-clamped and then reset to account for minor variations in the wheel hub clamping posture and enclosure structure during different testing cycles. The normal clearance distance values ​​obtained from the multiple measurements are summed and divided by the number of measurements to obtain the average calibration value. This average calibration value is used as the normal clearance distance at the location of the acoustic sensor and is stored in the storage unit of the testing equipment. The unit of the average calibration value is meters, accurate to the millimeter level.

[0034] The theoretical phase difference is obtained by dividing the normal gap distance by the wavelength of the sound wave propagating in air at the ambient temperature. The wavelength of the sound wave propagating in air at the ambient temperature is determined as follows: the ambient temperature is acquired in real time by a temperature sensor installed on the inner wall of the enclosure or in the annular cavity, and the temperature value acquired by the temperature sensor is in degrees Celsius. The relationship between the speed of sound in air and temperature is: the speed of sound is equal to the linear superposition of the speed of sound at 0℃ and the increase in the speed of sound for every 1℃ increase in temperature. The speed of sound at 0℃ is taken as 331.3 m / s, and the increase in the speed of sound for every 1℃ increase in temperature is taken as 0.606 m / s. The speed of sound at the ambient temperature is calculated by adding 331.3 m / s to the ambient temperature multiplied by 0.606 m / s / ℃. For example, when the ambient temperature is 25℃, the speed of sound is 331.3 + 25 × 0.606. The wavelength of sound propagating in air at the ambient temperature is calculated by dividing the speed of sound at the ambient temperature by the analysis frequency. The analysis frequency is taken as the center frequency corresponding to the time-frequency component to be processed in the structural acoustic component, and the center frequency is determined by the scaling factor sequence of the continuous wavelet transform in S3. The theoretical phase difference is calculated as follows: multiply the normal gap distance by 2, divide by the wavelength of sound propagating in air at the ambient temperature, and then multiply by 2π. The multiplication by 2 is because the propagation path of the shock wave from the inner wall of the enclosure to the outer wall of the hub and then reflected back to the inner wall of the enclosure is the round trip distance of the normal gap distance. The unit of the theoretical phase difference is radians. The theoretical phase difference reflects the phase delay corresponding to the round trip path of the shock wave generated by the leakage airflow impacting the inner wall of the enclosure from the inner wall of the enclosure to the outer wall of the hub and then reflected back to the inner wall of the enclosure.

[0035] The measured phase difference of each time-frequency component in the structural acoustic component is subtracted from the theoretical phase difference. Components with an absolute difference less than a preset phase deviation threshold are separated into direct impact structural acoustic components. The measured phase difference refers to the instantaneous phase value corresponding to each time-frequency point in the structural acoustic component identified in S3 on the instantaneous phase difference plane. The instantaneous phase value is read from the instantaneous phase difference plane. The absolute value of the difference is the absolute value of the difference between the measured phase difference and the theoretical phase difference, and the unit of the absolute value of the difference is radians. The preset phase deviation threshold is determined based on the uncertainty of the theoretical phase difference caused by fluctuations in the detection environment temperature and calibration errors in the normal gap distance. The preset phase deviation threshold is determined as follows: During the transition of the ambient temperature from the equipment warm-up stage to the thermal equilibrium stage, the ambient temperature is continuously collected and the temperature fluctuation range is recorded. The corresponding sound velocity fluctuation value is calculated based on the upper and lower limits of the temperature fluctuation range. The wavelength fluctuation value is then converted from the sound velocity fluctuation value. Finally, the theoretical phase difference fluctuation range is calculated by combining the wavelength fluctuation value with the calibrated average value of the normal gap distance. Half of the difference between the upper and lower limits of the theoretical phase difference fluctuation range is taken as the fluctuation half-width and used as the preset phase deviation threshold. For example, if the normal gap distance is 0.05m and the center frequency is 5000Hz, and the ambient temperature fluctuation range is ±2℃, the corresponding sound velocity fluctuation value is approximately ±1.212m / s. The theoretical phase difference fluctuation half-width is calculated and used as the preset phase deviation threshold. The time-frequency components with an absolute difference value less than a preset phase deviation threshold in the structural sound components are separated and marked as direct impact structural sound components. Direct impact structural sound components refer to the structural sound components generated by the direct impact of leaked airflow on the inner wall of the enclosure, propagating through the air medium to the measuring sound sensor. Direct impact structural sound components do not include structural sound components that are secondary radiated after conduction through the hub wall. The separated direct impact structural sound components are stored in the form of a time-frequency point set. Each time-frequency point in the set contains a time coordinate and a center frequency coordinate. The time-frequency point set is used by S5 to filter out the direct impact structural sound components.

