A pipeline vibration data acquisition method for pipeline leakage

Abstract

The present application relates to the field of data acquisition optimization technology, and in particular to a pipeline vibration data acquisition method for pipeline leakage, the method comprising the following steps: collecting vibration signals and pressure signals at multiple monitoring points in the pipeline, extracting the high-frequency component ratio of the vibration signal at any monitoring point; calculating the preliminary leakage coefficient of any monitoring point at a single moment based on the change in the high-frequency component ratio and the change in the pipeline pressure at the same moment; obtaining the maximum value of the preliminary leakage coefficient; obtaining a corrected leakage coefficient based on the maximum value of the preliminary leakage coefficient and the preliminary leakage coefficient at a single moment; calculating the difference from the mean of the corrected leakage coefficients of adjacent monitoring points to obtain the leakage coefficient of any monitoring point at a single moment; and dynamically adjusting the vibration data sampling rate of the corresponding monitoring point based on the leakage coefficient of any monitoring point at a single moment. The present application can adaptively increase the vibration data sampling rate when leakage occurs, thereby reducing the probability of missed detection of pipeline leakage.

Description

Technical Field

[0001] The present application relates to the technical field of data acquisition optimization, and in particular to a pipeline vibration data acquisition method for pipeline leakage. Background Art

[0002] Pipeline leakage is a major safety hazard in industrial transportation systems. It not only causes economic losses but may also lead to safety accidents. With the development of industrialization, the scale of pipeline systems continues to expand. Pipeline leakage monitoring is a necessary means of pipeline safety management. Pipeline leakage monitoring can detect pipeline safety risks in a timely manner.

[0003] Traditional pipeline leakage monitoring methods primarily rely on collecting vibration and pressure data at a fixed frequency. However, fixed-frequency acquisition has significant drawbacks: low-frequency sampling may fail to capture transient leakage characteristics, while high-frequency sampling generates a large amount of redundant data during non-leakage periods, increasing storage and computational burdens. While existing technologies can identify some leaks, they lack the ability to adaptively adjust the sampling rate, making it difficult to balance detection accuracy and system efficiency. Summary of the Invention

[0004] In order to solve the problem in the related art that the sampling rate cannot be adaptively adjusted, the present application provides a pipeline vibration data collection method for pipeline leakage.

[0005] This application provides a pipeline vibration data collection method for pipeline leakage, which adopts the following technical solutions:

[0006] Collect vibration signals and pressure signals from multiple monitoring points on the pipeline, and extract the high-frequency component ratio of the vibration signal at any monitoring point;

[0007] A time series is constructed based on the proportion of high-frequency components at any monitoring point at multiple moments, and the first-order difference cumulative value at a single moment in the time series is calculated. Based on the negative correlation between the first-order difference cumulative value and the pipeline pressure change at the same moment, the preliminary leakage coefficient of any monitoring point at a single moment is calculated;

[0008] Obtain the maximum value of the preliminary leakage coefficient of any monitoring point at multiple consecutive moments; obtain the corrected leakage coefficient of any monitoring point at a single moment based on the maximum value of the preliminary leakage coefficient of any monitoring point and the preliminary leakage coefficient at a single moment; calculate the difference between the corrected leakage coefficient of any monitoring point at a single moment and the corrected leakage coefficient of the adjacent monitoring points to obtain the leakage coefficient of any monitoring point at a single moment;

[0009] According to the leakage coefficient of any monitoring point at a single moment, the vibration data sampling rate of the corresponding monitoring point is dynamically adjusted.

[0010] Optionally, the step of obtaining the high-frequency component ratio value includes: setting a time window at a single moment, using the time window to construct a vibration data sequence of any monitoring point; using the empirical mode decomposition algorithm to decompose the vibration data sequence of any monitoring point to obtain all modal components, and using The test divides each modal component among all modal components into high and low frequencies to obtain the number of high-frequency modal components; the ratio of the number of high-frequency modal components to the number of all modal components is calculated to obtain the high-frequency component proportion of the vibration data at any monitoring point.

