A method and system for detecting the air tightness of a ball valve

Abstract

The application relates to the technical field of valve detection, and discloses a method and system for detecting the air tightness of a ball valve. The method comprises the following steps: in a pressurization and pressure maintenance test, a multi-channel pressure information flow is synchronously collected, and a three-dimensional pressure field dynamic evolution sequence reflecting a pressure transmission process of a sealing interface is constructed. By cross-comparing the sequence with a leakage signal simulation library in a time-frequency domain, a potential leakage characteristic signal is extracted, and reverse tracing analysis of a leakage energy propagation path is carried out, so that the spatial coordinates and energy intensity of a suspected leakage source are determined. Finally, an overall leakage state index is calculated, a sealing state evaluation atlas and a maintenance operation sequence are generated. The method realizes a change from single-point pressure monitoring to three-dimensional pressure field dynamic analysis, can accurately locate a leakage source position, and improves the intuitiveness of detection and the pertinence of maintenance.

Description

Technical Field

[0001] This invention relates to the field of valve testing technology, specifically to a method and system for testing the airtightness of ball valves. Background Technology

[0002] Currently, the mainstream method for testing the airtightness of ball valves is single-point or few-point pressure monitoring based on the pressure decay method. This method typically involves pressurizing one side of the valve and setting up one or a limited number of pressure sensors in the test circuit. By monitoring the rate of decrease or trend of pressure change at that point over a period of time, the presence and extent of leakage in the valve can be determined. This technical approach obtains a single-dimensional curve of pressure change over time, reflecting the overall pressure holding capacity of the test system.

[0003] Existing detection methods of this type have shortcomings. Due to the limitations of sensor placement, the acquired pressure signals are highly localized and integrated. They cannot distinguish whether the pressure drop originates from leakage at the valve body sealing surface or at the pipeline connection, let alone indicate the specific location of the leak. When a leak is detected, only a qualitative or quantitative conclusion of "a leak exists" can be given, but the crucial question of "where is the leak source" remains unanswered, causing difficulties for subsequent maintenance and requiring disassembly and step-by-step troubleshooting. A single time-series pressure curve loses the spatial information of pressure transmission and distribution at the complex three-dimensional sealing interface of the valve. It cannot reveal the dynamic process of how the pressure field is formed, evolves, and ultimately destabilizes between the sealing surfaces during pressurization and pressure holding, nor can it indirectly assess the contact state of the sealing interface and the uniformity of pressure distribution.

[0004] A detection method is needed that can overcome the limitations of single-point pressure monitoring. This method needs to acquire dynamic pressure distribution information within the three-dimensional space of the valve, not just changes over time. Furthermore, it needs the ability to reverse engineer complex signals to pinpoint the specific leak source, thus achieving a technological leap from "determining if there is a leak" to "locating where the leak is," providing direct evidence for precise maintenance. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for testing the airtightness of ball valves, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a method for testing the airtightness of a ball valve, the method comprising:

[0007] In the preset pressurization and pressure holding test process, the multi-channel pressure change information flow of the medium flowing through the target ball valve is collected simultaneously;

[0008] The multi-channel pressure change information stream is processed to construct a dynamic evolution sequence of the pressure field, generating a dynamic evolution sequence of the pressure field of the target ball valve in three-dimensional space. The dynamic evolution sequence of the pressure field consists of continuous spatiotemporal data frames that reflect the pressure transmission process at the sealing interface.

[0009] The dynamic evolution sequence of the pressure field is cross-compared with a preset leakage signal simulation library in the time and frequency domain to extract potential leakage feature signals.

[0010] The potential leakage characteristic signals are analyzed by reverse tracing of the leakage energy propagation path to determine the spatial coordinates and energy intensity of at least one suspected leakage source;

[0011] Based on the spatial coordinates and energy intensity of the suspected leakage source, calculate the quantitative index of the overall leakage status of the target ball valve.

[0012] Based on the overall leakage status quantification index, a sealing status assessment map and maintenance operation sequence for the target ball valve are generated.

[0013] Preferably, the step of constructing a dynamic evolution sequence of the pressure field for the multi-channel pressure change information stream to generate a dynamic evolution sequence of the pressure field for the target ball valve in three-dimensional space includes:

[0014] At multiple equally spaced time points during the preset pressurization and pressure holding test process, instantaneous pressure values ​​from different spatial sampling points are captured;

[0015] All instantaneous pressure values ​​at each time point are mapped onto a preset three-dimensional digital mesh model of the target ball valve to generate a snapshot of the static pressure distribution corresponding to the time point.

[0016] According to the time sequence, multiple consecutive static pressure distribution snapshots are smoothly stitched together and transition calculations are performed to form a dynamic image showing the process of pressure transmission and attenuation between the sealing surface, valve seat and valve body.

[0017] Pressure gradient vectorization calculation is performed on each frame of data in the dynamic image to obtain the direction and rate of the fastest pressure change in each frame.

[0018] The pressure gradient vectorization calculation results of all frames are arranged according to the time axis and combined to form the dynamic evolution sequence of the pressure field.

[0019] Preferably, the step of performing time-frequency domain cross-comparison processing between the dynamic evolution sequence of the pressure field and a preset leakage signal simulation library to extract potential leakage feature signals includes:

[0020] Establish a leakage signal simulation library containing multiple standard leakage modes. Each standard leakage mode corresponds to a standard pressure field dynamic evolution segment under a known leakage aperture, location, and leakage rate.

[0021] From the dynamic evolution sequence of the pressure field, extract a segment to be analyzed that has the same time length as the standard dynamic evolution segment of the pressure field;

[0022] The segment to be analyzed is cross-correlated with each standard leakage pattern in the leakage signal simulation library to calculate a series of correlation coefficients.