[0036] In step S5, the measured acoustic signal undergoes a time-frequency transformation, which converts the measured acoustic signal from a time-domain representation to a time-frequency domain representation. The time-frequency transformation uses a short-time Fourier transform (SFT). The SFT process is as follows: the measured acoustic signal is divided into multiple equal-length short-time data segments in chronological order. The segmentation method is consistent with the segmentation and windowing method in S2. The number of sampling points in each short-time data segment and the overlap ratio between adjacent segments are the same as the preset number of sampling points and overlap ratio in S2, ensuring that the time axis of the time-frequency plane is aligned with the time axis of the instantaneous phase difference plane in S3 on the time scale. Each short-time data segment is multiplied by a window function. The window function type is the same as the window function used in S2, and the length of the window function is equal to the number of sampling points in each short-time data segment. A fast Fourier transform is performed on each windowed short-time data segment to obtain the complex spectrum corresponding to that short-time data segment. Each frequency point in the complex spectrum contains both amplitude and phase. The complex spectra of all short-time data segments are arranged along the time axis according to their segment numbers, with the arrangement order consistent with the chronological order of the short-time data segments, forming the time-frequency surface of the measured acoustic signal. The horizontal axis of the time-frequency surface represents time, and the time scale is determined by the center time of the short-time data segment, which is equal to the start time of the short-time data segment plus half its duration. The vertical axis of the time-frequency surface represents frequency, and the frequency scale is determined by the frequency resolution of the Fast Fourier Transform, which is equal to the sampling clock frequency divided by the number of sampling points per segment. The value of each time-frequency point in the time-frequency surface is a complex number. The modulus of the complex number represents the amplitude of the measured acoustic signal at that time and frequency, and the argument of the complex number represents the phase of the measured acoustic signal at that time and frequency. The measured acoustic signal of each measurement channel is processed through a Short-Time Fourier Transform to obtain the corresponding time-frequency surface, and the time-frequency surface of each measurement channel is stored separately.

