Sensor-based pipeline defect location method and system

Abstract

Provided are a sensor-based pipeline defect location method and a system, which accurately locate the location of a pipeline defect point by means of multi-parameter fusion, related to the technical field of oil and gas pipeline maintenance. The method comprises: after a magnetic flux leakage detection sensor finishes performing magnetic flux leakage detection on a pipeline, inputting original data of a three-way odometer of the magnetic flux leakage detection sensor into an inertial measurement unit sensor, and acquiring first location data output by the inertial measurement unit sensor by means of performing data calculation on the original data; acquiring second location data recorded by at least one ground mark location sensor embedded in the running direction of the pipeline; fusing the first location data and the second location data by using time synchronization processing and spatial inference location processing, to obtain fused location data; determining defect data of the pipeline on the basis of content visually displayed by internal detection data analysis software for the original data; and fitting and comparing the defect data and the fused location data, to obtain the longitude and latitude of each defect point.

Description

Sensor-based pipeline defect positioning method and system

[0001] The present disclosure claims priority to the patent application with the application number 202410961933.0, which was filed with the China Patent Office on July 18, 2024, and the entire content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present disclosure relates to the technical field of oil and gas pipeline maintenance, and in particular to a sensor-based pipeline defect positioning method and system. BACKGROUND

[0003] Oil and gas pipeline transportation plays a crucial role in the energy field. Pipeline transportation is an efficient, safe, and economical way that plays an indispensable role in meeting global energy demand. Since oil and gas are the main sources of energy for modern industrial society, the smooth operation of the pipeline transportation system is crucial to maintaining social stability and economic development. If there is a defect in the oil and gas pipeline and a leak occurs, it will have serious impact and consequences.

[0004] Therefore, it is particularly important to prevent and respond to oil and gas pipeline leaks. However, in terms of preventing pipeline leaks, the most critical thing is to accurately locate the position of the pipeline defect point and repair the defect. However, it is not easy to determine the location of the defect point on a pipeline that is several kilometers long or even under the sea. The conventional excavation verification method is not only impractical, but also consumes a lot of manpower, material resources, and time.

[0005] Therefore, it is of great significance to accurately locate the position of the pipeline defect point in order to promptly carry out repair work. SUMMARY

[0006] The present disclosure provides a sensor-based pipeline defect positioning method and system, which can accurately locate the position of the pipeline defect point.

[0007] In a first aspect, a sensor-based pipeline defect positioning method is provided, comprising:

[0008] After the magnetic flux leakage detection sensor finishes the magnetic flux leakage detection on the pipeline, inputting the original data of the three-way odometer of the magnetic flux leakage detection sensor to the inertial measurement unit sensor;

[0009] Using the inertial measurement unit sensor to perform data calculation on the original data to obtain first positioning data, the first positioning data comprising at least one first time point, and a longitude and a latitude corresponding to each first time point;

[0010] The second positioning data recorded by at least one ground marker positioning sensor is acquired after the pipeline is detected by the magnetic flux leakage detection sensor, wherein the at least one ground marker positioning sensor is buried along the pipeline direction of the pipeline, and the second positioning data includes at least one second time point and the longitude and latitude corresponding to each second time point, and the second time point is the time point when the magnetic flux leakage detection sensor passes through the ground marker positioning sensor when the magnetic flux leakage detection sensor runs in the pipeline.

[0011] The first positioning data and the second positioning data are fused by using time synchronization processing and space inference positioning processing to obtain fused positioning data.

[0012] The original data is analyzed and visually displayed by using the internal detection data analysis software.

[0013] The defect data of the pipeline is determined according to the visual display content of the internal detection data analysis software, and the defect data includes detection time information and mileage information corresponding to each defect point.

[0014] The longitude and latitude of each defect point are obtained by fitting and comparing the defect data and the fused positioning data.