[0023] From the series of correlation coefficients, correlation coefficients that exceed a preset correlation threshold are selected, and the standard leakage patterns corresponding to the correlation coefficients are marked as matching patterns.

[0024] Based on all matching patterns, the segment to be analyzed is decomposed to separate the residual fluctuation signal that is independent of the background pressure field change. The residual fluctuation signal is the potential leakage characteristic signal.

[0025] Preferably, the step of performing reverse tracing analysis of the leakage energy propagation path on the potential leakage characteristic signal to determine the spatial coordinates and energy intensity of at least one suspected leakage source includes:

[0026] Obtain the material acoustic parameters and structural boundary conditions of the three-dimensional solid model of the target ball valve;

[0027] Using the location and time of the potential leakage characteristic signal in the dynamic evolution sequence of the pressure field as the endpoint, and based on the material acoustic parameters and structural boundary conditions, a wave equation inverse solution model for the propagation of leakage energy in the medium is established.

[0028] By solving the wave equation in reverse, the possible propagation paths of leakage energy inside the target ball valve are deduced in reverse, until the energy converges to one or more starting points.

[0029] Each energy convergence starting point is marked as a suspected leakage source, and its spatial coordinates in the three-dimensional digital mesh model are recorded.

[0030] The energy dissipated from each of the suspected leakage sources to the signal acquisition endpoint is calculated and used as the energy intensity of the suspected leakage source.

[0031] Preferably, the step of calculating the overall leakage status quantification index of the target ball valve based on the spatial coordinates and energy intensity of the suspected leakage source includes:

[0032] The total leakage energy is estimated by summing the energy intensities of all suspected leakage sources.

[0033] The number of suspected leakage sources is counted, and the distribution density of the suspected leakage sources on the key sealing surface of the target ball valve is calculated by combining the spatial coordinates of the suspected leakage sources.

[0034] Calculate the dispersion index of energy intensity based on the energy intensity of each suspected leak source;

[0035] The total leakage energy estimate, the distribution density, and the dispersion index are input into a preset leakage state comprehensive calculation function. The output of the leakage state comprehensive calculation function is the overall leakage state quantification index of the target ball valve.

[0036] Preferably, in the preset pressurization and pressure holding test process, the multi-channel pressure change information stream of the medium flowing through the target ball valve is simultaneously acquired, including:

[0037] Pressure sensors are arranged in at least three annular regions of the target ball valve: the upstream cavity, the downstream cavity, and the area where the valve seat contacts the ball.

[0038] The pressurization device is activated to inject the detection medium into the upstream cavity of the target ball valve at a constant rate, while recording the readings of all pressure sensors to form an information flow for the pressurization phase.

[0039] When the upstream chamber pressure reaches the preset test pressure, the pressurization device is stopped and the pressure holding stage begins. The readings of all pressure sensors are continuously recorded during the preset pressure holding time period to form a pressure holding stage information stream.

[0040] The information streams of the pressurization stage and the pressurization holding stage are merged in chronological order to form the multi-channel pressure change information stream.

[0041] Preferably, after extracting the potential leakage feature signals, the method further includes a step of removing environmental interference from the potential leakage feature signals:

[0042] Acquire background vibration and temperature drift signals in the test environment;

[0043] The background vibration signal and temperature drift signal are respectively compared with the potential leakage characteristic signal for coherence analysis;

[0044] The purified leakage characteristic signal is obtained by subtracting the signal component that is highly coherent with the background vibration signal and temperature drift signal from the potential leakage characteristic signal.

[0045] The purified leakage characteristic signal is used to update the potential leakage characteristic signal for subsequent reverse tracing analysis of the leakage energy propagation path.

[0046] Preferably, the generation of the sealing condition assessment map and maintenance operation sequence for the target ball valve includes:

[0047] Using a three-dimensional digital grid model as a carrier, the overall leakage status quantitative indicators, spatial coordinates and energy intensity of each suspected leakage source are visualized and rendered with different colors and brightness to generate the sealing status assessment map.

[0048] Based on the spatial coordinates of the suspected leak source, mark the areas that need to be prioritized for treatment on the sealing status assessment map;

[0049] For each area that needs to be prioritized, the corresponding sealing surface treatment process parameters and operation steps are matched from the preset operation knowledge base based on its corresponding energy intensity.

[0050] All matched operation steps are sorted according to the processing priority of their corresponding regions to form the maintenance operation sequence.

[0051] Preferably, after generating the sealing condition assessment map and maintenance operation sequence for the target ball valve, a verification and feedback step is also included:

[0052] After completing the maintenance according to the maintenance operation sequence, the same target ball valve is tested again using the above-described detection method to obtain a new quantitative index of the overall leakage status.

[0053] Compare the two overall leakage status quantitative indicators before and after maintenance, and calculate the improvement rate;

[0054] If the improvement rate is lower than a preset threshold, the operation effectiveness coefficient corresponding to the process parameters used in this maintenance is adjusted in the maintenance operation knowledge base, and detailed data of this maintenance is recorded to expand the preset leakage signal simulation library.

[0055] Preferably, the present invention also includes a ball valve airtightness testing system, the system including at least one processor and at least one memory; the at least one memory stores a computer program, which, when executed by the at least one processor, causes the system to perform the ball valve airtightness testing method as described above.

[0056] Compared with the prior art, the beneficial effects of the present invention are:

[0057] By simultaneously acquiring multi-channel pressure change information streams of the medium flowing through the target ball valve during a pre-defined pressurization and pressure holding process, and constructing a dynamic evolution sequence of the pressure field composed of continuous spatiotemporal data frames, this method achieves a three-dimensional dynamic digital reconstruction of the entire pressure transmission process at the sealing interface. Compared to obtaining only a single pressure-time curve, this approach generates a four-dimensional dataset containing both spatial and temporal dimensions. This allows for a visual representation of the pressure wavefront's expansion path between the ball valve's sealing pairs, its final stable distribution, and the distortion and collapse of the pressure field during leakage. This transforms detection from interpreting the final result to monitoring the dynamic process of seal formation and failure, enabling the capture of abnormal states such as uneven pressure distribution and local weak points that are undetectable by traditional methods. It provides an unprecedented full-field spatiotemporal information foundation for in-depth evaluation of sealing performance.