[0037] The amplitude of the direct impact structural acoustic component separated from S4 is either zeroed out or weighted attenuated on the time-frequency surface. The direct impact structural acoustic component separated from S4 is stored as a set of time-frequency points. Each time-frequency point in the set contains a time coordinate and a center frequency coordinate. The time coordinate is consistent with the time axis of the instantaneous phase difference plane in S3, and the center frequency coordinate is consistent with the center frequency sequence of the continuous wavelet transform in S3. Since the time axis of the instantaneous phase difference plane in S3 is aligned with the time axis of the time-frequency surface in S5 on the time scale, and both the center frequency sequence of the continuous wavelet transform in S3 and the frequency axis of the time-frequency surface in S5 are determined by the analysis frequency range, the mapping from the time-frequency point set of S4 to the time-frequency surface of S5 can be directly completed through the correspondence between the time coordinates and frequency coordinates. Zeroing the amplitude involves setting the amplitude of the time-frequency point corresponding to the coordinates of each time-frequency point in the time-frequency point set of the direct impact structural acoustic component separated from S4 to zero on the time-frequency surface. Weighted attenuation involves adjusting the amplitude of time-frequency points adjacent to the direct impact structural acoustic component separated from S4. These time-frequency points are those directly adjacent to the time-frequency points of the direct impact structural acoustic component separated from S4 in both the time and frequency dimensions of the time-frequency plane, but not belonging to the set of time-frequency points of the direct impact structural acoustic component separated from S4. The condition for determining whether adjacent time-frequency points need weighted attenuation is that the coherence value between the adjacent time-frequency point and the time-frequency point of the direct impact structural acoustic component separated from S4 exceeds a set coherence attenuation threshold. The coherence value is an indicator reflecting the similarity between the signal at an adjacent time-frequency point and the direct impact structural acoustic component signal separated from S4. The coherence value is calculated as follows: the amplitude of the current adjacent time-frequency point on the time-frequency plane is compared with the amplitude of the time-frequency point of the direct impact structural acoustic component separated from S4, and then multiplied by the cosine of the phase difference between the adjacent time-frequency points. The phase difference between adjacent time-frequency points is the absolute value of the difference between the phase of the adjacent time-frequency point and the phase of the time-frequency point of the direct impact structural acoustic component separated from S4. When an adjacent time-frequency point is adjacent to multiple time-frequency points of the direct impact structural acoustic component separated from S4, the maximum value among the coherence values ​​calculated with each of the time-frequency points of the direct impact structural acoustic component separated from S4 is taken as the coherence value of that adjacent time-frequency point. The coherence attenuation threshold is determined based on the distinguishability between the structure acoustic component energy and the direct wave energy of the leak hole jet on the time-frequency plane. The determination method is as follows: Within the time-frequency region of the structure acoustic component identified in S3, the coherence value distribution between the time-frequency points of the structure acoustic component and the adjacent time-frequency points of the non-structure acoustic component is statistically analyzed. The lower quartile of this coherence value distribution is taken as the coherence attenuation threshold. For adjacent time-frequency points whose coherence values ​​exceed the set coherence attenuation threshold, their amplitude is multiplied by 1 and subtracted from the attenuation coefficient for weighted attenuation. The attenuation coefficient is equal to the coherence value of that adjacent time-frequency point. For adjacent time-frequency points whose coherence values ​​do not exceed the set coherence attenuation threshold, their amplitude remains unchanged.For the remaining time-frequency points that are not adjacent to any time-frequency point of the direct impact structural acoustic component separated in S4, their time-frequency amplitudes are retained unchanged.

[0038] The residual acoustic signal is then recovered to the time domain through an inverse time-frequency transform. The inverse time-frequency transform is the inverse process of the short-time Fourier transform, transforming the time-frequency surface after amplitude zeroing and weighted attenuation back to the time domain. The execution process of the inverse time-frequency transform is as follows: An inverse fast Fourier transform is performed on the modified complex spectrum corresponding to each short-time data segment on the time-frequency surface to obtain the modified short-time data segment time-domain signal. Adjacent modified short-time data segment time-domain signals are overlapped and added together according to the overlap ratio used when dividing the short-time data segments in the short-time Fourier transform. The overlap and addition synthesis operation is the inverse of the overlap segmentation operation in the short-time Fourier transform. The continuous time-domain signal obtained after overlap and addition synthesis is the residual acoustic signal. The residual acoustic signal is the time-domain signal retained after filtering out the direct impact structure acoustic component separated in S4 from the measured acoustic signal. The residual acoustic signal retains the signal component of the direct wave from the leak hole jet. The duration of the residual acoustic signal is consistent with the duration of the original measured acoustic signal, and the sampling rate of the residual acoustic signal is the same as the sampling clock frequency of the original measured acoustic signal. For cases with multiple measurement channels, time-frequency transformation, amplitude zeroing or weighted attenuation, and inverse time-frequency transformation are performed on the measured acoustic signal of each measurement channel to obtain the residual acoustic signal of each measurement channel. The residual acoustic signal of each measurement channel is stored in the order of the measurement acoustic sensor number. The stored residual acoustic signal contains the time-domain sampling sequence and the corresponding sampling time point, which is used in S6 to calculate the time difference of the direct wave received by each measurement acoustic sensor.