[0015] In a second aspect, a pipeline defect positioning system based on a sensor is provided, including:

[0016] The basic information import module is configured to input the original data of the three-way mileage wheel of the magnetic flux leakage detection sensor to the inertial measurement unit sensor after the pipeline is detected by the magnetic flux leakage detection sensor.

[0017] The basic information import module is configured to use the inertial measurement unit sensor to perform data calculation on the original data to obtain first positioning data, and the first positioning data includes at least one first time point and the longitude and latitude corresponding to each first time point.

[0018] The basic information import module is configured to acquire second positioning data recorded by at least one ground marker positioning sensor after the pipeline is detected by the magnetic flux leakage detection sensor, wherein the at least one ground marker positioning sensor is buried along the pipeline direction of the pipeline, and the second positioning data includes at least one second time point and the longitude and latitude corresponding to each second time point, and the second time point is the time point when the magnetic flux leakage detection sensor passes through the ground marker positioning sensor when the magnetic flux leakage detection sensor runs in the pipeline.

[0019] The first fusion algorithm module is configured to fuse the first positioning data and the second positioning data by using time synchronization processing and space inference positioning processing to obtain fused positioning data.

[0020] a display module configured to analyze and visually display the raw data using internal detection data analysis software;

[0021] a defect data determination module configured to determine defect data of the pipeline according to the visual display of the internal detection data analysis software, the defect data including detection time information and mileage information corresponding to each defect point;

[0022] a second fusion algorithm module configured to fit and compare the defect data and the fusion positioning data to obtain the longitude and latitude of each defect point.

[0023] The raw data output by the three-way mileage wheel of the magnetic flux leakage detection sensor is greatly affected by the pipeline environment, which affects the accuracy of the first positioning data output by the inertial measurement unit sensor. Therefore, the accuracy of determining the position of the defect point according to the first positioning data is low. In order to improve the accuracy of positioning the defects of the pipeline, the second positioning data recorded by the ground marker positioning sensor is obtained, time synchronization processing and space inference positioning processing are adopted, the first positioning data is fused and corrected using the second positioning data, the accuracy of the first positioning data is improved through multi-parameter fusion from each sensor, and the fused and corrected first positioning data is the fusion positioning data. Therefore, fitting and comparing the defect data and the fusion positioning data can obtain the accurate longitude and latitude of each defect point, so as to timely carry out repair work. BRIEF DESCRIPTION OF DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the present disclosure, the drawings needed in the embodiments will be briefly introduced below. Obviously, for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0025] FIG. 1 is a schematic flowchart of a sensor-based pipeline defect positioning method according to an example embodiment of the present disclosure;

[0026] FIG. 2 is a schematic diagram of a sensor-based pipeline defect positioning system according to an example embodiment of the present disclosure. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present disclosure.

[0028] Currently, in the pipeline defect positioning, the traditional detection technology (such as ultrasonic, magnetic flaw detection, etc.) is usually used, and the traditional method usually needs manual operation, which is time-consuming and labor-consuming, and the positioning accuracy is greatly affected by human factors. In order to accurately locate the position of the pipeline defect point, the present disclosure adopts a multi-parameter fusion method to realize accurate positioning of the pipeline defect, solving the problems of resource waste and time delay caused by using traditional detection technology for pipeline maintenance.

[0029] Fig. 1 is a schematic flow chart of a sensor-based pipeline defect positioning method provided by an example embodiment of the present disclosure, as shown in Fig. 1, the method comprises:

[0030] S110, after the magnetic flux leakage detection sensor finishes the magnetic flux leakage detection on the pipeline, inputting the original data of the three-way odometer of the magnetic flux leakage detection sensor to the inertial measurement unit sensor.

[0031] When the magnetic flux leakage detection sensor detects the pipeline, the odometer on the magnetic flux leakage detection sensor will output 4096 pulses for each revolution, and will transmit 256 pulses of the 4096 pulses to the inertial measurement unit sensor in the form of digital signals through the 422 serial port.

[0032] S120, using the inertial measurement unit sensor to solve the original data to obtain first positioning data.