[0011] Optionally, the step of obtaining a preliminary leakage coefficient includes: constructing a time series based on the high-frequency component proportion values ​​of any monitoring point at multiple moments to obtain a high-frequency component proportion value sequence of any monitoring point; calculating a first-order difference cumulative sequence of the high-frequency component proportion value sequence of any monitoring point, and obtaining a first-order difference cumulative value in the first-order difference cumulative sequence at any monitoring point at a single moment; obtaining pipeline pressure data of any monitoring point, and calculating a change in pipeline pressure of any monitoring point at a single moment; and calculating a preliminary leakage coefficient of any monitoring point at a single moment based on a negative correlation between the first-order difference cumulative value of any monitoring point at a single moment and the change in pipeline pressure at the same moment.

[0012] Optionally, the step of obtaining the leakage coefficient includes: setting a time window at a single moment, obtaining the preliminary leakage coefficient of any monitoring point at continuous moments in the time window, and constructing a preliminary leakage coefficient sequence of any monitoring point based on the preliminary leakage coefficient of any monitoring point at continuous moments in the time window; obtaining the maximum value of the preliminary leakage coefficient in the preliminary leakage coefficient sequence of any monitoring point as the maximum value of the preliminary leakage coefficient of any monitoring point; adding the preliminary leakage coefficient at a single moment and the maximum value of the preliminary leakage coefficient to obtain the corrected leakage coefficient of any monitoring point at a single moment; obtaining the average of the corrected leakage coefficients of the neighboring monitoring points of any monitoring point at a single moment; calculating the difference between the corrected leakage coefficient of any monitoring point at a single moment and the average of the corrected leakage coefficients of the neighboring monitoring points to obtain the leakage coefficient of any monitoring point at a single moment.

[0013] Optionally, the step of dynamically adjusting the vibration data sampling rate of the corresponding monitoring point includes: obtaining the leakage coefficient at any monitoring point at a single moment, multiplying the leakage coefficient by the vibration sensor sampling rate at a single moment to obtain a new vibration sensor sampling rate, and updating the new sampling rate to the vibration sensor sampling rate to achieve dynamic adjustment of the vibration data sampling rate of the corresponding monitoring point.

[0014] Optionally, the step of obtaining a first-order difference cumulative sequence includes: obtaining a high-frequency component proportion value sequence of any monitoring point at a single moment, calculating a first-order difference sequence of the high-frequency component proportion value sequence of any monitoring point at a single moment, and accumulating the first-order difference sequence of any monitoring point at a single moment in chronological order to obtain a first-order difference cumulative sequence of any monitoring point at a single moment.

[0015] Optionally, the step of obtaining the pipeline pressure change includes: obtaining the pipeline pressure data of any monitoring point, obtaining the pipeline pressure value of any monitoring point at a single moment and the pipeline pressure value at the moment before the single moment, and calculating the difference between the pipeline pressure value of any monitoring point at a single moment and the pipeline pressure value at the moment before the single moment as the pipeline pressure change of any monitoring point at a single moment.

[0016] Optionally, the step of obtaining the corrected leakage coefficient of the adjacent monitoring points includes: when installing vibration sensors at different monitoring points on the same pipeline, recording the longitude and latitude coordinate values ​​of the vibration sensor installation at each monitoring point; obtaining the longitude and latitude coordinate values ​​of the remaining monitoring points on the same pipeline to the longitude and latitude coordinate values ​​of any monitoring point, using the Haversine formula to calculate the distance from the remaining monitoring points on the same pipeline to any monitoring point, and selecting the two monitoring points closest to any monitoring point on the same pipeline as the two adjacent monitoring points of any monitoring point; obtaining the average of the corrected leakage coefficients of the two adjacent monitoring points as the average of the corrected leakage coefficients of the adjacent monitoring points.