[0039] In step S6, the residual acoustic signal obtained in S5 is used as the direct wave of the leak hole jet, and the time difference between the arrival times of the direct waves received by each measuring acoustic sensor is calculated. The direct wave of the leak hole jet refers to the acoustic wave that radiates directly outward from the leak hole, propagates through the air medium, and reaches the measuring acoustic sensor. The direct wave of the leak hole jet is retained in the residual acoustic signal, while the impact sound on the structure has been filtered out. The time difference between the arrival times of the direct waves received by each measuring acoustic sensor is calculated using the cross-correlation method. The cross-correlation method performs cross-correlation operations on the residual acoustic signals of different measuring channels in the time domain, and uses the time delay corresponding to the peak value of the cross-correlation function as the relative arrival time difference between the two measuring channels. The specific execution method of cross-correlation operation is as follows: Residual acoustic signals from two measurement channels are selected, one as the reference channel residual acoustic signal and the other as the residual acoustic signal of the channel under test. A time-domain cross-correlation is performed on the reference channel residual acoustic signal and the residual acoustic signal of the channel under test. The cross-correlation function is the sliding inner product of the reference channel residual acoustic signal and the residual acoustic signal of the channel under test. The sliding variable of the sliding inner product is the time delay, which increases with the sampling time interval. The range of the time delay covers the maximum possible time difference between the reference channel residual acoustic signal and the residual acoustic signal of the channel under test. The maximum time difference is estimated by dividing the maximum path difference of sound wave propagation within the cavity by the speed of sound in air. The maximum path difference of sound wave propagation within the cavity is taken as the maximum geometric distance difference between the inner wall of the enclosure and the outer wall of the hub in different directions. The time delay corresponding to the peak value of the cross-correlation function is the relative arrival time difference between the reference channel residual acoustic signal and the residual acoustic signal of the channel under test. A positive relative arrival time difference indicates that the residual acoustic signal of the channel under test lags behind the residual acoustic signal of the reference channel, while a negative relative arrival time difference indicates that the residual acoustic signal of the channel under test leads the residual acoustic signal of the reference channel. Repeated cross-correlation operations are performed on each pair of measurement channels to obtain the relative arrival time difference matrix between each measurement channel. Each element in the relative arrival time difference matrix represents the relative arrival time difference of the residual acoustic signals between the two measurement channels.

[0040] A measurement channel is selected as the reference channel. The time when the residual acoustic signal of the reference channel arrives at the corresponding measurement acoustic sensor is used as the time reference. The absolute time difference of the direct wave received by each measurement acoustic sensor is calculated. The reference channel is selected as follows: from the relative arrival time difference matrix, find a row or column with the fewest negative values ​​in the relative arrival time differences between the corresponding measurement channel and other measurement channels. This row or column is then used as the reference channel. The sound transmission path of the measurement acoustic sensor corresponding to the reference channel is the shortest or closest to the shortest sound transmission path among all measurement acoustic sensors. The element values ​​of the row or column corresponding to the reference channel in the relative arrival time difference matrix are used as the time difference of the direct wave received by each measurement acoustic sensor. The time difference of the reference channel itself is set to zero. The time difference of the direct wave received by each measurement acoustic sensor represents the difference between the time it takes for the direct wave from the leak hole to propagate from the leak hole to each measurement acoustic sensor and the time it takes to propagate to the measurement acoustic sensor corresponding to the reference channel. The unit of time difference is seconds. When residual impact structure acoustic interference exists in the residual acoustic signal, the cross-correlation function may have multiple secondary peaks. When the height of the secondary peaks is close to the height of the main peak, the time difference is determined as follows: take the time delay of all peaks in the cross-correlation function that exceed the preset time difference peak threshold, substitute the time delay into the subsequent positioning equations and solve them, and take the time delay with the smallest residual of the positioning equations as the time difference. The preset time difference peak threshold is 80% of the maximum value of the cross-correlation function.