[0033] The first positioning data comprises at least one first time point, and the longitude and latitude corresponding to each first time point.

[0034] The inertial measurement unit sensor adopts inertial measurement unit (IMU) technology, which can fuse the three-way odometer information through a solving algorithm to output one-way positioning data (i.e. first positioning data).

[0035] For example, the first positioning data also contains elevation information, distance information from the initial point and timestamp information corresponding to each first time point.

[0036] It should be noted that the model of the magnetic flux leakage detection sensor is not limited by the present disclosure.

[0037] Each first time point is the time point when the IMU collects data. For example, the IMU collects data every 6us, and the first time point corresponding to the first data collection is 0.0000058301 seconds. The longitude and latitude corresponding to each first time point are the longitude and latitude of the position of the magnetic flux leakage detection sensor at the first time point.

[0038] S130, after the pipeline is detected by the magnetic flux leakage detection sensor, obtain the second positioning data recorded by the at least one ground marker positioning sensor.

[0039] The at least one ground marker positioning sensor is buried along the pipeline direction of the pipeline, and the second positioning data includes at least one second time point, and the longitude and latitude corresponding to each second time point. The second time point is the time point when the magnetic flux leakage detection sensor passes through the ground marker positioning sensor when the magnetic flux leakage detection sensor is running in the pipeline.

[0040] Exemplarily, the second positioning data is read and exported by the upper computer software, wherein the upper computer software is the ground marker positioning sensor upper computer software.

[0041] Before the pipeline is detected by the magnetic flux leakage detection sensor, a plurality of ground marker positioning sensors need to be buried along the pipeline direction of the pipeline, for example, one ground marker positioning sensor is placed every kilometer along the pipeline laying direction.

[0042] S140, the first positioning data and the second positioning data are fused by using time synchronization processing and space inference positioning processing to obtain fused positioning data.

[0043] In a feasible design, the first positioning data and the second positioning data are fused by using time synchronization processing and space inference positioning processing to obtain fused positioning data in the following way:

[0044] S141, input the first positioning data and the second positioning data to the first neural network model.

[0045] The network architecture of the first neural network model can be selected according to actual needs, and the present disclosure does not limit it. For example, the first neural network model can adopt a long short-term memory (LSTM) architecture.

[0046] S142, using the first neural network model, the first positioning data and the second positioning data are time-synchronized by recognizing the time features in the first positioning data and the second positioning data, so as to align the time points in the second positioning data with the time points in the first positioning data.

[0047] After the first positioning data and the second positioning data are input into the first neural network model, the first neural network model can recognize the time features in the first positioning data and the second positioning data by learning the time information in the data. Then, the first neural network model performs time synchronization processing on the first positioning data and the second positioning data according to the time features in the first positioning data and the second positioning data, so as to ensure that the two groups of data correspond at the same time point.

[0048] For example, the ground marker positioning sensor is arranged one per kilometer on the pipeline, and each point is one kilometer apart. Assuming that the first time point output by the first ground marker positioning sensor is January 24, 2024 12:50:07.195. However, the IMU collects data every 6us, and the first time point output is 0.0000058301 seconds. It can be seen that the two sets of data are organized according to their respective time sequences, and the corresponding relationship between the time points in the two sets of data cannot be seen, that is, the two sets of data are not synchronized. After synchronization, the corresponding time points in the two sets of data are corresponding.

[0049] The corresponding same time points of the two sets of data after synchronization can be understood by referring to the following time point corresponding formula (i.e. formula (1)): t sync = arg min t ||t-t i ||, formula (1);

[0050] Wherein, t is the second time point in the second positioning data, t sync is the corresponding same time point of the first positioning data and the second positioning data after time synchronization processing, t i is the i-th first time point in the first positioning data, arg min t represents the parameter that makes the function value minimum, and ||| represents the norm calculation, which is used to calculate the absolute value.

