Method and system for testing sealing performance of actuator

Abstract

The invention discloses an actuator sealing performance test method and system, particularly relates to the technical field of industrial equipment sealing performance detection, and is used for solving the problem of positioning failure caused by signal aliasing in a multi-leakage scene in an existing method. Pressure attenuation frequency spectrum characteristics and leakage impact frequency spectrum peaks are extracted respectively by pressurizing a sealed cavity and synchronously collecting dynamic pressure and vibration signals; calculating a path loss value based on the amplitude gradient and the propagation time difference, and screening candidate leakage nodes; constructing a space grid model to solve three-dimensional coordinates of the leakage source; after the coordinates are projected to the surface of the structure model, pressure modal coherence and vibration impact energy are coupled and analyzed to generate a detection map; and finally, outputting a multi-leakage-point identification graph through spatial clustering and intensity grading. Accurate separation and spatial positioning of multiple leakage signals are achieved, and the reliability of complex sealing structure detection is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial equipment sealing performance detection. More specifically, the present invention relates to a method and system for testing the sealing performance of an actuator. Background Art

[0002] In the field of industrial equipment manufacturing and maintenance, the sealing performance of an actuator is a key indicator to ensure its reliable operation. Currently, the sealing detection of an actuator mainly relies on conventional technical means such as the pressure decay method, the bubble method, or tracer gas detection. By monitoring the pressure change, observing the generation of bubbles, or detecting the concentration of tracer gas, these methods can effectively determine whether there is a leak in the sealing system. However, such technologies usually judge leaks based on a single signal dimension (such as the overall pressure value or the total gas concentration), and their application scenarios are mainly for the detection requirements of a single leak source.

[0003] When there are multiple leak points in the actuator sealing structure, the signals generated by different leak sources will be superimposed on each other, resulting in the detection data being unable to accurately reflect the actual state of each leak point. Due to the lack of the ability to separate multi-leak signals, the existing methods are difficult to locate the specific leak position, prone to misjudgment or repeated testing, and significantly reduce the detection efficiency and the reliability of maintenance guidance. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a method and system for testing the sealing performance of an actuator to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for testing the sealing performance of an actuator, comprising the following steps:

[0007] S1. Pressurize the sealing cavity of the actuator to be tested to a preset pressure value, and synchronously collect the dynamic pressure signal in the sealing cavity and the vibration waveform signal of the housing;

[0008] S2. Perform time-frequency analysis on the dynamic pressure signal to extract the first spectral peak containing the pressure decay feature, and perform spectral analysis on the vibration waveform signal to extract the second spectral peak related to the leakage impact;

[0009] S3. Calculate the path loss value according to the amplitude gradient of the first spectral peak and the propagation time difference of the second spectral peak, and select the sensor group with the path loss value less than the loss threshold as the candidate leakage node;

[0010] S4. Construct a spatial grid model with the candidate leakage nodes as vertices, and solve the initial three-dimensional coordinate set of the leak source through the time difference of arrival of the vibration waveform signal;

[0011] S5. Project the initial three-dimensional coordinate set onto the surface of the actuator three-dimensional structure model, and generate a detection map based on the coupled analysis of the pressure signal modal space coherence and the vibration transient impact energy;

[0012] S6. According to the coordinate distribution and energy proportion of the leakage area in the detection map, generate a detection identification map of multiple leakage points through spatial density clustering and leakage intensity grading analysis.

[0013] In a preferred embodiment, pressurize the sealing cavity of the actuator to be tested to a preset pressure value, and synchronously collect the dynamic pressure signal in the sealing cavity and the vibration waveform signal of the housing, including:

[0014] After fixing the actuator to be tested on the test bench, apply increasing pressure in stages until the sealing cavity reaches the preset pressure value;

[0015] When the pressure in the sealing cavity reaches the preset pressure value, continuously collect the dynamic pressure signal in the sealing cavity through the pressure sensor group at a preset sampling frequency, and at the same time collect the vibration waveform signal of the housing through the vibration sensor array attached to the surface of the housing with a synchronous sampling clock;

[0016] Perform time-domain alignment processing on the dynamic pressure signal and the vibration waveform signal to generate a multi-channel signal data set with consistent timestamps.

[0017] In a preferred embodiment, perform time-frequency analysis on the dynamic pressure signal to extract the first spectral peak containing pressure decay characteristics, and perform spectral analysis on the vibration waveform signal to extract the second spectral peak related to leakage impact, including:

[0018] Perform band-pass filtering on the dynamic pressure signal in the multi-channel signal data set to retain the signal components within the preset frequency band;

[0019] Perform time-frequency analysis on the filtered dynamic pressure signal, extract the time-frequency spectrogram through short-time Fourier transform, and identify the frequency components with an amplitude decay rate exceeding the first threshold in the time-frequency spectrogram as the first spectral peak;

[0020] Perform windowing on the vibration waveform signal in the multi-channel signal data set, generate the power spectral density diagram through fast Fourier transform, and locate the frequency components with an energy mutation amplitude exceeding the second threshold in the power spectral density diagram as the second spectral peak;

[0021] Perform waveform matching between the first spectral peak and the preset leakage decay characteristic template, and screen the candidate first spectral peaks with a matching degree exceeding the third threshold; perform correlation calculation between the second spectral peak and the preset mechanical impact characteristic template, and screen the candidate second spectral peaks with a correlation coefficient exceeding the fourth threshold.

[0022] In a preferred embodiment, a path loss value is calculated based on the amplitude gradient of the first spectral peak and the propagation time difference of the second spectral peak, and the sensor groups with path loss values less than the loss threshold are screened as candidate leakage nodes, including:

[0023] Based on the frequency distribution and amplitude gradient of the candidate first spectral peak, calculate the amplitude attenuation coefficient of the received signal of each sensor on the leakage propagation path;

[0024] Based on the arrival time difference of the candidate second spectral peak and the sensor spacing, calculate the propagation speed of the leakage impact signal between each sensor;

[0025] Generate a path loss value according to the product of the amplitude attenuation coefficient and the propagation speed;

[0026] Compare the path loss value with the preset loss threshold, and screen the sensor groups with path loss values less than the loss threshold as candidate leakage nodes;

[0027] Conduct a topological verification on the spatial distribution of the candidate leakage nodes, and retain the sensor groups that meet the adjacency density condition as valid candidate leakage nodes after removing the isolated nodes.

[0028] In a preferred embodiment, a spatial grid model with candidate leakage nodes as vertices is constructed, and the initial three-dimensional coordinate set of the leakage source is solved through the arrival time difference of the vibration waveform signal, including:

[0029] Construct a spatial grid model with the three-dimensional spatial coordinates of the valid candidate leakage nodes as vertices, and the edges between the vertices are the actual distances between adjacent nodes;

[0030] According to the arrival time difference of the vibration waveform signal between each valid candidate leakage node, calculate the propagation time difference vector of the leakage shock wave;

[0031] Based on the topological relationship of the vertices in the spatial grid model and the propagation time difference vector, establish a geometric constraint equation set for solving the leakage source position;

[0032] Use the least squares method to iteratively optimize the solution of the geometric constraint equation set, and generate an initial three-dimensional coordinate set of the leakage source that meets the preset error threshold.

[0033] In a preferred embodiment, project the initial three-dimensional coordinate set onto the surface of the actuator three-dimensional structure model, and generate a detection map based on the coupled analysis of the pressure signal modal space coherence and the vibration transient impact energy, including:

[0034] Orthogonally project the initial three-dimensional coordinate set of the leakage source along the normal direction of the surface of the actuator three-dimensional structure model to generate the surface projection coordinates of the candidate leakage points;

[0035] Extract the modal space coherence matrix of the pressure signal corresponding to the projection area of the candidate leakage point, and calculate the covariance ratio of the main diagonal elements to the off-diagonal elements in the modal space coherence matrix of the pressure signal as the modal coherence coefficient;

[0036] Synchronously extract the vibration transient impact energy value corresponding to the projection area of the candidate leakage point, and calculate the energy correction factor based on the propagation attenuation model of the leakage shock wave;

[0037] Fuse the modal coherence coefficient and the energy correction factor according to the preset weight to generate the leakage intensity factor; generate a two-dimensional detection map based on the distribution density of the surface projection coordinates of the candidate leakage point and the leakage intensity factor.