[0017] The present application has the following technical effects: when leakage starts, the preliminary leakage coefficient is quantified by utilizing the negative correlation between the proportion of high-frequency components in the collected vibration signal and the change in pressure data, thereby enhancing the sensitivity of leakage identification; and the preliminary leakage coefficient is corrected based on the change in the preliminary leakage coefficient after the leakage occurs, thereby achieving continuous effective monitoring of leakage characteristics while combining the difference in corrected leakage coefficients with adjacent monitoring points to reduce the probability of false triggering, so that when the vibration signal sampling rate is adaptively adjusted according to the leakage coefficient, the vibration signal sampling rate in the leakage period can be increased, and the vibration signal sampling rate in the non-leakage period can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of a method for collecting pipeline vibration data for pipeline leakage in the present application. DETAILED DESCRIPTION

[0019] The present application discloses a pipeline vibration data collection method for pipeline leakage, referring to Figure 1 A pipeline vibration data collection method for pipeline leakage includes steps S1 to S4, which are specifically as follows:

[0020] S1: Collect vibration signals and pressure signals from multiple monitoring points on the pipeline, and extract the high-frequency component ratio of the vibration signal from any monitoring point.

[0021] Vibration sensors and pressure sensors are installed at multiple monitoring points on the pipeline. The sampling rates of each monitoring point are synchronized through a clock synchronization protocol. The initial sampling rate of the vibration sensor and pressure sensor is set to 10 Hz (i.e., sampling once every 0.1 second). The clock synchronization protocol is the PTP / gPTP clock synchronization protocol, which can be adjusted by the implementer according to the specific implementation scenario. Different monitoring points are distributed along the pipeline.

[0022] The current sampling rate is too low, which will result in the inability to detect pipeline leakage problems in time when a pipeline leak occurs. When the sampling rate is too high, the amount of collected data will be too large, and since the pipeline is in normal operation most of the time, it will cause problems such as redundancy of collected data.

[0023] In this embodiment, the vibration sensor is an acceleration sensor. By placing the acceleration sensor in close contact with the pipeline, vibration data of the pipeline can be collected. The vibration data is an acceleration sequence with a time sequence.

[0024] In this embodiment, the pressure sensor is a piezoelectric pipe liquid pressure sensor. By placing the pressure sensor in close contact with the pipe, pressure data in the pipe can be collected. The pressure data is a time-series pressure value sequence.

[0025] By installing a vibration sensor and a pressure sensor at the same monitoring point on the pipeline, vibration and pressure data can be obtained at that monitoring point. Similarly, vibration and pressure data can be obtained at all sensor-installed locations on the pipeline.

[0026] When a pipeline leaks, the fluid flow state at the leaking location becomes more complicated, causing the flow state at the leaking location to be inconsistent with the flow state at the non-leaking location. When the leaking location vibrates, more high-frequency signals will be generated. At the same time, the structure of the pipeline itself changes, causing the pressure value in the pipeline to drop.

[0027] Get the first The monitoring point is Vibration data at all times , in Set the time window at the moment and get the The first time window in the past The continuous vibration data of the monitoring points are obtained Vibration sequence of monitoring points The time window value is 3 seconds, which can be adjusted by the implementer according to the specific implementation scenario.

[0028] right The empirical mode decomposition (EMD) algorithm is used to decompose the data and obtain multiple modal components. The test divides each modal component into a high-frequency modal component or a low-frequency modal component. Inspection is a known technical content.

[0029] The number of high-frequency modal components and the number of low-frequency modal components obtained by the division are statistically calculated, and the total number of all modal components is the sum of the number of high-frequency modal components and the number of low-frequency modal components obtained by the division.

[0030] Taking the number of high-frequency modal components as the numerator and the number of all modal components as the denominator, calculate the ratio of the number of high-frequency modal components to the number of all modal components to obtain the first The proportion of high-frequency components of vibration data at each monitoring point , The larger the value, the The higher the proportion of high-frequency components in the vibration data at each monitoring point, the higher the probability of leakage. The proportion of high-frequency components in the vibration data at each monitoring point will continue to rise.

[0031] S2: Construct a time series based on the proportion of high-frequency components of any monitoring point at multiple moments, calculate the first-order difference cumulative value at a single moment in the time series, and calculate the preliminary leakage coefficient of any monitoring point at a single moment based on the negative correlation between the first-order difference cumulative value and the pipeline pressure change at the same moment.

[0032] In the event of leakage, the The proportion of high-frequency components in the vibration data at each monitoring point will continue to rise, so the The time window at different times The high-frequency component ratio of each monitoring point is obtained The high-frequency component ratio of each monitoring point at consecutive moments is used to construct a time series of high-frequency component ratios, and the high-frequency component ratio ratio sequence is obtained. .