[0041] A set of location equations is constructed based on the spatial coordinates of each acoustic sensor and the time difference between the direct waves received by each sensor. This set of equations describes the geometric relationship between the leak hole location and the spatial coordinates and time differences of each acoustic sensor. The location equations are constructed as follows: For the i-th measurement channel, the equation is: ||P-Si||=||P-S0||+τi×c; where P is the leak hole location coordinate vector, Si is the spatial coordinate vector of the i-th acoustic sensor, S0 is the spatial coordinate vector of the acoustic sensor corresponding to the reference channel, τi is the time difference between the direct waves received by the i-th acoustic sensor, c is the speed of sound in air, and ||·|| represents the Euclidean distance. The constructed location equations are nonlinear, with the unknowns being the three components of the leak hole location coordinate vector P. The number of equations in the set is equal to the number of measurement channels minus one.

[0042] The location of the leak hole is obtained by solving the system of localization equations, using an iterative optimization algorithm. The execution process of the iterative optimization algorithm is as follows: The initial estimated value Pcur of the leak hole location is set as the geometric center of the spatial coordinates of each measuring acoustic sensor. Each coordinate component of the geometric center is equal to the arithmetic mean of the corresponding coordinate components of each measuring acoustic sensor. In each iteration, the estimated time difference τi' of each measuring channel under the current estimated leak hole location is calculated. τi' is calculated as: τi'=(||Pcur-Si||-||Pcur-S0||) / c; where Pcur is the current estimated leak hole location vector, Si is the spatial coordinate vector of the i-th measuring acoustic sensor, S0 is the spatial coordinate vector of the measuring acoustic sensor corresponding to the reference channel, and c is the speed of sound in air. The difference between the estimated time difference τi' and the actual time difference τi is taken to obtain the time difference residual vector r, where the i-th component ri of r is: ri=τi'-τi. Construct the Jacobian matrix J. The element Jij in the i-th row and j-th column of J is calculated as: Jij = ((Pcurj - Sij) / ||Pcur - Si|| - (Pcurj - S0j) / ||Pcur - S0||) / c; where Pcurj is the j-th coordinate component of the current leak hole location estimate Pcur, Sij is the j-th component of the spatial coordinate Si of the i-th measuring acoustic sensor, and S0j is the j-th component of the spatial coordinate S0 of the measuring acoustic sensor corresponding to the reference channel. Calculate the update amount ΔP of the leak hole location estimate, which is: ΔP = pinv(J) × r; where pinv(J) is the pseudo-inverse of the Jacobian matrix J, and r is the time difference residual vector. Add the update amount ΔP to the current leak hole location estimate Pcur to obtain the new leak hole location estimate: Pnew = Pcur + ΔP. The iteration terminates when the L2 norm ||ΔP||² of the update quantity ΔP is less than a preset position convergence threshold ε. The preset position convergence threshold ε is set to, for example, 0.1 mm or 0.5 mm based on the positioning accuracy requirements. Alternatively, iteration stops when the number of iterations reaches a preset maximum number of iterations, for example, 20 or 30. The estimated location of the leak hole at the time of iteration termination is the solved location of the leak hole, output in three-dimensional coordinates, with the coordinate components in millimeters.

[0043] Example 2: Figure 2 A schematic diagram of a real-time analysis and leak point location system for wheel hub airtightness testing data is provided. This system includes the following modules: The signal acquisition module is used to set a reference acoustic sensor on the inner wall of the enclosure of the wheel hub airtightness testing equipment, and to set multiple measuring acoustic sensors in the annular cavity between the wheel hub and the enclosure to simultaneously acquire the reference acoustic signal and the measuring acoustic signal during the pressure holding stage. The cross-spectral analysis module is used to perform cross-power spectrum analysis on the reference acoustic signal and each measured acoustic signal to obtain the cross-power spectral density of each measurement channel; The structure recognition module is used to construct the instantaneous phase difference plane of each measurement channel based on the phase spectrum of the cross power spectral density, identify the time-frequency components on the instantaneous phase difference plane where the phase undergoes a step-like abrupt change and the abrupt change time is aligned with the arrival time of the shock wave packet in the reference acoustic signal, and identify them as the structure acoustic components caused by the leakage airflow impacting the inner wall of the enclosure. The component separation module is used to separate the components in the structural acoustic component whose phase difference matches the geometric distance relationship, based on the geometric distance relationship between the inner wall of the enclosure and the outer wall of the hub, into direct impact structural acoustic components. The filtering processing module is used to filter out the direct impact structural acoustic components from each measured acoustic signal to obtain the residual acoustic signal; The positioning calculation module is used to use the residual acoustic signal as the direct wave of the leak hole jet, and to calculate the location of the leak hole by using the spatial coordinates of each measuring acoustic sensor for time difference positioning.