[0038] In a preferred embodiment, the color depth of the coordinate points in the two-dimensional detection map represents the leakage intensity level.

[0039] In a preferred embodiment, according to the coordinate distribution and energy proportion of the leakage area in the detection map, through spatial density clustering and leakage intensity grading analysis, generate a detection identification map for multiple leakage points, including:

[0040] Based on the coordinate distribution of the leakage area in the two-dimensional detection map, use the spatial density clustering algorithm to divide the spatial clusters of the candidate leakage points;

[0041] Extract the maximum value and the mean value of the leakage intensity factor in each spatial cluster, and map the leakage intensity factor to the preset grading interval based on the dynamic interval division strategy;

[0042] Assign a unique identifier to each spatial cluster according to the leakage intensity grading interval, and the graphic size of the identifier is proportional to the total value of the leakage intensity factor in the cluster;

[0043] Overlay the identifier on the surface of the actuator three-dimensional structure model to generate a detection identification map for multiple leakage points, and different identifiers in the detection identification map are colored to distinguish the leakage intensity levels.

[0044] In a preferred embodiment, the neighborhood radius when the spatial density clustering algorithm divides the spatial clusters of the candidate leakage points is adaptively adjusted according to the energy proportion of the leakage area.

[0045] On the other hand, the present invention provides an actuator sealing performance test system, including the following modules:

[0046] A pressurization and acquisition module, which is used to pressurize the sealing cavity of the actuator to be tested to a preset pressure value, and synchronously acquire the dynamic pressure signal in the sealing cavity and the vibration waveform signal of the housing;

[0047] A signal analysis module, which is used to perform time-frequency analysis on the dynamic pressure signal to extract the first frequency spectrum peak containing the pressure attenuation characteristics, and perform spectrum analysis on the vibration waveform signal to extract the second frequency spectrum peak related to the leakage shock;

[0048] A feature processing module, configured to calculate a path loss value according to the amplitude gradient of the first spectral peak and the propagation time difference of the second spectral peak, and screen the sensor groups with path loss values less than the loss threshold as candidate leakage nodes;

[0049] A model construction module, configured to construct a spatial grid model with candidate leakage nodes as vertices, and solve the initial three-dimensional coordinate set of the leakage source through the time difference of arrival of vibration waveform signals;

[0050] A spectrum generation module, configured to project the initial three-dimensional coordinate set onto the surface of the actuator three-dimensional structure model, and generate a detection spectrum based on the coupled analysis of the pressure signal modal space coherence and the vibration transient shock energy;

[0051] An identification generation module, configured to generate a detection identification map of multiple leakage points through spatial density clustering and leakage intensity grading analysis according to the coordinate distribution and energy proportion of the leakage area in the detection spectrum.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. Through multi-dimensional signal fusion and spatial correlation analysis, accurate identification and visual positioning of multiple leakage points of the actuator are realized. Based on the joint feature extraction of dynamic pressure signals and vibration waveforms, the signal interference superimposed by different leakage sources can be effectively separated. By screening candidate nodes through a path loss model and combining three-dimensional grid modeling technology, the positioning accuracy of the leakage source is significantly improved. Compared with traditional single-dimensional detection methods, by using the physical correlation between pressure attenuation characteristics and vibration shock propagation, a coupled distribution model of leakage energy in space is constructed, greatly improving the spatial resolution of weak leakage signals in complex structures. The generation of the detection spectrum integrates the dynamic weights of modal coherence and transient energy, intuitively presenting the leakage intensity level and position distribution, providing high-confidence data support for maintenance decisions.

[0054] 2. By means of adaptive threshold adjustment and hierarchical clustering algorithms, the problems of high false positive rate and ambiguous positioning in multi-leakage point detection are solved. Based on the spatial density analysis and path loss calculation of the sensor network, noise interference can be automatically eliminated and real leakage clusters can be identified, effectively avoiding repeated tests. The detection identification map integrates leakage intensity, spatial distribution, and energy proportion information into the same visualization interface through multi-dimensional coding of color and size, greatly improving the interpretability of the detection results. While maintaining the convenience of implementation in industrial sites, a technical leap from single-leakage determination to precise positioning of multiple sources of leakage is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a flowchart of a method for testing the sealing performance of an actuator according to the present invention;

[0056] Figure 2 This is a schematic structural diagram of a testing system for the sealability of an actuator according to the present invention. Specific embodiments

[0057] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0058] Embodiment 1: Figure 1 A method for testing the sealability of an actuator according to the present invention is provided, which includes the following steps:

[0059] S1. Pressurize the sealing cavity of the actuator to be tested to a preset pressure value, and synchronously collect the dynamic pressure signal in the sealing cavity and the vibration waveform signal of the housing.

[0060] S2. Perform time-frequency analysis on the dynamic pressure signal to extract the first spectral peak with pressure decay characteristics, and perform spectral analysis on the vibration waveform signal to extract the second spectral peak related to leakage impact.

[0061] S3. Calculate the path loss value according to the amplitude gradient of the first spectral peak and the propagation time difference of the second spectral peak, and select the sensor group with a path loss value less than the loss threshold as the candidate leakage node.

[0062] S4. Construct a spatial grid model with the candidate leakage nodes as vertices, and solve the initial three-dimensional coordinate set of the leakage source through the time difference of arrival of the vibration waveform signal.

[0063] S5. Project the initial three-dimensional coordinate set onto the surface of the three-dimensional structure model of the actuator, and generate a detection map based on the coupling analysis of the modal space coherence of the pressure signal and the vibration transient impact energy.

[0064] S6. According to the coordinate distribution and energy ratio of the leakage area in the detection map, generate a detection identification map of multiple leakage points through spatial density clustering and leakage intensity grading analysis.

[0065] S1. Pressurize the sealing cavity of the actuator to be tested to a preset pressure value, and synchronously collect the dynamic pressure signal in the sealing cavity and the vibration waveform signal of the housing. Specifically, it can be implemented as follows:

[0066] After fixing the actuator to be tested to the test bench, apply increasing pressure in stages until the sealing cavity reaches the preset pressure value. The test bench includes a rigid support and an adjustable fixture. For example, the rigid support is fixed to the ground by bolts, and the adjustable fixture adjusts the clamping force through a knob to adapt to actuators of different sizes. When applying increasing pressure in stages, a pipeline system including multiple pressurization stages is used. For example, the pipeline system includes a low-pressure air pump, a medium-pressure air pump, and a high-pressure air pump connected in series. The outlet pressure of the low-pressure air pump is, for example, 0.1 MPa, the outlet pressure of the medium-pressure air pump is, for example, 0.5 MPa, and the outlet pressure of the high-pressure air pump is, for example, 1.0 MPa. The preset pressure value is set according to the rated working pressure of the actuator. For example, for an actuator with a rated working pressure of 0.8 MPa, the preset pressure value can be set to 1.2 times the rated pressure (e.g., 0.96 MPa). During the pressure application process, the holding time for each stage of pressure can be set according to actual needs. For example, it can be set to 30 seconds to avoid impact on the structure of the sealing cavity caused by sudden pressure changes.

[0067] When the pressure in the sealing cavity reaches the preset pressure value, continuously collect the dynamic pressure signals in the sealing cavity through the pressure sensor group at the preset sampling frequency. The pressure sensor group includes multiple pressure sensors. For example, at least three pressure sensors are evenly distributed on the inner wall of the sealing cavity, and the spacing between adjacent pressure sensors is set according to the size of the sealing cavity. For example, it is not greater than one-fourth of the circumference of the sealing cavity to ensure the uniformity of signal coverage. The preset sampling frequency is set according to the frequency characteristics of the leakage signal. For example, when the pressure fluctuations caused by leakage are mainly concentrated in the low-frequency band, the preset sampling frequency can be set to not less than 400 Hz to meet the requirements of signal integrity. The acquisition duration of the dynamic pressure signal covers at least three complete pressure fluctuation cycles. For example, when the leakage characteristic frequency is 50 Hz, the acquisition duration can be set to 60 milliseconds to completely capture the periodic characteristics.