[0033] calculate The first-order difference sequence value of , and Accumulate in time sequence to obtain the first-order difference accumulation sequence The larger the cumulative value in the first-order difference cumulative sequence, the more obvious the change in the proportion of high-frequency components. Middle The accumulated value of the first-order difference at time , among which, The time is the current time.

[0034] Get the Moment The pressure data value at each monitoring point Hedi Moment The pressure data value at each monitoring point .

[0035] Calculate the Moment Preliminary leakage coefficient at each monitoring point :

[0036]

[0037] Where, is an exponential function.

[0038] For the Moment The first-order difference accumulation value of the vibration data at each monitoring point, The larger the value, the more obvious the change in the proportion of high-frequency components, and the greater the probability of leakage.

[0039] For the Moment The pressure data value at each monitoring point, For the Moment The pressure data value at each monitoring point, Indicates the Moment The change in pipeline pressure at each monitoring point. When leakage occurs, the pressure in the pipeline will drop, thus The value will be less than 0, indicating that the pressure in the pipeline is decreasing.

[0040] So yes Using exponential function Perform negative correlation mapping and get ,when When it becomes smaller, When the value of increases and the pressure in the pipeline decreases, it may be that the pipeline pressure drops when the leakage just occurs.

[0041] The larger the value, the Moment The larger the initial leakage coefficient at a monitoring point, the greater the probability that leakage has just occurred.

[0042] S3: Obtain the maximum value of the preliminary leakage coefficient of any monitoring point at multiple consecutive moments; obtain the corrected leakage coefficient of any monitoring point at a single moment based on the maximum value of the preliminary leakage coefficient of any monitoring point and the preliminary leakage coefficient at a single moment; calculate the difference between the corrected leakage coefficient of any monitoring point at a single moment and the corrected leakage coefficient of the adjacent monitoring points to obtain the leakage coefficient of any monitoring point at a single moment.

[0043] After leakage occurs, the pressure value will gradually stabilize again and the high-frequency signal will also gradually stabilize, resulting in a decrease in the initial leakage coefficient.

[0044] Get the The monitoring point is The initial leakage coefficient at each moment in the time window of the moment constitutes the sequence , get The maximum value among The maximum value of the initial leakage coefficient of a monitoring point at multiple consecutive moments in the time window indicates the The maximum probability of leakage occurring at each monitoring point within the time window.

[0045] As the pressure value gradually stabilizes again and the high-frequency signal gradually stabilizes, the initial leakage coefficient decreases, making it impossible to continuously monitor the pipeline leakage characteristics. When the sampling rate is too low, leakage detection is likely to occur. When the sampling rate is too high, data redundancy problems will occur. When there is no leakage risk, the sampling rate should be lowered to reduce data redundancy. When there is a leakage risk, the sampling rate should be increased to reduce the probability of leakage detection.

[0046] The first Moment The preliminary leakage coefficient at each monitoring point is The maximum values ​​of the preliminary leakage coefficients of the monitoring points are accumulated to obtain the corrected leakage coefficient.

[0047] When leakage occurs, it is usually in a local part of the pipeline and the distance between monitoring points is large, which leads to the The leakage at the monitoring point will have little impact on the vibration and pressure data of the adjacent monitoring points. Therefore, the difference between the mean value of the corrected leakage coefficient of the monitoring point and the adjacent monitoring points is calculated to obtain the value of the first Moment Leakage coefficient of each monitoring point.

[0048] Among them, The process of obtaining the adjacent monitoring points of each monitoring point is as follows: when installing vibration sensors at different monitoring points on the same pipeline, record the longitude and latitude coordinates of the vibration sensor installation at each monitoring point, and obtain the longitude and latitude coordinates of the remaining monitoring points on the same pipeline to the first monitoring point. The latitude and longitude coordinates of the first monitoring point are calculated using the Haversine formula to calculate the distance from the remaining monitoring points on the same pipeline to the The distance between the first and second monitoring points is selected on the same pipeline. The two nearest monitoring points are used as the first The Haversine formula is a general formula for calculating the great circle distance between two points on the earth's surface and is well known.