[0051] It can be seen that formula (1) is a mathematical implementation for calculating the corresponding same time points of the two sets of data, which helps to understand the concept of the same time point. The meaning expressed by formula (1) is that after determining the second time point in the second positioning data with the minimum time difference t i , t sync is determined as t i and the time point corresponding to the second time point with the minimum time difference t i . t sync may be t i , that is, each second time point in the second positioning data is replaced by the first time point with the minimum time difference, so as to realize the time synchronization processing of the two sets of data.

[0052] The above example is based on the following consideration: taking the time information in the first positioning data as the standard, the time information of the second positioning data is re-determined.

[0053] (1) The second positioning data needs to be used for interpolation processing of the first positioning data in the future. Taking the time information in the first positioning data as the standard, the time information of the second positioning data is re-determined to prepare for subsequent interpolation processing;

[0054] (2) Compared with the first positioning data, the second positioning data includes several discontinuous second time points, and the interval between two adjacent second time points is too long, making it unsuitable as a reference for time synchronization processing.

[0055] It should be understood that, given that the first positioning data includes elevation information corresponding to each first time point and mileage information from the initial point, the fused positioning data will also include elevation information corresponding to each time point and mileage information from the initial point.

[0056] S142 uses a second neural network model to predict longitude and latitude functions based on the first positioning data.

[0057] Among them, the longitude function is used to represent the correspondence between time point series and longitude, and the latitude function is used to represent the correspondence between time point series and latitude.

[0058] The network architecture used in the second neural network model can be selected according to actual needs, and this disclosure does not impose any restrictions on it. For example, the second neural network model can adopt a long short-term memory network architecture.

[0059] S143 uses a second neural network model to correct the first positioning data based on longitude and latitude functions.

[0060] For a certain first time point t in the first positioning data i During the time period [t] i-1 , t i Within this timeframe, a second neural network model can be used to predict the latitude and longitude of that period. This second neural network model, through learned patterns and utilizing historical data and contextual information, infers longitude and latitude on a time series basis, thereby predicting the latitude (lon). i and latitude lat i The estimated value of longitude is given by the second neural network model. The estimated longitude is shown in the following formula (2): lon i =f NN (IMU data , t i ), formula (2);

[0061] The second neural network model predicts the latitude estimate as shown in the following formula (3): lat i =g NN (IMU data , t i ), formula (3);

[0062] Among them, f NN With g NN These represent the longitude and latitude functions calculated by the second neural network model, respectively. (IMU) data This indicates the first location data.

[0063] In this example, the second neural network model re-predicts the longitude of each first time point t i corresponding to the longitude, re-predicts the latitude of each first time point t i corresponding to the latitude, and then updates the longitude and latitude of the first positioning data at each first time point, to achieve correction of the first positioning data.

[0064] S144, linearly interpolates the corrected first positioning data using the second positioning data to obtain fused positioning data.

[0065] For example, the longitude of the second positioning data is used to linearly interpolate the longitude of the corrected first positioning data, as shown in the following formula (4): lon interp = loni IMU + (loni MK -loni IMU ) × (t i -t i-1 ) / (t i+1 -t i-1 ), formula (4);

[0066] The latitude of the second positioning data is used to linearly interpolate the latitude of the corrected first positioning data, as shown in the following formula (5): lat interp = lati IMU + (lati MK -lati IMU ) × (t i -t i-1 ) / (t i+1 -t i-1 ), formula (5);

[0067] wherein loni IMU is the longitude estimate of the corrected first positioning data at the first time point t i , lati IMU is the latitude estimate of the corrected first positioning data at the first time point t i , loni MK is the actual recorded longitude value of the second positioning data at the second time point t i after time synchronization processing, lati MK is the actual recorded latitude value of the second positioning data at the second time point t i after time synchronization processing, lon interp is the interpolated longitude value at the first time point t i in the first positioning data, and lat interp is the interpolated latitude value at the first time point ti the latitude value after interpolation.