[0068] At the same time, collect the vibration waveform signals of the housing through the vibration sensor array attached to the housing surface with a synchronous sampling clock. The vibration sensor array includes multiple vibration sensors. For example, at least four vibration sensors are staggered along the axial and circumferential directions of the housing, and the spacing between adjacent vibration sensors is set according to the size of the housing. For example, the axial spacing is one-fifth of the housing length, and the circumferential spacing is one-sixth of the housing circumference. The synchronous sampling clock is generated by an external clock source. For example, the clock source outputs a pulse signal to trigger the sampling actions of the pressure sensor group and the vibration sensor array simultaneously, and the rising edge accuracy of the pulse signal can be set to 1 microsecond to ensure the synchronization accuracy. The sampling frequency of the vibration waveform signal is the same as the preset sampling frequency of the dynamic pressure signal. For example, both are set to 1 kHz to ensure the signal alignment requirements for subsequent analysis.

[0069] Perform time-domain alignment processing on the dynamic pressure signal and the vibration waveform signal to generate a multi-channel signal dataset with consistent timestamps. The time-domain alignment processing includes the following steps: Extract the original sampling data of the pressure sensor group and the vibration sensor array, and mark the start timestamps of each signal channel according to the trigger time of the synchronous sampling clock; Perform interpolation processing on the sampling time series of the pressure signal and the vibration signal. For example, when the sampling interval of the pressure signal is 1 millisecond and the sampling interval of the vibration signal is 0.5 millisecond, adjust the sampling interval of the pressure signal to 0.5 millisecond by interpolation; Align the adjusted pressure signal and vibration signal along the time axis according to the start timestamps to generate a multi-channel signal dataset with a unified time reference. The data format of the multi-channel signal dataset can be set as a two-dimensional matrix. For example, the rows of the matrix correspond to the time series, and the columns correspond to each sensor channel. The first column is the data of channel 1 of the pressure sensor group, the second column is the data of channel 1 of the vibration sensor array, and so on.

[0070] When applying increasing pressure in stages, if the leakage amount of the sealing cavity exceeds the temporary threshold, suspend pressurization and record the current pressure value as the leakage critical pressure. The temporary threshold is set according to the relationship between the volume of the sealing cavity and the leakage rate. For example, for a sealing cavity with a volume of 0.1 cubic meters, the temporary threshold can be set as the pressure drop not exceeding 0.01 MPa per second. The leakage critical pressure is used to evaluate the sealing performance level of the actuator. For example, when the leakage critical pressure reaches 80% of the preset pressure value, it is determined that the sealing performance is qualified, and when it is lower than this ratio, it is determined to be unqualified.

[0071] The layout positions of the pressure sensor group and the vibration sensor array are optimized according to the structural characteristics of the actuator. For example, for an actuator with an axisymmetric structure, the pressure sensors can be evenly distributed at equal angles along the circumference of the sealing cavity. For example, a pressure sensor is arranged every 120 degrees; the vibration sensors can be symmetrically arranged along the axial center line of the housing. For example, a vibration sensor is arranged at both ends and the middle of the housing. For an actuator with a non-axisymmetric structure, the pressure sensors can be arranged in the stress concentration areas of the sealing cavity, such as welds or flange joints; the vibration sensors can be arranged in the areas with significant vibration responses of the housing, such as the vibration peak positions determined by modal analysis.

[0072] The time-domain alignment processing of the dynamic pressure signal and the vibration waveform signal also includes the elimination of abnormal data segments. Abnormal data segments include distorted waveforms caused by transient overload of sensors or environmental interference. The determination method for abnormal data segments is: Calculate the average energy and standard deviation of the signal. When the signal energy within a certain time window exceeds a certain multiple (such as three standard deviations) of the average energy, determine that this window is an abnormal data segment. The abnormal data segment can be repaired by interpolation of adjacent normal data segments. For example, for an abnormal data segment consisting of 5 sampling points, fill it with the mean values of the first 10 and the last 10 sampling points.

[0073] After the multi-channel signal dataset is generated, the alignment accuracy is ensured by verifying the timestamp deviation of each sensor channel. The allowable range of the timestamp deviation can be set to not exceed one sampling interval. For example, when the sampling frequency is 1 kHz, the timestamp deviation does not exceed 1 millisecond. When the timestamp deviation of a certain sensor channel exceeds the limit, the synchronous sampling clock is re-triggered and the data acquisition process is repeated until the timestamp deviations of all channels meet the requirements.

[0074] S2. Perform time-frequency analysis on the dynamic pressure signal to extract the first spectral peak containing the pressure decay feature, and perform spectral analysis on the vibration waveform signal to extract the second spectral peak related to the leakage impact. Specifically, it can be implemented as follows:

[0075] Perform band-pass filtering on the dynamic pressure signal in the multi-channel signal dataset to retain the signal components within the preset frequency band. The preset frequency band for the band-pass filtering is set according to the typical frequency range of the leakage pressure fluctuation. For example, the pressure decay feature caused by the leakage of the sealed cavity is mainly concentrated in the frequency band of 20 Hz to 200 Hz, and the preset frequency band can be set to 20 Hz to 200 Hz. The band-pass filter uses a Butterworth filter, such as a fourth-order Butterworth filter, with a passband ripple set to 1 dB and a stopband attenuation set to 40 dB to ensure suppressing the out-of-band noise while retaining the signals in the target frequency band. Only the signal components within the preset frequency band are retained in the filtered dynamic pressure signal, such as the pressure fluctuation data in the range of 20 Hz to 200 Hz.

[0076] Perform time-frequency analysis on the filtered dynamic pressure signal. Extract the time-frequency spectrogram through the short-time Fourier transform, and identify the frequency components with an amplitude decay rate exceeding the first threshold in the time-frequency spectrogram as the first spectral peak. The window length of the short-time Fourier transform is set according to the non-stationary characteristics of the signal. For example, when the time scale of the leakage pressure fluctuation is 10 milliseconds, the window length is set to 20 milliseconds to balance the time-frequency resolution. The horizontal axis of the time-frequency spectrogram is time, the vertical axis is frequency, and the amplitude is represented by the shade of color. The calculation method of the amplitude decay rate is as follows: at each frequency point of the time-frequency spectrogram, calculate the downward slope of the amplitude within adjacent time windows. For example, the amplitude drops from A1 to A2 within the time period from the t-th second to the (t + 0.1)-th second, and the decay rate is (A1 - A2) / 0.1. The first threshold is set according to the statistical results of historical leakage data. For example, when the amplitude decay rate exceeds 0.5 MPa per second, it is determined as an effective leakage feature, and values lower than this are regarded as noise.

[0077] Window the vibration waveform signals in the multi-channel signal dataset, generate a power spectral density map through fast Fourier transform, and locate the frequency components in the power spectral density map where the energy mutation amplitude exceeds the second threshold as the second spectral peak. The windowing process uses a Hanning window function. For example, the window function length is 1024 sampling points, and the overlap rate is set to 50% to reduce spectral leakage. The number of points for the fast Fourier transform is set according to the frequency resolution requirement. For example, when the sampling frequency is 1 kHz, a 1024-point fast Fourier transform can obtain a resolution of approximately 0.98 Hz. The horizontal axis of the power spectral density map is frequency, and the vertical axis is the energy amplitude. The energy mutation amplitude is defined as the ratio of the energy value of a local frequency point to the average energy of its adjacent frequency points. For example, if the energy of a certain frequency point is E, and the average energy of 5 frequency points before and after it is E_avg, the mutation amplitude is E / E_avg. The second threshold is set according to typical mechanical shock characteristics. For example, when the mutation amplitude exceeds 3 times, it is determined as a component related to leakage shock, and values below this are ignored.

[0078] Perform waveform matching between the first spectral peak and a preset leakage attenuation feature template, and screen out the candidate first spectral peaks with a matching degree exceeding the third threshold. The leakage attenuation feature template is constructed based on the time-frequency spectrograms of known leakage conditions. For example, in a standard leakage experiment, collect the time-frequency spectrograms corresponding to different leakage apertures, and extract their amplitude attenuation rate and frequency distribution characteristics as the template. Waveform matching is achieved by calculating a similarity index. For example, the cosine similarity is used to measure the closeness between the first spectral peak and the template in the frequency-amplitude space. The third threshold is set according to the distribution of the matching degree. For example, when the cosine similarity exceeds 0.8, it is determined as an effective match, otherwise it is excluded.