[0058] By cross-comparing the constructed dynamic evolution sequence of the pressure field with a pre-set leakage signal simulation library in the time and frequency domains, potential leakage characteristic signals are extracted. Further reverse tracing analysis of the leakage energy propagation path of these signals can determine the spatial coordinates and energy intensity of suspected leakage sources. The collected full-field pressure disturbance is considered as a complex field formed by energy waves emitted from the leakage point propagating in the medium and reflecting and superimposing at the boundaries. The wave source is solved inversely using an algorithm. This allows for the precise three-dimensional location and relative leakage intensity of the microscopic leakage point causing the macroscopic pressure drop phenomenon. This realizes the transformation of leakage detection from system-level judgment to precise component-level location. The output is no longer an abstract leakage rate value, but a suspected fault point with clear spatial coordinates, guiding maintenance personnel to directly check and handle the specified location, improving the guidance value of detection and maintenance efficiency. Attached Figure Description

[0059] Figure 1 This is a schematic diagram illustrating the working principle of the ball valve airtightness testing method described in this invention.

[0060] Figure 2 A flowchart for constructing the dynamic evolution sequence of the pressure field;

[0061] Figure 3 A flowchart for extracting potential leakage characteristic signals;

[0062] Figure 4 A comparison diagram of pressure changes between the upstream and downstream chambers of the ball valve and the valve seat sensor during the pressure holding stage;

[0063] Figure 5 This is the dynamic evolution curve of the pressure field for the ball valve airtightness test. Detailed Implementation

[0064] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0065] Please see Figure 1 This invention provides a method for detecting the airtightness of a ball valve. The method includes: simultaneously acquiring multi-channel pressure change information streams of the medium flowing through the target ball valve during a preset pressurization and pressure holding test process; constructing a dynamic evolution sequence of the pressure field from the multi-channel pressure change information stream to generate a dynamic evolution sequence of the pressure field of the target ball valve in three-dimensional space, which consists of continuous spatiotemporal data frames reflecting the pressure transmission process at the sealing interface; subsequently, performing time-frequency domain cross-comparison processing between the dynamic evolution sequence of the pressure field and a preset leakage signal simulation library to extract potential leakage characteristic signals; performing reverse tracing analysis of the leakage energy propagation path on the potential leakage characteristic signals to determine the spatial coordinates and energy intensity of at least one suspected leakage source; calculating a quantitative index of the overall leakage state of the target ball valve based on the spatial coordinates and energy intensity of the suspected leakage source; and finally, generating a sealing state evaluation map and maintenance operation sequence for the target ball valve based on the quantitative index of the overall leakage state.

[0066] Example 1: See Figure 2 At multiple equally spaced time points during the pre-defined pressurization and pressure holding test process, instantaneous pressure values ​​from different spatial sampling points are captured. All instantaneous pressure values ​​at each time point are mapped onto a pre-defined three-dimensional digital mesh model of the target ball valve, generating a static pressure distribution snapshot corresponding to that time point. Following a chronological order, multiple consecutive static pressure distribution snapshots are smoothly stitched together and transition calculations are performed to form a dynamic image demonstrating the pressure transmission and attenuation process between the sealing surface, valve seat, and valve body. Pressure gradient vectorization calculations are performed on each frame of the dynamic image to obtain the direction and rate of the fastest pressure change in each frame. The pressure gradient vectorization calculation results of all frames are arranged along the time axis and combined to form a dynamic evolution sequence of the pressure field.

[0067] In practical implementation, instantaneous pressure values ​​from different spatial sampling points in the upstream cavity, downstream cavity, and valve seat region need to be captured at multiple equally spaced time nodes during the preset pressurization and pressure holding test process. In some embodiments, the interval between the equally spaced time nodes is 10 milliseconds, and the instantaneous pressure values ​​of all eight pressure sampling points are collected synchronously at each time node. All instantaneous pressure values ​​at each time node are mapped onto a preset three-dimensional digital mesh model of the target ball valve. This preset three-dimensional digital mesh model contains 100,000 mesh nodes, and the instantaneous pressure values ​​are allocated to the nearest mesh node through a spatial interpolation algorithm to generate a snapshot of the static pressure distribution corresponding to the time node.

[0068] Following a chronological order, one thousand consecutive snapshots of static pressure distribution are smoothly stitched together and transitionally calculated. This smoothing and transition calculation employs a cubic spline interpolation algorithm in the time domain to fit the pressure values ​​of each grid node between adjacent snapshots, forming a dynamic image demonstrating the transmission and attenuation of pressure between the sealing surface, valve seat, and valve body. Pressure gradient vectorization is performed on each frame of the dynamic image, calculating the direction and rate of the fastest pressure change based on the pressure difference between each grid node and its neighboring nodes. The pressure gradient vectorization results of all frames are arranged along the time axis and combined to form a dynamic evolution sequence of the pressure field. This dynamic evolution sequence is a four-dimensional data array, with dimensions including time, three-dimensional spatial coordinates, and the direction and magnitude of the pressure gradient vector.

[0069] In some embodiments, the formula for calculating pressure gradient vectorization is:

[0070] ;

[0071] in: Indicates time Spatial coordinates Pressure gradient vector at that point This indicates the pressure value at that point. , , These represent the unit direction vectors of the three-dimensional coordinate axes. In the formula... , , It is the pressure partial derivative calculated on a preset three-dimensional digital mesh model using the central difference method.