[0044] All calculations involved in the embodiments are dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0045] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0046] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and inventive constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0047] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0048] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0049] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0050] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for real-time analysis of wheel hub airtightness test data and leakage point location, characterized in that, Includes the following steps: S1: A reference acoustic sensor is installed on the inner wall of the enclosure of the wheel hub airtightness testing equipment, and multiple measuring acoustic sensors are installed in the annular cavity between the wheel hub and the enclosure to simultaneously collect the reference acoustic signal and the measuring acoustic signal during the pressure holding stage. S2: Perform cross-power spectrum analysis on the reference acoustic signal and each measured acoustic signal to obtain the cross-power spectral density of each measurement channel; S3: Construct the instantaneous phase difference plane for each measurement channel based on the phase spectrum of the cross-power spectral density. Identify the time-frequency components on the instantaneous phase difference plane where the phase undergoes a step-like abrupt change and the abrupt change time is aligned with the arrival time of the shock wave packet in the reference acoustic signal. Identify these components as structural acoustic components caused by the impact of leaked airflow on the inner wall of the enclosure; specifically including: The instantaneous phase difference of each measurement channel is extracted from the phase spectrum of the cross power spectral density; A plane of instantaneous phase difference is constructed on a two-dimensional plane consisting of time and frequency. Continuous wavelet transform is performed on the phase spectrum of the cross-power spectral density to obtain the instantaneous phase at different center frequencies. Arranged according to center frequency and time to form an instantaneous phase difference plane; On the instantaneous phase difference plane, time-frequency components whose phase transitions between adjacent time-frequency points exceed a preset angle and whose transition times coincide with the arrival time of the shock wave packet in the reference acoustic signal are identified as structure acoustic components, including: The amplitude envelope of the reference acoustic signal is extracted by envelope detection. The peak time of the amplitude envelope that exceeds the set amplitude threshold is taken as the arrival time of the shock wave packet. The time-frequency region that coincides with the arrival time of the shock wave packet is searched on the instantaneous phase difference plane. The phase jump variable of adjacent time-frequency points is detected in the time-frequency region. The time-frequency points whose phase jump variable exceeds the preset angle are classified into the structure sound component. S4: Based on the geometric distance relationship between the inner wall of the enclosure and the outer wall of the hub, the components in the structural acoustic component whose phase difference matches the geometric distance relationship are separated into direct impact structural acoustic components; specifically including: Obtain the normal clearance distance between the inner wall of the enclosure and the outer wall of the hub at the location of the measuring acoustic sensor; The normal clearance distance between the inner wall of the enclosure and the outer wall of the hub at the location of the sound sensor was calibrated multiple times, and the average value of the calibration was taken as the normal clearance distance. The theoretical phase difference is obtained by dividing the normal gap distance by the wavelength of the sound wave propagating in air at the ambient temperature of the detection environment; The measured phase difference and theoretical phase difference of each time-frequency component in the structural acoustic component are subtracted, and the components whose absolute value of the difference is less than the preset phase deviation threshold are separated into direct impact structural acoustic components. S5: Filter out the direct impact structural acoustic component from each measured acoustic signal to obtain the residual acoustic signal; S6: The residual acoustic signal is used as the direct wave of the leak hole jet. The spatial coordinates of each measuring acoustic sensor are used for time difference positioning to calculate the location of the leak hole.