[0079] Perform correlation calculation between the second spectral peak and a preset mechanical shock feature template, and screen out the candidate second spectral peaks with a correlation coefficient exceeding the fourth threshold. The mechanical shock feature template is constructed based on the differences in vibration signals between normal conditions and leakage conditions. For example, collect the power spectral density map of the vibration signal when there is no leakage as the baseline template, and the difference region between the power spectral density map under the leakage condition and the baseline template is the mechanical shock feature. The correlation calculation uses the Pearson correlation coefficient. For example, calculate the correlation of the energy distribution between the second spectral peak and the template at the same frequency points. The fourth threshold is set according to the significance level of the correlation. For example, when the Pearson correlation coefficient exceeds 0.7, it is determined as an effective leakage shock component, otherwise it is excluded.

[0080] In band-pass filtering, if the signal-to-noise ratio of the signal components within the preset frequency band is lower than the temporary signal-to-noise ratio threshold, the passband range of the filter is automatically expanded. The temporary signal-to-noise ratio threshold is set according to the signal quality requirements. For example, when the signal-to-noise ratio is lower than 10 dB, it is determined as a low-quality signal. The step size for expanding the passband range is set to 10 Hz. For example, if the initial preset frequency band is from 20 Hz to 200 Hz, after expansion, it is adjusted to 10 Hz to 210 Hz until the signal-to-noise ratio reaches the threshold or the maximum expansion limit is reached.

[0081] The window type of the short-time Fourier transform can be adjusted according to the signal characteristics. For example, for transient leakage signals, a Gaussian window is used to improve the time-domain resolution; for steady-state leakage signals, a rectangular window is used to reduce the computational complexity. The selection of the window type is achieved through preset rules. For example, when the non-stationarity index of the signal exceeds 0.5, it is switched to a Gaussian window; otherwise, a rectangular window is used. The non-stationarity index is calculated as the ratio of the time-domain variance to the frequency-domain variance of the signal.

[0082] The calculation of the energy mutation amplitude in the power spectral density diagram also includes the verification of the frequency band energy integration. For example, after locating the energy mutation frequency point, calculate the total energy integration value of the frequency band where the frequency point is located (for example, within the range of ±5 Hz). If the integration value exceeds the preset integration threshold, the second spectral peak is retained. The preset integration threshold is set according to the calibration result of the sensitivity of the vibration sensor. For example, when the integration value exceeds 10 mV² / Hz, it is determined as an effective energy mutation.

[0083] The results of waveform matching and correlation calculation are weighted and fused to generate the final candidate spectral peaks. For example, the matching degree weight of the first spectral peak is set to 0.6, and the correlation weight of the second spectral peak is set to 0.4. The spectral peaks with a weighted total score exceeding 0.75 are used as effective leakage features. The weight assignment is dynamically adjusted according to the signal-to-noise ratio of the pressure signal and the vibration signal. For example, when the signal-to-noise ratio of the pressure signal is lower than that of the vibration signal, the weight of the vibration signal is increased to 0.6.

[0084] The time alignment verification of the candidate first spectral peak and the candidate second spectral peak is achieved by checking whether the appearance time windows of the two overlap. For example, if the candidate first spectral peak appears within the time window from the t-th second to the (t + 0.2)-th second, the appearance time window of the candidate second spectral peak needs to overlap with the former by at least 50% to be considered an effective association. The time window overlap ratio is set according to the leakage propagation delay. For example, when the physical transmission delay between the pressure fluctuation caused by the leakage and the mechanical shock is less than 0.1 second, the overlap ratio threshold is set to 50%.

[0085] S3. Calculate the path loss value based on the amplitude gradient of the first spectral peak and the propagation time difference of the second spectral peak, and select the sensor group with a path loss value less than the loss threshold as the candidate leakage node. Specifically, it can be implemented as follows:

[0086] Based on the frequency distribution and amplitude gradient of the candidate first spectral peak, calculate the amplitude attenuation coefficient of the received signal of each sensor on the leakage propagation path. The frequency distribution of the candidate first spectral peak is determined by statistically analyzing the main frequency positions of the candidate first spectral peaks detected by each sensor. For example, if the main frequency of the candidate first spectral peak detected by sensor A is 50 Hz and that detected by sensor B is 52 Hz, the frequency distribution difference reflects the dispersion characteristics of the leakage signal propagation path. The amplitude gradient is obtained by calculating the ratio of the amplitude difference between the candidate first spectral peaks of different sensors to the sensor spacing. For example, if the amplitude of the candidate first spectral peak of sensor A is 1.2 V, the amplitude of sensor B is 0.9 V, the spacing between them is 0.5 m, and the amplitude difference is 0.3 V, then the amplitude gradient is 0.3 V divided by 0.5 m, resulting in 0.6 V / m. The amplitude attenuation coefficient is calculated based on the product of the amplitude gradient and the signal propagation distance. For example, if the amplitude gradient is 0.6 V / m and the propagation distance is 2 m, then the amplitude attenuation coefficient is 0.6 V / m multiplied by 2 m, obtaining 1.2 V.

[0087] Based on the time difference of arrival and sensor spacing of the candidate second spectral peak, calculate the propagation speed of the leakage impact signal between each sensor. The time difference of arrival of the candidate second spectral peak is determined by comparing the time stamps of the same candidate second spectral peak detected by different sensors. For example, if the time when sensor A detects the candidate second spectral peak is t1 and the time detected by sensor B is t2, the time difference of arrival is t2 minus t1, resulting in Δt. The sensor spacing is calculated based on the three-dimensional coordinate differences of the sensor layout positions. For example, if the coordinates of sensor A are (x1, y1, z1) and the coordinates of sensor B are (x2, y2, z2), the coordinate differences are x2 - x1, y2 - y1, and z2 - z1 respectively, and the spacing is the square root of the sum of the squares of each coordinate difference. The propagation speed is the ratio of the sensor spacing to the time difference of arrival. For example, if the spacing is 0.5 m and Δt is 0.001 s, then the propagation speed is 0.5 m divided by 0.001 s, obtaining 500 m / s.

[0088] Generate a path loss value based on the product of the amplitude attenuation coefficient and the propagation speed. The path loss value characterizes the comprehensive attenuation characteristics of the leakage signal on the propagation path. For example, if the amplitude attenuation coefficient is 1.2 V and the propagation speed is 500 m / s, then the path loss value is 1.2 V multiplied by 500 m / s, obtaining 600 V·m / s. The dimensional combination of the path loss value reflects the energy dissipation rate of the leakage signal, and the larger its value indicates that the leakage point is farther from the sensor or the medium attenuation is stronger.

[0089] Compare the path loss value with a preset loss threshold, and select the sensor groups with path loss values less than the loss threshold as candidate leakage nodes. The loss threshold is set according to the typical propagation characteristics of the leakage signal. For example, in a standard leakage experiment, the average path loss value at a distance of 1 meter from the leakage point is 400 V·m / s, so the loss threshold can be set to 400 V·m / s. During the screening, if the path loss value of a certain sensor group is 300 V·m / s and less than the threshold of 400 V·m / s, it is determined as a candidate leakage node; if the path loss value is 600 V·m / s and greater than the threshold of 400 V·m / s, it is excluded.

[0090] Conduct topological verification on the spatial distribution of candidate leakage nodes, and retain the sensor groups that meet the adjacency density condition after removing isolated nodes as valid candidate leakage nodes. The topological verification includes the following steps: construct a spatial position matrix of candidate leakage nodes, and calculate the three-dimensional Euclidean distance between each node; if the minimum distance between a certain node and other nodes exceeds the preset adjacency threshold, it is determined as an isolated node. The adjacency threshold is set according to the sensor layout density. For example, when the average sensor spacing is 0.5 meters, the adjacency threshold is set to 1.0 meter. The adjacency density condition requires that at least three nodes in the valid candidate leakage nodes are spaced less than the adjacency threshold from each other. For example, if the distances between nodes A, B, and C are 0.8 meters, 0.7 meters, and 0.9 meters respectively and are all less than 1.0 meter, they are retained as valid candidate leakage nodes; if the distances between node D and other nodes are all 1.2 meters and greater than 1.0 meter, it is excluded.

[0091] When calculating the amplitude attenuation coefficient, if there are significant differences in the frequency distribution of the candidate first spectral peak, initiate dispersion compensation correction. The dispersion compensation correction adjusts the amplitude gradient by introducing a relationship model between frequency and propagation speed. For example, when the main frequency difference of the candidate first spectral peak exceeds 10 Hz, the amplitude gradient is weighted and corrected according to the experimentally calibrated dispersion curve. The corrected amplitude gradient is used to recalculate the amplitude attenuation coefficient to improve the accuracy of the path loss value.