[0049] According to Moment The leakage coefficient of each monitoring point is Moment The sampling rate of the vibration sensor of each monitoring point is adjusted to obtain the adjusted sampling rate of the vibration sensor.

[0050] No. Moment The leakage coefficient at each monitoring point is :

[0051]

[0052] Where, is an exponential function, Indicates taking the maximum value, The adjustment coefficient is a hyperparameter. Take the experience value 0.86, The value can be adjusted by the implementer according to the specific implementation scenario.

[0053] For the The monitoring point is at The initial leakage coefficient at time The larger the value, the The monitoring point is at The greater the probability of leakage occurring at any time.

[0054] For the The preliminary leakage coefficient sequence of each monitoring point within the time window, For the The maximum value of the initial leakage coefficient of each monitoring point within the time window, The larger the The greater the probability that leakage has just occurred at a monitoring point within the time window.

[0055] As the pressure value gradually stabilizes and the high-frequency signal gradually stabilizes, the initial leakage coefficient decreases, which affects the subsequent continuous monitoring of leakage data. and Add together and get , as the first Moment The corrected leakage coefficient at each monitoring point is used to correct the preliminary leakage coefficient at subsequent moments.

[0056] For all adjacent monitoring points The mean value of the corrected leakage coefficient at the moment, and then by calculating , as the Moment The difference between the mean value of the corrected leakage coefficient of the monitoring point and all adjacent monitoring points can be used to represent the current The relative significance of the leakage characteristics of a monitoring point compared to the adjacent monitoring points, so The larger the value is, the greater the probability of leakage is. Moment The higher the vibration sensor sampling rate of each monitoring point, the higher the sampling rate should be.

[0057] when When it is greater than 0, it means that the leakage feature is relatively significant. When it is less than or equal to 0, it means that the leakage is relatively low, and then the exp exponential function is used to Mapping is performed. When the leakage characteristics are obvious, When it is greater than 1, the relative significance of the leakage feature is low. Less than or equal to 1, used for vibration sensor sampling rate adjustment.

[0058] S4. Dynamically adjust the vibration data sampling rate of any monitoring point according to the leakage coefficient of any monitoring point at a single moment.

[0059] Get the Moment Leakage coefficient at each monitoring point , the leakage coefficient With the The vibration sensor sampling rate is multiplied at every moment to obtain a new vibration sensor sampling rate, and the new sampling rate is updated as the vibration sensor sampling rate to optimize the collection of pipeline leakage vibration data.

[0060] The above are all preferred embodiments of the present application, and are not intended to limit the scope of protection of the present application. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A pipeline vibration data acquisition method for pipeline leakage, characterized in that: Including steps: Collect vibration signals and pressure signals from multiple monitoring points on the pipeline, and extract the high-frequency component ratio of the vibration signal at any monitoring point; A time series is constructed based on the proportion of high-frequency components at any monitoring point at multiple moments, and the first-order difference cumulative value at a single moment in the time series is calculated. Based on the negative correlation between the first-order difference cumulative value and the pipeline pressure change at the same moment, the preliminary leakage coefficient of any monitoring point at a single moment is calculated; Obtain the maximum value of the preliminary leakage coefficient of any monitoring point at multiple consecutive moments; obtain the corrected leakage coefficient by summing the preliminary leakage coefficient at any monitoring point at a single moment with the maximum value of the preliminary leakage coefficient at any monitoring point; calculate the difference between the corrected leakage coefficient of any monitoring point at a single moment and the corrected leakage coefficient of the adjacent monitoring points to obtain the leakage coefficient of any monitoring point at a single moment; According to the leakage coefficient of any monitoring point at a single moment, the vibration data sampling rate of the corresponding monitoring point is dynamically adjusted.

2. The pipeline vibration data acquisition method for pipeline leakage according to claim 1, characterized in that: The steps for obtaining the high-frequency component ratio include: Set a time window at a single moment and use the time window to construct a vibration data sequence for any monitoring point; The vibration data sequence of any monitoring point is decomposed using the empirical mode decomposition algorithm to obtain all modal components. The test divides each modal component in all modal components into high and low frequencies to obtain the number of high-frequency modal components; The ratio of the number of high-frequency modal components to the number of all modal components is calculated to obtain the proportion of high-frequency components in the vibration data at any monitoring point.