[0072] In practical implementation, generating a static pressure distribution snapshot requires a precise match between a pre-defined 3D digital mesh model and the geometric dimensions and sensor locations of the physical ball valve. The mapping process maps the physical coordinates of the pressure sensor to the coordinates of the nearest mesh node in the pre-defined 3D digital mesh model, using this node as a constraint point for the known pressure value. It can be understood that for areas in the pre-defined 3D digital mesh model where no sensors are located, the pressure values ​​of the mesh nodes need to be calculated from the pressure values ​​of the known constraint points using a spatial interpolation algorithm based on radial basis functions. Optionally, a Gaussian kernel function is used as the radial basis function. The specific implementation of using a Gaussian kernel function for the radial basis function is as follows: using the spatial coordinates of each known sensor point as the center point, the Euclidean distance between the unsampled mesh nodes and each center point is calculated, and this distance is used as the input to the Gaussian kernel function. By adjusting the bandwidth parameter of the kernel function, the decay rate of the weights is controlled, so that closer sensor points contribute a larger weight to the interpolation result, thereby generating a smooth and continuous pressure distribution estimate. This logic for constructing the dynamic evolution sequence of the pressure field can be adapted to the detection requirements of high-pressure ball valves in energy storage systems and liquid-cooled loop ball valves in data centers. Ball valves in these scenarios often face pressure fluctuations caused by the circulation of the medium. By accurately mapping the pressure transmission process through a three-dimensional digital mesh model, subtle signs of leakage caused by long-term dynamic pressure on the sealing surface can be captured, providing adaptability support for monitoring the sealing performance of ball valves in different application scenarios.

[0073] Example 2: See Figure 3 A leakage signal simulation library containing multiple standard leakage modes was established. Each standard leakage mode corresponds to a standard pressure field dynamic evolution segment with a known leakage aperture, location, and leakage rate. From the pressure field dynamic evolution sequence, a segment of the same time length as the standard pressure field dynamic evolution segment was extracted for analysis. This segment was then subjected to convolutional cross-correlation with each standard leakage mode in the leakage signal simulation library, yielding a series of correlation coefficients. Correlation coefficients exceeding a preset correlation threshold were selected from this series, and the corresponding standard leakage modes were marked as matching modes. Based on all matching modes, the segment was decomposed to separate residual fluctuation signals unrelated to background pressure field changes; these residual fluctuation signals are the potential leakage characteristic signals. After extracting the potential leakage characteristic signal, the following steps are performed to remove environmental interference: background vibration signal and temperature drift signal are collected from the test environment, and coherence analysis is performed between the background vibration signal and the temperature drift signal and the potential leakage characteristic signal. The signal components with high coherence with the background vibration signal and the temperature drift signal are subtracted from the potential leakage characteristic signal to obtain the purified leakage characteristic signal. The purified leakage characteristic signal is used to update the potential leakage characteristic signal for subsequent reverse tracing analysis of the leakage energy propagation path.

[0074] In practical implementation, a leakage signal simulation library containing multiple standard leakage modes is established. This library is constructed by performing the same testing procedure on calibrated ball valves with known leakage orifice diameters, locations, and leakage rates. Each standard leakage mode corresponds to a standard pressure field dynamic evolution segment under known leakage conditions. In some embodiments, the leakage signal simulation library contains fifty standard leakage modes, each standard pressure field dynamic evolution segment lasting five hundred milliseconds. From the pressure field dynamic evolution sequence, a segment of the same duration as the standard pressure field dynamic evolution segment is extracted for analysis. The extraction operation starts from the beginning of the pressure field dynamic evolution sequence and is performed using a sliding time window with a sliding step size of ten milliseconds.

[0075] In practice, the segment to be analyzed is cross-correlated with each standard leakage mode in the leakage signal simulation library using convolutional cross-correlation to calculate a series of correlation coefficients. These correlation coefficients quantify the similarity between the segment to be analyzed and each standard leakage mode in the time-frequency domain. In some embodiments, the Pearson correlation coefficient is calculated as the correlation coefficient. The calculation formula is:

[0076] ;

[0077] in: This represents the calculated correlation coefficient. This represents the total number of data points contained in the segment to be analyzed or the standard pressure field dynamic evolution segment. Indicates the first segment in the segment to be analyzed. The magnitude of the pressure gradient vector for each data point. This represents the average magnitude of the pressure gradient vector across all data points in the segment to be analyzed. This represents the first segment of the standard pressure field dynamic evolution corresponding to the standard leakage mode. The magnitude of the pressure gradient vector for each data point. This represents the average magnitude of the pressure gradient vector across all data points in the standard pressure field dynamic evolution segment. Correlation coefficients exceeding a preset correlation threshold (set to 0.7) are selected from a series of correlation coefficients, and the corresponding standard leakage patterns are marked as matching patterns.

[0078] Understandably, based on all matching patterns, signal decomposition is performed on the segment to be analyzed. This decomposition employs a matching pursuit-based algorithm, iteratively subtracting components linearly correlated with each matching pattern from the segment to separate residual fluctuation signals unrelated to changes in the background pressure field. These residual fluctuation signals are the potential leakage characteristic signals. After extracting the potential leakage characteristic signals, an environmental interference stripping step is performed. Background vibration and temperature drift signals from the test environment are acquired. The background vibration signal is obtained using an accelerometer mounted on the test platform base, with a sampling frequency of 1 kHz. The temperature drift signal is obtained using a temperature sensor attached to the surface of the target ball valve body, with a sampling frequency of 10 Hz.