[0092] The calculation of the propagation speed also includes the calibration of the medium attenuation factor. The medium attenuation factor is obtained through a calibration experiment at a known leakage point. For example, in a no-leakage condition, a standard impact signal is injected, and the propagation speed between sensors is measured as a reference value; during actual detection, the ratio of the measured speed to the reference speed is used as the attenuation factor. The attenuation factor is used to correct the path loss value. For example, if the reference speed is 500 m / s and the measured speed is 450 m / s, the attenuation factor is 450 m / s divided by 500 m / s to get 0.9, and the corrected path loss value is the original value multiplied by 0.9.

[0093] The screening process of path loss values includes dynamic threshold adjustment. When the ambient noise level is high, the loss threshold is adjusted in real time according to the signal-to-noise ratio. For example, the basic threshold is 400 V·m / s. If the current noise energy is 6 dB higher than the standard working condition, the threshold is adjusted proportionally to 600 V·m / s. The dynamic adjustment is achieved by the look-up table method, and the mapping relationship between the noise energy and the threshold adjustment coefficient is pre-calibrated through experiments.

[0094] The spatial distribution verification of effective candidate leakage nodes also includes the direction consistency check. The direction consistency is achieved by calculating the difference in the direction angles of the node connection lines. For example, the direction angles of the position connection lines of effective candidate leakage nodes A, B, and C are 30°, 35°, and 40° respectively. The difference in direction angles within the preset threshold is considered consistent; if the difference in the direction angle of a certain node exceeds the threshold, it is determined as an abnormal node and excluded. The threshold of the direction angle difference is set according to the physical size of the leakage source. For example, when the leakage aperture is less than 1 mm, the threshold is set to 10°; when the aperture is greater than 5 mm, the threshold is set to 30°.

[0095] S4. Construct a spatial grid model with candidate leakage nodes as vertices, and solve the initial three-dimensional coordinate set of the leakage source through the time difference of arrival of vibration waveform signals. Specifically, it can be implemented as follows:

[0096] Construct a spatial grid model with the three-dimensional spatial coordinates of effective candidate leakage nodes as vertices, and the edges between the vertices are the actual distances between adjacent nodes. The three-dimensional spatial coordinates of effective candidate leakage nodes are determined according to the sensor layout positions. For example, the coordinates of effective candidate leakage node A are determined by the installation position of sensor A, and the coordinates of node B are determined by the installation position of sensor B. The edges between the vertices are generated by calculating the square root of the sum of the squares of the differences in the respective components of the three-dimensional coordinates of adjacent nodes. For example, the edge length between node A and node B is the square root of the sum of the squares of the differences in the x coordinates, y coordinates, and z coordinates of node A and node B. The topological relationship of the spatial grid model is represented by an adjacency matrix. For example, if node A is adjacent to node B, it is marked as 1 at the corresponding position in the matrix, and non-adjacent is marked as 0.

[0097] Calculate the propagation time difference vector of the leakage shock wave based on the arrival time differences of the vibration waveform signals among the effective candidate leakage nodes. The arrival time differences are obtained by comparing the detection timestamps of the same leakage shock wave at different nodes. For example, if the timestamp when node A detects the vibration waveform signal is t1 and the timestamp detected by node B is t2, the time difference between node A and node B is the result of t2 minus t1. The propagation time difference vector is composed of the time differences of all adjacent nodes arranged in sequence. For example, nodes A, B, and C form a triangular grid, and the time difference vector includes the time difference between node A and B, the time difference between node B and C, and the time difference between node A and C. The directionality of the time difference vector is determined by the spatial position relationship of the sensors. For example, if node A is located upstream of the leakage propagation path and node B is located downstream, the time difference from node A to node B is positive.

[0098] Based on the topological relationship of the vertices in the spatial grid model and the propagation time difference vector, establish a geometric constraint equation set for solving the leakage source location. Each equation in the geometric constraint equation set represents the relationship between the propagation time difference of a vertex pair and the spatial distance. For example, for nodes A and B, the equation is that the propagation speed multiplied by the time difference is equal to the three-dimensional spatial distance between node A and node B. The propagation speed is set according to the medium characteristics. For example, through a sound speed experiment in a metal shell, the measured propagation speed is 5000 m / s, and in the gas medium in the sealed cavity, it is set to the standard sound speed of 340 m / s. The leakage source location coordinates are introduced into the equation set as unknowns. For example, the distance from node A to the leakage source is the square root of the sum of the squares of the differences in the x, y, and z coordinates between the x coordinate of the leakage source and the x coordinate of node A. The distance from node B to the leakage source is calculated similarly, and the difference between the two distances is equal to the propagation speed multiplied by the time difference between node A and node B.

[0099] Use the least squares method to iteratively optimize the solution of the geometric constraint equation set to generate an initial three-dimensional coordinate set of the leakage source that meets the preset error threshold. The least squares method minimizes the sum of the squares of the residuals of all equations by gradually adjusting the leakage source coordinates. For example, the initial guessed coordinates are the geometric center point of the actuator, and the coordinates are adjusted in each iteration to reduce the residuals. The preset error threshold is set according to the positioning accuracy requirements. For example, the maximum allowable residual is 0.1 m, and the iteration stops when the sum of the squares of the residuals is less than 0.1. During the iteration process, if the sum of the squares of the residuals of a certain iteration is less than the threshold, the current coordinates are output. If convergence is not achieved after exceeding the maximum number of iterations, the solution is discarded. The initial three-dimensional coordinate set includes all coordinates that pass the threshold verification. For example, after 10 iterations, 3 sets of coordinates are obtained, and 2 of them have residuals of 0.08 m and 0.09 m, which are retained as valid solutions.

[0100] When constructing a spatial grid model, if the number of effective candidate leakage nodes is insufficient, virtual nodes are generated by interpolating adjacent nodes to supplement the model. The coordinates of the virtual nodes are calculated as the average of the coordinates of adjacent nodes. For example, if the coordinates of node A are (1.0, 2.0, 3.0) and the coordinates of node B are (2.0, 3.0, 4.0), the coordinates of the inserted virtual node are (1.5, 2.5, 3.5). The time difference of the virtual node is estimated by linear interpolation of the time differences of adjacent nodes. For example, if the time difference between node A and node B is 0.001 seconds, the time difference between the virtual node and node A is set to 0.0005 seconds. Virtual nodes are only used for model construction and do not participate in the output of the final leakage source coordinates.

[0101] The establishment of the geometric constraint equations also includes medium attenuation correction. The medium attenuation correction adjusts the propagation speed by introducing an attenuation factor. For example, at the interface between the sealed cavity and the shell, the propagation speed is set as the weighted average of the speeds of the two media according to the acoustic impedance difference at the interface. The attenuation factor is calibrated through experiments. For example, if the measured propagation speed at the interface is 2500 m / s and the originally set speed is 5000 m / s, the attenuation factor is 2500 / 5000 = 0.5, and the corrected propagation speed is the original speed multiplied by the attenuation factor.

[0102] During the least squares iterative optimization process, if the initial guessed coordinates deviate too far from the true leakage source, a multi-start initialization strategy is adopted. Multi-start initialization uniformly selects multiple initial coordinate points within the spatial grid model. For example, the center point, vertex positions, and midpoints of the edges of the model are selected as the initial guessed points, and the iterative optimization is performed respectively, and then the solution with the smallest residual is selected. The number of initial points is set according to the model complexity. For example, at least 3 initial points are selected for a triangular grid, and at least 4 initial points are selected for a tetrahedral grid.

[0103] After generating the initial three-dimensional coordinate set of the leakage source, physically infeasible solutions are excluded through coordinate validity verification. The validity verification includes boundary check and propagation consistency check. The boundary check requires that the coordinates are within the geometric range of the three-dimensional structure model of the actuator. For example, the x component of the coordinates is between 0 and the length L of the actuator; the propagation consistency check requires that the deviation between the theoretical time difference from the leakage source to each node and the measured time difference is less than the allowable error. For example, the deviation shall not exceed 10%. If a certain coordinate does not meet any of the conditions, it is excluded.