[0068] As the detected pipeline is used for a long time, a large amount of dirt and impurities such as sediments, rust, silt and the like are accumulated inside. With the passage of time, these dirt continuously accumulates to form a solid fouling layer, which reduces the flow efficiency of the pipeline. Influenced by the environment inside the pipeline, the odometer on the magnetic flux leakage detection sensor running in the pipeline is prone to skidding. The accuracy of the IMU output data is largely dependent on whether the odometer is running normally. Therefore, only relying on the IMU data to locate the actual position of the defect on the pipeline may lead to inaccurate positioning, and thus the data recorded by the ground marker positioning sensor needs to be fused and corrected.

[0069] In the above example, the first neural network model can accurately perform time synchronization processing on the first positioning data and the second positioning data, ensuring that the two sets of data correspond at the same time point, so as to facilitate subsequent interpolation processing of the first positioning data using the second positioning data, and realize fusion correction of the first positioning data. Since the first positioning data output by the IMU is scattered data points, and the accuracy is greatly affected by the environment inside the pipeline, the present disclosure uses the second neural network model to predict the longitude function and the latitude function, and to re-estimate the longitude and latitude corresponding to each first time point, so as to realize correction of the first positioning data and further improve the accuracy of the first positioning data. On this basis, linear interpolation of the corrected first positioning data using the second positioning data can obtain accurate fusion positioning data.

[0070] In a feasible design, before the first positioning data and the second positioning data are fused by using time synchronization processing and spatial inference positioning processing, the first positioning data and / or the second positioning data are subjected to first preprocessing, and the first preprocessing includes one or more of the following modes: data cleaning processing, time calibration synchronization processing and format conversion processing, so as to ensure the accuracy and consistency of the data.

[0071] Among them, the data cleaning processing is to remove redundant data. The first positioning data output by the IMU corresponds to a set of time data every 6us, and a set of time data includes mileage data, longitude data, latitude data and elevation data. Since the odometer may have uncertain conditions on the pipeline, such as skidding, in this case the data recorded by the odometer is invalid. In addition, there may be a situation that several groups of data are the same, so except for the first group of data, the following several groups of data are redundant and need to be eliminated. In addition, a ground marker positioning sensor may record several data at the same time, and the most accurate data needs to be selected to eliminate useless data.

[0072] The format conversion processing refers to converting the unit of the second time point in the second positioning data into the time length of the distance from the magnetic flux leakage detection sensor inlet pipe, which is in seconds, in order to make the unit of the second time point in the second positioning data the same as the unit of the first time point in the first positioning data. For example, the magnetic flux leakage detection sensor starts detecting at 12:50, and the second time point of the output of the first ground marker positioning sensor is 12:52. That is, the magnetic flux leakage detection sensor experiences 2 minutes through the first ground marker positioning sensor, and changes the second time point recorded by the first ground marker positioning sensor, "12:52", to 120 seconds.

[0073] The time calibration synchronization processing includes performing accurate sorting and deduplication operations on the timestamp data of the first positioning data and the timestamp data of the second positioning data, and a timestamp correction operation.

[0074] In order to guarantee the timing and accuracy of the first positioning data and the second positioning data, accurate sorting and deduplication operations are performed on the timestamp data of the two groups of data, which is the basis for eliminating potential data conflicts and data duplication.

[0075] The timestamp correction operation is to identify and correct possible time deviations, which may be caused by factors such as clock drift or time delay caused by signal transmission delay. The implementation of the timestamp correction operation ensures that the time information of different data sources can run synchronously under the same time reference.

[0076] The timestamp correction operation includes:

[0077] (1) The timestamp of the first positioning data and the second positioning data is checked to verify the consistency and integrity of the data.

[0078] (2) Check whether the time interval between the timestamps conforms to the expected time sequence rule, and exclude abnormal data or incomplete data.

[0079] (3) Perform timestamp consistency detection to ensure that the time information provided by the IMU and the ground marker positioning sensor does not occur misalignment or incorrect alignment in the matching process.