3. The pipeline vibration data acquisition method for pipeline leakage according to claim 1, characterized in that: The steps to obtain a preliminary leakage coefficient include: Based on the high-frequency component ratio values ​​of any monitoring point at multiple moments, a time series is constructed to obtain a high-frequency component ratio value sequence of any monitoring point; Calculate the first-order difference cumulative sequence of the high-frequency component ratio value sequence of any monitoring point, and obtain the first-order difference cumulative value in the first-order difference cumulative sequence of any monitoring point at a single moment; Obtain pipeline pressure data at any monitoring point and calculate the pipeline pressure change at any monitoring point at a single moment; According to the negative correlation between the first-order difference cumulative value of any monitoring point at a single moment and the pipeline pressure change at the same moment, the preliminary leakage coefficient of any monitoring point at a single moment is calculated.

4. The pipeline vibration data acquisition method for pipeline leakage according to claim 1, characterized in that: The steps to obtain the leakage coefficient include: A time window is set at a single moment, and a preliminary leakage coefficient of any monitoring point at consecutive moments within the time window is obtained. Based on the preliminary leakage coefficient of any monitoring point at consecutive moments within the time window, a preliminary leakage coefficient sequence of any monitoring point is constructed; Obtain the maximum value of the preliminary leakage coefficient in the preliminary leakage coefficient sequence of any monitoring point as the maximum value of the preliminary leakage coefficient of any monitoring point; Add the preliminary leakage coefficient at a single moment and the maximum preliminary leakage coefficient to obtain the corrected leakage coefficient of any monitoring point at a single moment; Obtain the average value of the corrected leakage coefficient of the neighboring monitoring points of any monitoring point at a single moment; Calculate the difference between the corrected leakage coefficient of any monitoring point at a single moment and the average of the corrected leakage coefficients of adjacent monitoring points to obtain the leakage coefficient of any monitoring point at a single moment.

5. The pipeline vibration data acquisition method for pipeline leakage according to claim 1, characterized in that: The steps of dynamically adjusting the vibration data sampling rate of the corresponding monitoring point include: obtaining the leakage coefficient at any monitoring point at a single moment, multiplying the leakage coefficient by the vibration sensor sampling rate at a single moment to obtain a new vibration sensor sampling rate, and updating the new sampling rate to the vibration sensor sampling rate to achieve dynamic adjustment of the vibration data sampling rate of the corresponding monitoring point.

6. The pipeline vibration data collection method for pipeline leakage according to claim 3, characterized in that: The steps of obtaining the first-order difference cumulative sequence include: obtaining a high-frequency component proportion value sequence of any monitoring point at a single moment, calculating a first-order difference sequence of the high-frequency component proportion value sequence of any monitoring point at a single moment, and accumulating the first-order difference sequence of any monitoring point at a single moment in chronological order to obtain a first-order difference cumulative sequence of any monitoring point at a single moment.

7. The pipeline vibration data collection method for pipeline leakage according to claim 3, characterized in that: The step of obtaining the pipeline pressure change includes: obtaining the pipeline pressure data of any monitoring point, obtaining the pipeline pressure value of any monitoring point at a single moment and the pipeline pressure value at the moment before the single moment, and calculating the difference between the pipeline pressure value of any monitoring point at a single moment and the pipeline pressure value at the moment before the single moment as the pipeline pressure change of any monitoring point at a single moment.

8. The pipeline vibration data collection method for pipeline leakage according to claim 4, characterized in that: The steps for obtaining the average value of the modified leakage coefficient of adjacent monitoring points include: When installing vibration sensors at different monitoring points on the same pipeline, record the longitude and latitude coordinates of the vibration sensor installation at each monitoring point; Obtain the latitude and longitude coordinates of the remaining monitoring points on the same pipeline to any monitoring point, calculate the distance from the remaining monitoring points to any monitoring point using the Haversine formula, and select the two monitoring points closest to any monitoring point on the same pipeline as the two neighboring monitoring points of any monitoring point; The average of the corrected leakage coefficients of the two adjacent monitoring points is obtained as the average of the corrected leakage coefficients of the adjacent monitoring points.

Images