[0104] The dynamic adjustment of the preset error threshold is performed adaptively according to the environmental noise level. The dynamic adjustment rule is implemented through a mapping table of noise energy and threshold coefficient. For example, when the noise energy is 6 dB higher than the standard working condition, the error threshold is relaxed from 0.1 m to 0.15 m; when the noise energy is lower than the standard working condition, the threshold is tightened to 0.05 m. The mapping table is pre-calibrated through experiments. For example, the positioning error is tested under different noise levels to establish the correspondence between noise energy and the threshold.

[0105] S5. Project the initial three-dimensional coordinate set onto the surface of the actuator three-dimensional structure model, and generate a detection map based on the coupled analysis of the pressure signal modal space coherence and the vibration transient impact energy. Specifically, it can be implemented as follows:

[0106] Orthogonally project the initial three-dimensional coordinate set of the leakage source along the normal direction of the surface of the actuator three-dimensional structure model to generate the surface projection coordinates of the candidate leakage points. The normal direction of the surface of the actuator three-dimensional structure model is determined according to the geometric characteristics of the model. For example, the normal direction of the plane area is perpendicular to the plane, and the normal direction of the curved surface area points outward along the center of curvature. In the orthogonal projection operation, draw a perpendicular line from the initial three-dimensional coordinate to the model surface, and the coordinates of the foot of the perpendicular are the surface projection coordinates. For example, a point with an initial coordinate of (2.0, 3.0, 4.0) is projected to a surface coordinate of (2.0, 3.0, 3.8). When multiple initial coordinates are projected onto the same surface area, retain the point with the smallest projection spacing as the candidate leakage point. For example, when the projection spacing of two initial coordinates is less than 0.01 meters, they are merged into a single candidate point.

[0107] Extract the pressure signal modal space coherence matrix corresponding to the projection area of the candidate leakage points, and calculate the covariance ratio of the main diagonal elements to the non-diagonal elements in the pressure signal modal space coherence matrix as the modal coherence coefficient. The pressure signal modal space coherence matrix is constructed through the cross-correlation analysis of multi-sensor pressure signals. For example, the elements of the modal coherence matrix of sensors A, B, and C represent the modal correlations between sensor A and A, A and B, and A and C, respectively. The main diagonal elements are the self-modal correlations of the sensors. For example, the self-modal correlation of sensor A is set to 1.0; the non-diagonal elements are the cross-sensor modal correlations. For example, the modal correlation between sensor A and B is 0.8. The calculation method of the covariance ratio is: the sum of the covariances of the main diagonal elements divided by the sum of the covariances of the non-diagonal elements. For example, if the sum of the main diagonal covariances is 1.5 and the sum of the non-diagonal covariances is 0.6, then the ratio is 1.5 / 0.6 = 2.5. The larger the modal coherence coefficient, the stronger the modal correlation of the leakage signal between the sensors.

[0108] Synchronously extract the vibration transient shock energy value corresponding to the projected area of the candidate leakage point, and calculate the energy correction factor based on the propagation attenuation model of the leakage shock wave. The vibration transient shock energy value is obtained by integrating the square of the amplitude of the vibration signal within the projected time window of the candidate leakage point. For example, the energy value is obtained by accumulating the square of the amplitude of the vibration signal within a 0.1-second time window. The propagation attenuation model describes the characteristic that the vibration energy decays exponentially with the increase of the propagation distance. For example, the energy correction factor is inversely proportional to the propagation distance. For every 1-meter increase in the propagation distance, the energy value decays to 50% of the original value. The attenuation coefficient is calibrated through a standard leakage experiment. For example, if the measured energy value at a distance of 1 meter is 50% of the initial value, the attenuation coefficient is set to 0.693. The corrected vibration energy value is the result of multiplying the original energy value by the exponential function of the product of the attenuation coefficient and the propagation distance.

[0109] Fuse the modal coherence coefficient and the energy correction factor according to a preset weight to generate a leakage intensity factor. The preset weight is dynamically allocated according to the signal-to-noise ratio of the pressure signal and the vibration signal. For example, when the signal-to-noise ratio of the pressure signal is higher than that of the vibration signal, the weight of the modal coherence coefficient is set to 0.7, and the weight of the energy correction factor is set to 0.3; otherwise, the weights are set to 0.4 and 0.6. The weight allocation table is pre-calibrated through experiments. For example, when the signal-to-noise ratio is 20 dB, the weights are set to 0.6:0.4. The leakage intensity factor is calculated by weighted summation. For example, if the modal coherence coefficient is 2.5, the energy correction factor is 0.8, and the weight is 0.6:0.4, then the leakage intensity factor is 2.5 multiplied by 0.6 plus 0.8 multiplied by 0.4, resulting in 1.82.

[0110] Generate a two-dimensional detection map based on the distribution density of the surface projection coordinates of the candidate leakage points and the leakage intensity factor. The distribution density is calculated through the kernel density estimation algorithm. For example, a Gaussian distribution kernel function is constructed with each candidate leakage point as the center, and the kernel bandwidth is set to 0.05 meters according to the minimum feature size of the actuator surface. The leakage intensity factor participates in the calculation as the density weight. For example, there are three candidate leakage points in a certain area, and their leakage intensity factors are 1.5, 1.8, and 2.0 respectively. Then the weighted density of this area is the sum of the three, which is 5.3. The horizontal and vertical axes of the two-dimensional detection map correspond to the coordinates of the actuator surface, and the color depth is mapped to different color scales according to the weighted density value. For example, a density of 0-1 corresponds to light yellow, 1-3 corresponds to orange, 3-5 corresponds to red, and above 5 corresponds to dark red.

[0111] During the orthogonal projection process, if there are holes or discontinuous regions on the model surface, compensation coordinates are generated by interpolating the coordinates of adjacent projection points. The interpolation method is based on local surface fitting. For example, the coordinates of four projection points on the edge of the hole are selected, and the coordinates inside the hole are generated by fitting a quadratic surface equation. The interpolated coordinates are only used for density calculation and do not participate in the generation of the leakage intensity factor.

[0112] The calculation of the coherence matrix of the pressure signal modal space includes noise floor subtraction. The noise floor is determined by the coherence matrix of the pressure signal mode under the no-leakage condition. For example, when there is no leakage, the covariance of the main diagonal is 0.1, and the covariance of the non-diagonal is 0.05. During actual calculation, the noise floor value is subtracted from the measured covariance. For example, when the measured covariance of the main diagonal is 0.5, the effective covariance after subtracting 0.1 is 0.4.

[0113] The calculation of the energy correction factor takes into account the influence of medium anisotropy. The anisotropic attenuation model adjusts the attenuation coefficient according to the propagation direction. For example, the attenuation coefficient along the axial direction of the shell is 0.6, and the radial direction is 0.8. The correction factor is the sum of the product of the axial attenuation coefficient and the propagation distance plus the product of the radial attenuation coefficient and the propagation distance, and then takes the result of the exponential function. The anisotropy coefficient is obtained through multi-direction calibration experiments. For example, the distance when the energy decays to 50% along the axial direction is 1 meter, and the radial direction is 0.8 meter.

[0114] The fusion weight of the leakage intensity factor is dynamically adjusted according to the signal stability. The signal stability is evaluated by the coefficient of variation of the pressure and vibration signals. For example, when the coefficient of variation of the pressure signal is 0.1, it is determined to be highly stable, and the weight is increased to 0.8; when the coefficient of variation of the vibration signal is 0.3, it is determined to be low stable, and the weight is decreased to 0.2. The coefficient of variation threshold is set according to historical data statistics. For example, less than 0.2 is high stability.

[0115] After the two-dimensional detection map is generated, isolated noise points are eliminated by morphological filtering. The morphological filtering uses the opening operation of erosion first and then dilation. For example, the erosion operation removes isolated color blocks with an area less than 1 square centimeter, and the dilation operation restores the original contour of the effective area. Only the continuously distributed color blocks in the filtered detection map are retained as the effective leakage area.