[0080] The purpose of the time calibration synchronization processing is to ensure that the time sequence of the two groups of data remains consistent, and to prepare for subsequent time synchronization processing of the two groups of data using the first neural network model.

[0081] Exemplarily, the coordinate system conversion is performed on the first positioning data to eliminate data noise and errors.

[0082] Exemplarily, the second positioning data is standardized in format for subsequent calculation and analysis.

[0083] The above example effectively ensures that the time information from different data sources can be correctly matched and aligned through the first preprocessing, thereby enabling subsequent accurate data fusion and extraction of pipeline trend information.

[0084] In a feasible design, the methods include:

[0085] The longitude and latitude in the fused positioning data are used as one observation input to the first Kalman filter, and the corrected first positioning data is used as another observation input to the first Kalman filter.

[0086] The first Kalman filter is used to fit the fused positioning data and the first positioning data to obtain the fitted fused positioning data.

[0087] The fused localization data obtained through linear interpolation is used as the first observation in the Kalman filter, and the corrected first localization data is used as the second observation. The constructed observation matrix is, for example, H = [first observation; second observation], where H represents the observation matrix. The first observation is used to update the state estimate of the Kalman filter system. The output of the Kalman filter is the complete data after comparing the two observations and removing redundant data.

[0088] The above example considers the correlation between two observations and optimizes the state estimation of the Kalman filter system. By fitting the fused positioning data and the first positioning data, a complete set of fused positioning data is formed, thereby improving the accuracy of the fused positioning data.

[0089] The S150 uses internal detection data analysis software to analyze and visualize the raw data.

[0090] The internal detection data analysis software is used to parse the signals (i.e., raw data) of various types of magnetic flux leakage detection sensors, making it convenient for data readers to view the raw data and mark pipeline characteristic signals such as defect signals.

[0091] In a feasible design, before using internal detection data analysis software to analyze and visualize the raw data, the methods include:

[0092] The internal detection data analysis software employs a synchronization mechanism for the mileage wheel to synchronize the three mileage wheels.

[0093] For example, the synchronization mechanism for the application mileage wheel is implemented to synchronize the three mileage wheels in the following way:

[0094] Every 5 seconds, the distance traveled by each of the three mileage wheels is calculated, and the data of the mileage wheel with the highest distance traveled is taken as the data of the other two mileage wheels.

[0095] In order to reduce the impact of the odometer jam on the detection as much as possible, three odometers are usually deployed for each detector when the magnetic flux leakage detection sensor detects in the pipeline. Therefore, it is necessary to synchronize the three odometers on the internal detection data software, in order to ensure the consistency and accuracy of the data.

[0096] In the above example, the original data generated during the operation of the magnetic flux leakage detection sensor can be obtained by time sampling. According to the odometer synchronization processing mechanism, the data of the three odometers in the original data is synchronized for subsequent data analysis and interpretation.

[0097] Exemplarily, after the synchronized original data is differentially analyzed, it is spread in the internal detection data analysis software for further data analysis and visual display. Through the spread data, the user can intuitively view the trend of the odometer data in time and space, helping to identify possible problems or abnormal situations in the pipeline.

[0098] Among them, differential analysis refers to converting the time-acquired mileage information into mileage-ordered information. Spreading in the software is to perform mileage concatenation, which combines multiple scattered files into a continuous file according to the mileage information for data reading.

[0099] S160, according to the content of the visual display of the internal detection data analysis software, determines the defect data of the pipeline.

[0100] Among them, the defect data includes detection time information and mileage information corresponding to each defect point. The detection time information includes time stamp and other time information.

[0101] Specifically, after the personnel mark the defect signal on the internal detection data analysis software, the detection time information and mileage information corresponding to the defect point at the marked position are exported from the software, thereby obtaining the defect data.

[0102] S170, fitting and comparing the defect data and the fusion positioning data to obtain the longitude and latitude of each defect point.