2. The method for real-time analysis of wheel hub airtightness test data and leakage point location according to claim 1, characterized in that, S1 includes: The reference acoustic sensor is rigidly mounted on the inner wall of the enclosure in the area directly opposite the outer wall of the wheel hub. Multiple acoustic sensors are arranged at intervals along the circumferential direction of the outer wall of the wheel hub within the annular cavity; After the wheel hub is fully inflated and enters the pressure holding stage, the sound signals generated by the leakage are collected synchronously by a reference sound sensor and multiple measurement sound sensors to obtain the reference sound signal and the measurement sound signal respectively.

3. The method for real-time analysis of wheel hub airtightness test data and leakage point location according to claim 1, characterized in that, S2 include: Segmented windowing and Fourier transform are applied to the reference acoustic signal and the individual measured acoustic signal; Calculate the cross-power spectrum of the reference acoustic signal and the measured acoustic signal in the frequency domain; The cross power spectrum of multiple segments is averaged to obtain the cross power spectral density of the measurement channel, and the cross power spectral density of all measurement channels is obtained by traversing each measurement acoustic signal.

4. The method for real-time analysis of wheel hub airtightness test data and leakage point location according to claim 1, characterized in that, S5 include: Perform time-frequency transformation on the measured acoustic signal; The amplitude of the separated direct impact structural acoustic components is set to zero or weighted attenuation on the time-frequency plane; The signal is then restored to the time domain by inverse time-frequency transformation, yielding the residual acoustic signal.

5. The method for real-time analysis of wheel hub airtightness test data and leakage point location according to claim 4, characterized in that, The amplitude of the separated direct impact structural acoustic components is zeroed or weighted attenuated on the time-frequency plane, including: on the time-frequency plane, the amplitude of the time-frequency points belonging to the direct impact structural acoustic components is zeroed, the amplitude of the time-frequency points adjacent to the direct impact structural acoustic components and whose coherence values ​​exceed a set threshold is weighted attenuated according to the coherence value ratio, and the amplitude of the remaining time-frequency points remains unchanged.

6. The method for real-time analysis of wheel hub airtightness test data and leakage point location according to claim 1, characterized in that, S6 include: The residual acoustic signal is used as the direct wave of the leak hole jet, and the time difference of the direct wave received by each measuring acoustic sensor is calculated. A set of positioning equations is constructed based on the spatial coordinates of each acoustic sensor and the time difference between the direct waves received by each acoustic sensor. The location of the leak hole is obtained by solving the system of positioning equations.

7. A system for real-time analysis of wheel hub airtightness test data and leakage point location, used to implement the method for real-time analysis of wheel hub airtightness test data and leakage point location as described in any one of claims 1-6, characterized in that, Includes the following modules: The signal acquisition module is used to set a reference acoustic sensor on the inner wall of the enclosure of the wheel hub airtightness testing equipment, and to set multiple measuring acoustic sensors in the annular cavity between the wheel hub and the enclosure to simultaneously acquire the reference acoustic signal and the measuring acoustic signal during the pressure holding stage. The cross-spectral analysis module is used to perform cross-power spectrum analysis on the reference acoustic signal and each measured acoustic signal to obtain the cross-power spectral density of each measurement channel; The structure recognition module is used to construct the instantaneous phase difference plane of each measurement channel based on the phase spectrum of the cross power spectral density, identify the time-frequency components on the instantaneous phase difference plane where the phase undergoes a step-like abrupt change and the abrupt change time is aligned with the arrival time of the shock wave packet in the reference acoustic signal, and identify them as the structure acoustic components caused by the leakage airflow impacting the inner wall of the enclosure. The component separation module is used to separate the components in the structural acoustic component whose phase difference matches the geometric distance relationship, based on the geometric distance relationship between the inner wall of the enclosure and the outer wall of the hub, into direct impact structural acoustic components. The filtering processing module is used to filter out the direct impact structural acoustic components from each measured acoustic signal to obtain the residual acoustic signal; The positioning calculation module is used to use the residual acoustic signal as the direct wave of the leak hole jet, and to calculate the location of the leak hole by using the spatial coordinates of each measuring acoustic sensor for time difference positioning.

Images