[0116] S6. According to the coordinate distribution and energy proportion of the leakage area in the detection map, through spatial density clustering and leakage intensity grading analysis, a detection identification map of multiple leakage points is generated. Specifically, it can be implemented as:

[0117] Based on the coordinate distribution of the leakage areas in the two-dimensional detection map, a spatial density clustering algorithm is used to divide the spatial clusters of candidate leakage points, and the neighborhood radius is adaptively adjusted according to the energy proportion of the leakage areas. The spatial density clustering algorithm determines the neighborhood radius by calculating the weighted value of the spatial distance and the energy proportion between candidate leakage points. For example, when the energy proportion of a certain area exceeds 20% of the total energy, the neighborhood radius is set to 0.1 meter; when the energy proportion is 10%-20%, the radius is set to 0.05 meter; when it is less than 10%, the radius is set to 0.02 meter. The dynamic adjustment of the neighborhood radius is achieved by the look-up table method. For example, a mapping table of the preset energy proportion threshold and the radius is set, and the radius value is obtained by looking up the table according to the real-time energy proportion. During the clustering process, if the number of candidate leakage points within the neighborhood of a certain point exceeds the minimum point threshold (such as 5), it is determined as an independent spatial cluster; otherwise, it is marked as a noise point and eliminated.

[0118] Extract the maximum value and the mean value of the leakage intensity factors within each spatial cluster, and map the leakage intensity factors to the preset grading intervals based on the dynamic interval division strategy. The maximum value and the mean value of the leakage intensity factors are obtained by traversing the leakage intensity factors of all points within the spatial cluster. For example, a certain spatial cluster contains 10 candidate leakage points, and the leakage intensity factors are 1.5, 1.8, 2.0, etc. The maximum value is 2.0 and the mean value is 1.7. The dynamic interval division strategy determines the range of the grading intervals according to the ratio of the maximum value to the mean value. For example, when the maximum value / mean value ≥ 1.5, the interval is divided into three levels: high, medium, and low; when the ratio is 1.2-1.5, it is divided into two levels: high and low; when it is less than 1.2, it is regarded as a single level. The boundary values of the preset grading intervals are set according to the statistics of historical leakage data. For example, the high-grade interval is above the mean + standard deviation, the medium-grade is from the mean to the mean + standard deviation, and the low-grade is below the mean.

[0119] Assign a unique identifier to each spatial cluster according to the leakage intensity grading interval, and the graphic size of the identifier is proportional to the total value of the leakage intensity factors within the cluster. The graphic size of the unique identifier is determined by the product of the reference size and the total value of the leakage intensity factors. For example, the reference size is 5 millimeters, and the total value of the leakage intensity factors of a certain spatial cluster is 15.0, then the identifier size is 5×15 = 75 millimeters. The color of the identifier is set according to the grading interval. For example, the high grade is red, the medium grade is orange, and the low grade is yellow. The shape of the identifier is a circle, and the center position is the weighted centroid coordinates of the spatial cluster. For example, the centroid coordinates are calculated by the weighted average of the leakage intensity factors of each point within the cluster.

[0120] Overlay the identifier on the surface of the three-dimensional structure model of the actuator to generate a detection identification map with multiple leakage points. The overlay operation is achieved through coordinate mapping. For example, the centroid coordinates in the two-dimensional detection map are converted to the corresponding positions on the surface of the three-dimensional structure model of the actuator. The transparency of the identifier is dynamically adjusted according to the reliability of the leakage intensity factor. For example, when the standard deviation of the leakage intensity factor of the points within the spatial cluster exceeds 0.5, the transparency is set to 50%; when it is lower than 0.5, it is set to 100%. The detection identification map is output in vector graphics format, supporting zooming and hierarchical display. For example, when the user clicks on the identifier, detailed leakage intensity data can be viewed.

[0121] During the spatial density clustering process, if there are noise interference regions in the detection map, the noise filtering mechanism is activated. The noise filtering mechanism is achieved by calculating the energy proportion of the spatial cluster and its correlation with neighboring clusters. For example, when the energy proportion of a certain spatial cluster is lower than 1% and the minimum distance from neighboring clusters exceeds 0.2 meters, it is determined as noise and removed. After filtering, the neighborhood radius of the spatial cluster is recalculated. For example, the neighborhood radius of the remaining clusters after noise removal is increased by 10% according to the energy proportion.

[0122] The dynamic interval division strategy also includes smoothing the interval boundaries. The boundary smoothing is achieved by introducing a transition interval. For example, a ±5% transition band is set between the high-grade and medium-grade intervals, and the leakage intensity factors within the transition band are assigned colors by linear interpolation. The smoothing process avoids the color jump of the identifier caused by interval mutations. For example, when the leakage intensity factor is 1.49 (the lower limit of the transition band), the color is close to the high-grade red, and when it is 1.51 (the upper limit of the transition band), it is completely red.

[0123] The graphic size of the identifier is adaptively adjusted and dynamically scaled according to the resolution of the display device. For example, on a screen with a resolution of 1920×1080, the reference size is set to 5 pixels; at 4K resolution, the reference size is adjusted to 10 pixels. The scaling ratio is calculated based on the device pixel density. For example, when the pixel density is 300PPI, the scaling coefficient is 1.0, and when it is 600PPI, it is 2.0.

[0124] The generation of the detection identification map also includes overlaying the user interaction layer. The user interaction layer displays the historical detection data of the leakage source through a transparent layer. For example, when the user hovers over the identifier, the change curve of the leakage intensity factor for the last three detections of this cluster is displayed. The timestamps of the historical data are associated and stored with the detection results, supporting playback of the leakage evolution process along the time axis.

[0125] The unique identifier of the spatial cluster maintains position consistency during multiple detections. The position consistency is achieved by binding the identifier coordinates to the key feature points of the three-dimensional structure model of the actuator. For example, the center coordinates of the identifier are aligned with the bolt holes or weld positions on the model surface, with a deviation not exceeding 0.01 meters. The binding relationship is stored through a hash table. For example, the coordinate hash value of the key feature point is used as the position index of the identifier.

[0126] The output format of the detection identification map is compatible with industrial standard visualization tools. For example, a DWG format file is generated for editing by CAD software, or exported as a PNG image for embedding in reports. The format conversion is achieved through an intermediate data interface. For example, when converting vector graphics to bitmaps, hierarchical metadata is retained to ensure traceability for subsequent analysis.

[0127] Embodiment 2: Figure 2 The structural schematic diagram of an actuator sealing performance testing system according to the present invention is given. An actuator sealing performance testing system includes the following modules:

[0128] A pressurization and acquisition module, which is used to pressurize the sealing cavity of the actuator to be tested to a preset pressure value, and synchronously acquire the dynamic pressure signal in the sealing cavity and the vibration waveform signal of the housing;

[0129] A signal analysis module, which is used to perform time-frequency analysis on the dynamic pressure signal to extract the first spectral peak containing the pressure decay feature, and perform spectral analysis on the vibration waveform signal to extract the second spectral peak related to leakage impact;

[0130] A feature processing module, which is used to calculate the path loss value according to the amplitude gradient of the first spectral peak and the propagation time difference of the second spectral peak, and screen the sensor groups with path loss values less than the loss threshold as candidate leakage nodes;

[0131] A model construction module, which is used to construct a spatial grid model with candidate leakage nodes as vertices, and solve the initial three-dimensional coordinate set of the leakage source through the time difference of arrival of the vibration waveform signal;

[0132] A map generation module, which is used to project the initial three-dimensional coordinate set onto the surface of the actuator three-dimensional structure model, and generate a detection map based on the coupling analysis of the modal space coherence of the pressure signal and the vibration transient impact energy;

[0133] An identification generation module, which is used to generate a detection identification map of multiple leakage points through spatial density clustering and leakage intensity grading analysis according to the coordinate distribution and energy ratio of the leakage area in the detection map.

[0134] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and threshold selections in the calculations are set by those skilled in the art according to the actual situation.

[0135] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0136] Those of ordinary skill in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application of the technical solution and the invention constraints. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0137] In addition, in each embodiment of this application, the various functional modules can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0138] In several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there can be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces, and the indirect coupling or communication connection of the devices or modules can be in an electrical, mechanical, or other form.

[0139] As described above, this is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0140] Finally: The above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should all be included in the protection scope of the present invention.