[0079] In practical implementation, the background vibration signal and temperature drift signal are respectively subjected to coherence analysis with the potential leakage characteristic signal. The coherence analysis calculates the coherence function values ​​of the background vibration signal, temperature drift signal, and potential leakage characteristic signal in multiple frequency bands. From the potential leakage characteristic signal, signal components with high coherence to the background vibration signal and temperature drift signal are subtracted. The criterion for high coherence is a coherence function value greater than 0.8 in the main frequency band. This subtraction operation is implemented through an adaptive filtering algorithm, yielding the purified leakage characteristic signal. For ball valves in cooling systems in the new energy field, the medium characteristics may cause leakage signals to exhibit special fluctuation patterns. The environmental interference stripping steps and signal decomposition logic can be specifically adapted to such scenarios, effectively eliminating the influence of background vibration and temperature drift, and improving the accuracy of leakage characteristic signal extraction.

[0080] Example 3: Obtain the material acoustic parameters and structural boundary conditions of the three-dimensional solid model of the target ball valve. Using the location and time of the potential leakage characteristic signal appearing in the dynamic evolution sequence of the pressure field as the endpoint, a wave equation inverse solution model for the propagation of leakage energy in the medium is established based on the material acoustic parameters and structural boundary conditions. By solving the wave equation inverse solution model, the propagation path of leakage energy inside the target ball valve is deduced in reverse until the energy converges to one or more starting points. Each energy convergence starting point is marked as a suspected leakage source, and its spatial coordinates in the three-dimensional digital mesh model are recorded. The energy dissipated from each suspected leakage source to the signal acquisition endpoint is calculated, and this is used as the energy intensity of the suspected leakage source. The energy intensities of all suspected leakage sources are summed to obtain the total leakage energy estimate. The number of suspected leakage sources is counted, and the distribution density of suspected leakage sources on the key sealing surface of the target ball valve is calculated based on the spatial coordinates of the suspected leakage sources. Based on the energy intensity of each suspected leakage source, the dispersion index of energy intensity is calculated. The total leakage energy estimate, distribution density, and dispersion index are input into the preset leakage state comprehensive calculation function. The output of the leakage state comprehensive calculation function is the overall leakage state quantitative index of the target ball valve.

[0081] In practical implementation, the material acoustic parameters and structural boundary conditions of the target ball valve's three-dimensional solid model are obtained. The material acoustic parameters include medium density, medium sound velocity, valve body material density, and valve body material acoustic impedance. The structural boundary conditions include the geometry of the ball valve's internal cavity, the spatial orientation of the sealing surface, and the precise coordinates of the pressure sensors. Using the location and time of the potential leakage characteristic signal appearing in the dynamic evolution sequence of the pressure field as the endpoint, the location information is the coordinates of the grid nodes where the pressure gradient vector undergoes abnormal fluctuations in the dynamic evolution sequence of the pressure field, and the time information is the timestamp of the recorded abnormal fluctuation. Based on the material acoustic parameters and structural boundary conditions, an inverse solution model for the wave equation of leakage energy propagation in the medium is established.

[0082] In some embodiments, the inverse solution model of the wave equation is constructed based on the three-dimensional non-homogeneous acoustic wave equation, and its inverse solution process is performed in the time domain. By solving the inverse solution model of the wave equation, the propagation path of leakage energy inside the target ball valve is deduced in reverse. The solution method employs time-reversal mirror technology, which uses the acquired potential leakage characteristic signal as the source signal and performs time-reversal calculations in the numerical model until the energy converges to one or more starting points at the zero point of the inversion time. Each energy convergence starting point is marked as a suspected leakage source, and its spatial coordinates in the three-dimensional digital grid model are recorded. These coordinates are expressed with millimeter precision in the three-dimensional Cartesian coordinate system. The energy dissipated from each suspected leakage source to the signal acquisition endpoint is calculated, and this is used as the energy intensity of the suspected leakage source. The energy intensity is calculated based on the integral attenuation value of the energy flux density along the propagation path in the inverse solution model of the wave equation.

[0083] In practical implementation, based on the spatial coordinates and energy intensity of suspected leakage sources, a quantitative index of the overall leakage state of the target ball valve is calculated. The energy intensities of all suspected leakage sources are summed to obtain a total leakage energy estimate, expressed in microjoules. The number of suspected leakage sources is counted, and combined with their spatial coordinates, the distribution density of these sources on the key sealing surface of the target ball valve is calculated. The key sealing surface is predefined as the annular contact area between the ball and the upstream and downstream valve seats, and the distribution density is expressed in units per square centimeter. Based on the energy intensity of each suspected leakage source, an energy intensity dispersion index is calculated, characterized by the coefficient of variation.

[0084] It can be understood that the total leakage energy estimate, distribution density, and dispersion index are input into a preset leakage state comprehensive calculation function, which is a multivariate scalar function. In some embodiments, the leakage state comprehensive calculation function... It has the following form:

[0085] ;

[0086] in: This represents a quantitative index indicating the overall leakage status of the target ball valve in the calculated output. This represents the estimated total leakage energy. This indicates the distribution density of suspected leak sources on the critical sealing surface of the target ball valve; The coefficient of variation represents the energy intensity, which is the ratio of the standard deviation to the mean of the energy intensity. , , The preset weighting coefficients are determined by fitting calibration data from a historical leakage case database. The output of the leakage state comprehensive calculation function is a quantitative index of the overall leakage state of the target ball valve. Optional, the weighting coefficients... The value is 0.5. The value is 0.3. The value is set to 0.2. Cooling circuit ball valves in the power industry require high operational stability. Quantitative indicators of overall leakage status can directly reflect the impact of the ball valve's sealing performance on the system. Accurate determination of the spatial coordinates and energy intensity of the leakage source can provide reliable detection data support for safe operation in such scenarios.

[0087] Example 4: Pressure sensors are arranged in at least three annular areas of the target ball valve: the upstream cavity, the downstream cavity, and the area where the valve seat contacts the ball. A pressurization device is activated to inject the detection medium into the upstream cavity of the target ball valve at a constant rate, while simultaneously recording the readings of all pressure sensors to form a pressurization phase information stream. When the pressure in the upstream cavity reaches the preset test pressure, the pressurization device is stopped, and the pressure holding phase begins. The readings of all pressure sensors are continuously recorded during the preset pressure holding time period to form a pressure holding phase information stream. The pressurization phase information stream and the pressure holding phase information stream are merged in chronological order to form a multi-channel pressure change information stream.

[0088] In practical implementation, pressure sensors are arranged in at least three annular regions of the target ball valve: the upstream cavity, the downstream cavity, and the area where the valve seat contacts the ball. One pressure sensor is arranged in each of the upstream and downstream cavities, and six pressure sensors are evenly distributed around the annular area where the valve seat contacts the ball, forming a multi-channel data acquisition array of eight pressure sensors. The pressure sensors are connected to a high-speed data acquisition card, whose synchronous sampling frequency is set to 2kHz to ensure strict time alignment of pressure readings from all channels. Refer to Table 1 for the specific arrangement parameters of the pressure sensors.

[0089] Table 1. Pressure Sensor Arrangement Parameters:

[0090]

[0091] In some embodiments, a pressurization device is activated to inject a detection medium into the upstream cavity of the target ball valve at a constant rate, which is set to increase by 0.5 MPa per minute. Simultaneously, the readings of all pressure sensors are recorded, forming a pressurization phase information stream, which is an eight-channel time-series data set. When the upstream cavity pressure reaches a preset test pressure (6.0 MPa), the pressurization device is stopped, and the pressure holding phase begins. The readings of all pressure sensors are continuously recorded during a preset pressure holding time period of 120 seconds, forming a pressure holding phase information stream.

[0092] In practice, the information streams from the pressurization and holding phases are merged chronologically to form a multi-channel pressure change information stream. This multi-channel pressure change information stream forms the original data basis for subsequently constructing the dynamic evolution sequence of the pressure field. It can be understood that the stability of the information stream during the holding phase is the foundation for evaluation, and the pressure readings during the holding phase... The stability criterion formula is:

[0093] ;

[0094] in: The stability coefficient representing the pressure reading during the pressure holding phase. This represents the standard deviation of any pressure sensor reading sequence within a preset pressure holding time period. This represents the average value of the same sequence. In some embodiments, the stability coefficient of all pressure sensors is required. The pressure holding stage is considered effective only if all pressure fluctuations are less than 0.5%. Optionally, the detection medium is dry, oil-free nitrogen. Optionally, the pressurization device is a precision gas booster pump, with output pressure fluctuations controlled within ±0.5%. Ball valves in data centers and new energy storage systems have stringent requirements for detection accuracy due to their installation environment. A flexible arrangement of multi-channel pressure sensors can adapt to ball valves of different structures, ensuring accurate acquisition of pressure change information during the pressurization and pressure holding stages, providing a reliable data foundation for subsequent detection and analysis.

[0095] See Figure 4The data presented during the pressure holding phase primarily demonstrates the dynamic pressure changes in the upstream (S1), downstream (S2), and valve seat sensor average values ​​of the target ball valve. Specifically, this phase corresponds to the pressure holding process at a preset test pressure of 6.0 MPa. Data acquisition was based on an 8-channel pressure sensor array (including S1, S2, and 6 valve seat annular area sensors), with a synchronous sampling frequency of 2 kHz. The time axis (720-724 seconds) in the figure represents a local time period within the pressure holding phase. From the pressure fluctuation characteristics, the average values ​​of S1, S2, and the valve seat sensors all fluctuate slightly around the 6.0 MPa benchmark, meeting the core requirements for stability assessment during the pressure holding phase (stability coefficient must be less than 0.5%). The fluctuation amplitude of the valve seat sensor average value is more gradual, reflecting the suppression effect of multi-sensor data averaging on local interference. Furthermore, the pressure changes in S1 and S2 show weak correlation fluctuations, reflecting the dynamic transmission process of the medium at the ball valve sealing interface. This type of fluctuation is one of the key spatiotemporal data frame sources for subsequently constructing the dynamic evolution sequence of the pressure field.

[0096] Example 5: Using a 3D digital mesh model as a carrier, the overall leakage status quantification index, spatial coordinates and energy intensity of each suspected leakage source are visualized and rendered with different colors and brightness to generate a sealing status assessment map. Based on the spatial coordinates of the suspected leakage sources, areas requiring priority treatment are marked on the sealing status assessment map. For each area requiring priority treatment, the corresponding sealing surface treatment process parameters and operation steps are matched from a preset operation knowledge base according to its corresponding energy intensity. All matched operation steps are sorted according to the treatment priority of their corresponding areas to form a maintenance operation sequence. After generating the sealing status assessment map and maintenance operation sequence for the target ball valve, a verification and feedback step is performed: After completing the maintenance according to the maintenance operation sequence, the detection method is executed again on the same target ball valve to obtain a new overall leakage status quantification index. The two overall leakage status quantification indices before and after maintenance are compared, and the improvement rate is calculated. If the improvement rate is lower than a preset threshold, the operation effectiveness coefficient corresponding to the process parameters used in this maintenance is adjusted in the maintenance operation knowledge base, and detailed data of this maintenance is recorded to expand the preset leakage signal simulation library.

[0097] In practical implementation, a three-dimensional digital mesh model is used as the carrier to visualize the overall leakage status quantification indicators, spatial coordinates, and energy intensity of each suspected leakage source using different colors and brightness levels. The color mapping rule maps the pressure gradient values ​​to a continuous color spectrum from blue to red, and the brightness mapping rule is proportional to the energy intensity value of the suspected leakage source, generating a sealing status assessment map. Based on the spatial coordinates of the suspected leakage sources, areas requiring priority treatment are marked on the sealing status assessment map. The marking method involves overlaying a bright, flashing 3D marker box at the corresponding mesh node of the three-dimensional digital mesh model.