Claims

1. A method for testing the sealing performance of an actuator, characterized in that: The steps include: S1. Pressurize the sealed cavity of the actuator to be tested to a preset pressure value, and simultaneously collect the dynamic pressure signal in the sealed cavity and the vibration waveform signal of the shell; S2, performing time-frequency analysis on the dynamic pressure signal to extract a first spectrum peak containing pressure attenuation characteristics, and performing spectrum analysis on the vibration waveform signal to extract a second spectrum peak related to leakage impact; S3, calculating the path loss value according to the amplitude gradient of the first spectrum peak and the propagation time difference of the second spectrum peak, and screening the sensor group whose path loss value is less than the loss threshold as the candidate leakage node; S4, constructing a spatial grid model with candidate leakage nodes as vertices, and solving the initial three-dimensional coordinate set of the leakage source through the arrival time difference of the vibration waveform signal; S5, projecting the initial three-dimensional coordinate set onto the surface of the actuator three-dimensional structure model, and generating a detection spectrum based on the coupling analysis of the pressure signal modal spatial coherence and the vibration transient impact energy; S6. According to the coordinate distribution and energy proportion of the leakage area in the detection map, a detection identification map of multiple leakage points is generated through spatial density clustering and leakage intensity classification analysis.

2. The actuator sealing test method according to claim 1, characterized in that: Pressurize the sealed cavity of the actuator to be tested to a preset pressure value, and simultaneously collect the dynamic pressure signal in the sealed cavity and the vibration waveform signal of the shell, including: After the actuator to be tested is fixed to the test bench, increasing pressure is applied in stages until the sealed chamber reaches a preset pressure value; When the pressure in the sealed cavity reaches a preset pressure value, the dynamic pressure signal in the sealed cavity is continuously collected at a preset sampling frequency through the pressure sensor group, and at the same time, the vibration waveform signal of the shell is collected at a synchronous sampling clock through the vibration sensor array attached to the surface of the shell; The dynamic pressure signal and the vibration waveform signal are aligned in the time domain to generate a multi-channel signal data set with consistent timestamps.

3. The actuator sealing test method according to claim 1, characterized in that: Perform time-frequency analysis on the dynamic pressure signal to extract the first spectrum peak containing the pressure attenuation characteristics, and perform spectrum analysis on the vibration waveform signal to extract the second spectrum peak related to the leakage impact, including: Performing bandpass filtering on the dynamic pressure signal in the multi-channel signal data set to retain the signal components within the preset frequency band; Performing time-frequency analysis on the filtered dynamic pressure signal, extracting a time-frequency spectrum through short-time Fourier transform, and identifying a frequency component in the time-frequency spectrum whose amplitude decay rate exceeds a first threshold as a first spectrum peak; Performing windowing processing on the vibration waveform signal in the multi-channel signal data set, generating a power spectrum density map by fast Fourier transform, and locating the frequency component in the power spectrum density map where the energy mutation amplitude exceeds the second threshold as the second spectrum peak; The first spectrum peak is waveform matched with the preset leakage attenuation feature template, and candidate first spectrum peaks whose matching degree exceeds the third threshold are screened; the second spectrum peak is correlated with the preset mechanical impact feature template, and candidate second spectrum peaks whose correlation coefficient exceeds the fourth threshold are screened.

4. The method for testing the sealing performance of an actuator according to claim 1, characterized in that: Calculating the path loss value according to the amplitude gradient of the first spectrum peak and the propagation time difference of the second spectrum peak, and screening the sensor group whose path loss value is less than the loss threshold as the candidate leakage node, including: Based on the frequency distribution and amplitude gradient of the candidate first spectrum peak, the amplitude attenuation coefficient of the received signal of each sensor on the leakage propagation path is calculated; Based on the arrival time difference of the candidate second spectrum peak and the sensor spacing, the propagation speed of the leakage impulse signal between the sensors is calculated; Generate a path loss value based on the product of the amplitude attenuation coefficient and the propagation speed; Compare the path loss value with a preset loss threshold, and select a sensor group whose path loss value is less than the loss threshold as a candidate leakage node; The spatial distribution of candidate leakage nodes is topologically verified, and after removing isolated nodes, sensor groups that meet the adjacency density conditions are retained as valid candidate leakage nodes.

5. The actuator sealing test method according to claim 1, characterized in that: Construct a spatial grid model with candidate leakage nodes as vertices, and solve the initial three-dimensional coordinate set of the leakage source through the arrival time difference of the vibration waveform signal, including: Construct a spatial grid model with the three-dimensional spatial coordinates of valid candidate leakage nodes as vertices, and the edges between vertices are the actual distances between adjacent nodes; According to the arrival time difference of the vibration waveform signal between each valid candidate leakage node, the propagation time difference vector of the leakage shock wave is calculated; Based on the topological relationship of vertices in the spatial grid model and the propagation time difference vector, a set of geometric constraint equations for solving the leakage source position is established; The least square method is used to iteratively optimize the solution of the geometric constraint equations to generate an initial three-dimensional coordinate set of the leakage source that meets the preset error threshold.

6. The method for testing the sealing performance of an actuator according to claim 1, characterized in that: The initial three-dimensional coordinate set is projected onto the surface of the actuator three-dimensional structure model. Based on the coupling analysis of the pressure signal modal spatial coherence and the vibration transient impact energy, a detection map is generated, including: Orthogonally projecting the initial three-dimensional coordinate set of the leakage source along the normal direction of the surface of the three-dimensional structure model of the actuator to generate the surface projection coordinates of the candidate leakage point; The modal space coherence matrix of the pressure signal corresponding to the projection area of ​​the candidate leakage point is extracted, and the covariance ratio of the main diagonal elements to the non-diagonal elements in the modal space coherence matrix of the pressure signal is calculated as the modal coherence coefficient; The vibration transient impact energy value corresponding to the projection area of ​​the candidate leakage point is extracted synchronously, and the energy correction factor is calculated based on the propagation attenuation model of the leakage shock wave; The modal coherence coefficient and the energy correction factor are fused according to preset weights to generate a leakage intensity factor; a two-dimensional detection map is generated according to the distribution density of the surface projection coordinates of the candidate leakage points and the leakage intensity factor.

7. The method for testing the sealing performance of an actuator according to claim 6, characterized in that: The color depth of the coordinate points in the two-dimensional detection map represents the leakage intensity level.

8. The method for testing the sealing performance of an actuator according to claim 1, characterized in that: According to the coordinate distribution and energy proportion of the leakage area in the detection map, through spatial density clustering and leakage intensity classification analysis, a detection identification map of multiple leakage points is generated, including: Based on the coordinate distribution of the leakage area in the two-dimensional detection map, the spatial density clustering algorithm is used to divide the spatial clusters of candidate leakage points; Extract the maximum and mean values ​​of the leakage intensity factor in each spatial cluster, and map the leakage intensity factor to a preset classification interval based on a dynamic interval partitioning strategy; A unique identifier is assigned to each spatial cluster according to the leakage intensity classification interval, and the graphic size of the identifier is proportional to the total value of the leakage intensity factor within the cluster; The identifiers are superimposed on the surface of the actuator three-dimensional structure model to generate a detection identification map of multiple leakage points. The colors of different identifiers in the detection identification map distinguish the leakage intensity levels.

9. The method for testing the sealing performance of an actuator according to claim 8, characterized in that: The neighborhood radius of the spatial density clustering algorithm when dividing the spatial clusters of candidate leakage points is adaptively adjusted according to the energy proportion of the leakage area.

10. An actuator sealing test system, used to implement an actuator sealing test method according to any one of claims 1 to 9, characterized in that: Includes the following modules: A pressurization acquisition module is used to pressurize the sealed cavity of the actuator to be tested to a preset pressure value, and simultaneously collect the dynamic pressure signal in the sealed cavity and the vibration waveform signal of the shell; A signal analysis module, used to perform time-frequency analysis on the dynamic pressure signal to extract a first spectrum peak containing pressure attenuation characteristics, and to perform spectrum analysis on the vibration waveform signal to extract a second spectrum peak related to leakage impact; A feature processing module, used to calculate a path loss value according to the amplitude gradient of the first spectrum peak and the propagation time difference of the second spectrum peak, and screen a sensor group whose path loss value is less than a loss threshold as a candidate leakage node; A model building module is used to build a spatial grid model with candidate leakage nodes as vertices, and solve the initial three-dimensional coordinate set of the leakage source through the arrival time difference of the vibration waveform signal; A map generation module is used to project the initial three-dimensional coordinate set onto the surface of the actuator three-dimensional structure model, and generate a detection map based on the coupling analysis of the pressure signal modal spatial coherence and the vibration transient impact energy; The identification generation module is used to generate a detection identification map of multiple leakage points based on the coordinate distribution and energy proportion of the leakage area in the detection map, through spatial density clustering and leakage intensity classification analysis.

Images