[0098] In some embodiments, for each area requiring priority treatment, the corresponding sealing surface treatment process parameters and operation steps are matched from a preset operation knowledge base based on its corresponding energy intensity. The matching logic involves searching for a preset range of energy intensity values ​​in the operation knowledge base and extracting the polishing grit size, grinding pressure, and reciprocating frequency parameters associated with that range. All matched operation steps are sorted according to the processing priority of their corresponding areas. The processing priority is calculated based on a combination of energy intensity values ​​and whether the area is located on a primary or secondary sealing surface, forming a maintenance operation sequence. The maintenance operation sequence is output in the form of a structured chemical list.

[0099] It is understandable that after generating the sealing condition assessment map and maintenance operation sequence for the target ball valve, verification and feedback steps are also included. After completing the maintenance according to the maintenance operation sequence, the detection method is performed again on the same target ball valve to obtain a new overall leakage condition quantification index. The two overall leakage condition quantification indices before and after maintenance are compared to calculate the improvement rate. The calculation formula is:

[0100] ;

[0101] in: Indicates the maintenance improvement rate; This represents a quantitative indicator of the overall leakage status of the target ball valve before maintenance. This is a quantitative indicator representing the overall leakage status of the target ball valve after maintenance.

[0102] In practice, if the improvement rate is lower than a preset threshold (set at 70%), the operational effectiveness coefficient corresponding to the process parameters used in this maintenance is adjusted in the maintenance operation knowledge base. The initial value of the operational effectiveness coefficient is 1.0, and the adjustment method is to multiply it by a decay factor less than 1 calculated based on the improvement rate. Detailed data for this maintenance is recorded, including the maintenance operation sequence, the actual executed process parameters, and quantitative indicators of the overall leakage status before and after maintenance. This data is used to expand the preset leakage signal simulation library. The expansion method involves adding the post-maintenance test data as a new training sample to the update queue of the leakage signal simulation library. For high-pressure ball valves in new energy storage systems, the sealing status of their key sealing surfaces directly affects system safety. The sealing status assessment map can intuitively display the risk distribution, facilitating maintenance personnel to quickly locate high-risk areas. Maintenance of ball valves in cooling systems needs to be combined with scenario characteristics; the maintenance operation sequence can accurately match the sealing surface treatment process, improving the targeted nature of maintenance.

[0103] See Figure 5In the pressurization and holding stages of ball valve airtightness testing, the dynamic evolution of the pressure field relies on the synchronous acquisition of multi-channel pressure signals. Specifically, the pressure field data consists of pressure change information streams from the upstream cavity, downstream cavity, and valve seat region. The horizontal axis represents test time (seconds), and the vertical axis represents pressure (MPa). During the pressurization stage (0-60 seconds), the pressure in each region increases linearly and synchronously, reflecting the uniform transmission process of the medium inside the ball valve. During the holding stage (after 60 seconds), the pressure in each region enters a dynamic fluctuation state, and the amplitude and coordination of these fluctuations reflect the pressure transmission characteristics of the sealing interface. In the parameter configuration process, the stage boundary is set to 60 seconds, covering the complete transition from pressurization to holding in the preset test procedure.

[0104] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0105] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for detecting the gas tightness of a ball valve, characterized in that, The method comprises the following steps: In a preset pressurization and pressure maintenance test process, a multi-channel pressure change information stream of a medium flowing through the target ball valve is synchronously collected; A pressure field dynamic evolution sequence of the target ball valve in a three-dimensional space is generated by constructing and processing the multi-channel pressure change information stream, and the pressure field dynamic evolution sequence is composed of continuous time-space data frames reflecting the transmission process of pressure at the sealing interface; The pressure field dynamic evolution sequence is cross-compared with a preset leakage signal simulation library in time and frequency domains, and a potential leakage characteristic signal is extracted; A suspected leakage source is determined by performing reverse tracing analysis on a leakage energy propagation path of the potential leakage characteristic signal, and spatial coordinates and energy intensity of the suspected leakage source are determined; Based on the spatial coordinates and energy intensity of the suspected leakage source, an overall leakage state quantitative index of the target ball valve is calculated; According to the overall leakage state quantitative index, a sealing state evaluation map and a maintenance operation sequence for the target ball valve are generated.

2. The method for detecting the gas tightness of a ball valve according to claim 1, characterized in that, The constructing and processing of the multi-channel pressure change information stream to generate the pressure field dynamic evolution sequence of the target ball valve in the three-dimensional space comprises: At multiple equally spaced time nodes of the preset pressurization and pressure maintenance test process, instantaneous pressure values from different spatial sampling points are intercepted; All instantaneous pressure values at each time node are mapped to a preset three-dimensional digital grid model of the target ball valve to generate a static pressure distribution snapshot corresponding to the time node; In time sequence, continuous multiple static pressure distribution snapshots are smoothly spliced and transition calculated to form a dynamic image showing the transmission and attenuation process of pressure between the sealing surface, the valve seat and the valve body; Pressure gradient vectorization calculation is performed on each frame of data in the dynamic image to obtain the direction and rate of the fastest pressure change in each frame; The pressure gradient vectorization calculation results of all frames are arranged in time axis to form the pressure field dynamic evolution sequence.

3. The method for detecting the gas tightness of a ball valve according to claim 2, characterized in that, The cross comparison of the pressure field dynamic evolution sequence with the preset leakage signal simulation library in time and frequency domains to extract the potential leakage characteristic signal comprises: A leakage signal simulation library containing multiple standard leakage modes is established, and each standard leakage mode corresponds to a standard pressure field dynamic evolution segment under a known leakage aperture, position and leakage rate; From the pressure field dynamic evolution sequence, a to-be-analyzed segment with the same time length as the standard pressure field dynamic evolution segment is intercepted; Convolution cross-correlation operation is performed between the to-be-analyzed segment and each standard leakage mode in the leakage signal simulation library, and a series of correlation coefficients are calculated; From the series of correlation coefficients, correlation coefficients exceeding a preset correlation threshold are screened out, and the standard leakage mode corresponding to the correlation coefficient is marked as a matching mode; According to all matching modes, signal decomposition is performed on the to-be-analyzed segment to separate out a residual fluctuation signal irrelevant to background pressure field change, and the residual fluctuation signal is the potential leakage characteristic signal.

4. The method for detecting the gas tightness of a ball valve according to claim 1, characterized in that, The reverse tracing analysis of the potential leakage characteristic signal on the leakage energy propagation path determines the spatial coordinates and energy intensity of at least one suspected leakage source, comprising: Obtaining the material acoustic parameters and structural boundary conditions of the three-dimensional entity model of the target ball valve; Taking the position and time of the occurrence of the potential leakage characteristic signal in the dynamic evolution sequence of the pressure field as the terminal point, and based on the material acoustic parameters and structural boundary conditions, establishing a wave equation inverse solving model of the leakage energy propagation in the medium; By solving the wave equation inverse solving model, the possible propagation path of the leakage energy inside the target ball valve is deduced in reverse until the energy converges to one or more starting points; Each energy convergent starting point is marked as a suspected leakage source, and its spatial coordinates in the three-dimensional digital grid model are recorded; The energy dissipated from each suspected leakage source to the signal collection terminal is calculated as the energy intensity of the suspected leakage source.

5. The method for detecting the gas tightness of a ball valve according to claim 4, characterized in that, Based on the spatial coordinates and energy intensity of the suspected leakage source, the overall leakage state quantitative index of the target ball valve is calculated, comprising: Summing up the energy intensity of all suspected leakage sources to obtain the total leakage energy estimate; Counting the number of suspected leakage sources and combining the spatial coordinates of the suspected leakage sources to calculate the distribution density of the suspected leakage sources on the key sealing surface of the target ball valve; According to the energy intensity of each suspected leakage source, the dispersion degree index of the energy intensity is calculated; The total leakage energy estimate, the distribution density and the dispersion degree index are input into a preset leakage state comprehensive calculation function, and the output of the leakage state comprehensive calculation function is the overall leakage state quantitative index of the target ball valve.

6. The method for detecting the gas tightness of a ball valve according to claim 1, characterized in that, In the preset pressurization and pressure maintenance test process, the multi-channel pressure change information stream of the medium flowing through the target ball valve is synchronously collected, comprising: Pressure sensors are arranged in at least three annular regions of the upstream cavity, the downstream cavity, and the contact between the valve seat and the ball of the target ball valve; Start the pressurizing device to inject the detection medium into the upstream cavity of the target ball valve at a constant rate, and record the readings of all pressure sensors to form the pressurization stage information stream; When the upstream cavity pressure reaches the preset test pressure, stop the pressurizing device and enter the pressure maintenance stage, and continuously record the readings of all pressure sensors within the preset pressure maintenance time period to form the pressure maintenance stage information stream; The pressurization stage information stream and the pressure maintenance stage information stream are combined in chronological order to form the multi-channel pressure change information stream.

7. The method for detecting the gas tightness of a ball valve according to claim 3, characterized in that, After the potential leakage characteristic signal is extracted, the potential leakage characteristic signal is subjected to an environmental interference stripping step: Collect the background vibration signal and temperature drift signal in the test environment; Perform coherence analysis on the background vibration signal and temperature drift signal and the potential leakage characteristic signal respectively; Subtract the signal components with high coherence with the background vibration signal and temperature drift signal from the potential leakage characteristic signal to obtain the purified leakage characteristic signal; The purified leakage characteristic signal is used to update the potential leakage characteristic signal for subsequent reverse tracing analysis of the leakage energy propagation path.

8. The method for detecting the gas tightness of a ball valve according to claim 1, characterized in that, The generating of the sealing state evaluation atlas and the maintenance operation sequence for the target ball valve comprises: Taking a three-dimensional digital grid model as a carrier, the overall leakage state quantitative indicator, the spatial coordinates and energy intensity of each suspected leakage source are visualized and rendered in different colors and brightness, and the sealing state evaluation atlas is generated; According to the spatial coordinates of the suspected leakage source, the area needing priority processing is marked on the sealing state evaluation atlas; For each area needing priority processing, the corresponding sealing surface processing process parameters and operation steps are matched from the preset operation knowledge base according to the corresponding energy intensity; All matched operation steps are sorted according to the processing priority of the corresponding area to form the maintenance operation sequence.

9. The method for detecting the gas tightness of a ball valve according to claim 8, characterized in that, After the generating of the sealing state evaluation atlas and the maintenance operation sequence for the target ball valve, a verification and feedback step is further included: After the maintenance according to the maintenance operation sequence is completed, the detection method of claim 1 is performed again on the same target ball valve to obtain a new overall leakage state quantitative indicator; The overall leakage state quantitative indicators before and after the maintenance are compared to calculate the improvement rate; If the improvement rate is lower than a preset threshold, the operation effectiveness coefficient corresponding to the process parameters used in this maintenance in the maintenance operation knowledge base is adjusted, and the detailed data of this maintenance is recorded for expanding the preset leakage signal simulation library.

10. A system for ball valve gas tightness testing, comprising: The system comprises at least one processor and at least one memory; the at least one memory stores a computer program, and when the computer program is executed by the at least one processor, the system executes the ball valve air tightness detection method according to any one of claims 1 to 9.

Images