[0103] In a feasible design, before aligning the defect data and the fusion positioning data by using the time stamps of the defect data and the fusion positioning data, the method comprises:

[0104] The second preprocessing of the defect data comprises one or more of the following ways:

[0105] Format conversion processing, standardization and normalization processing.

[0106] Among them, the format conversion processing is used to convert the unit of the time information of the defect data into the same unit as the unit of the time information of the fusion positioning data.

[0107] For example, the standardization and normalization process is performed using the following formula (6):

[0108] Where x is the original defect data, σ is the average value of the defect data, σ is the standard deviation of the defect data, and x' is the defect data after standardization and normalization.

[0109] The above example demonstrates how standardizing and normalizing data can eliminate dimensional differences between different features.

[0110] In a feasible design, the longitude and latitude of each defect point are obtained by fitting and comparing the defect data and the fused location data in the following way:

[0111] Align the defect data and the fused location data using their timestamps;

[0112] Using fused positioning data and mileage information for each defect point, cubic spline interpolation is performed on the mileage interval of each defect point to obtain the longitude and latitude of each defect point.

[0113] Specifically, by aligning the defect data with the location data based on the timestamps in the defect data and the fused location data, it is possible to find the time point in the fused location data that is closest in time to each defect point, i.e., to find the time point that satisfies the condition |t. d -t p The smallest t p , where t p =arg min tp |t d -t p |,t d Let t be the time point corresponding to the d-th defect point. p For the p-th time point in the fused positioning data, arg min tp This represents the parameter that minimizes the function value. After aligning the defect data and the fused location data, cubic spline interpolation is performed on each defect point within its mileage interval.

[0114] The mileage interval is the distance between the upstream ground marker positioning sensor and the downstream ground marker positioning sensor at the defect point.

[0115] The starting position of the mileage interval where the defect point is located is defined as the location of the upstream ground marker positioning sensor adjacent to the defect point. The upstream ground marker positioning sensor refers to the nearest ground marker positioning sensor that was buried before the defect point in the pipeline extension direction.

[0116] The end position of the mileage interval where the defect point is located is defined as the position of the downstream ground marker positioning sensor adjacent to the defect point, wherein the downstream ground marker positioning sensor refers to the closest ground marker positioning sensor buried after the defect point in the extension direction of the pipeline.

[0117] Since in the detection of pipeline defects, it is considered that the defect distribution is very sparse on a pipeline with good quality, and cubic spline interpolation fitting requires at least 4 parameters, the mileage segment distance cannot be set too small. In addition, since a ground marker positioning sensor is usually set every kilometer, the distance between the defect point and the ground marker positioning sensor needs to be included in the mileage information of the output defect point. Therefore, the mileage interval is set as the position interval between the upstream ground marker positioning sensor and the downstream ground marker positioning sensor of the defect point.

[0118] Exemplarily, the cubic spline interpolation fitting of the mileage interval where each defect point is located is realized by the following formula (7) and formula (8):

[0119] The longitude interpolation function corresponding to the dth defect point is defined as S_lon_d(m), and the latitude interpolation function is defined as S_lat_d(m), wherein: S_lon_d(m)=a 1d m 3 +a 2d m 2 +a 3d m+a 4d , formula (7); S_lat_d(m)=b 1d m 3 +b 2d m 2 +b 3d m+b 4d , formula (8);

[0120] Wherein, a 1d , a 2d , a 3d , a 4d , b 1d , b 2d , b 3d , b 4d are obtained by fitting the fusion positioning data, and m represents the mileage of each defect point.

[0121] Two linear equation groups can be constructed according to the mileage information, longitude information and latitude information provided in the fusion positioning data, and each equation group contains N equations corresponding to the data at N time points. The parameters of the two linear equation groups are solved to obtain a 1d , a 2d , a 3d , a 4d , and b 1db 2d b 3d b 4d .

[0122] The value interpolated according to the mileage of each defect point obtained by formula (7) is taken as the longitude corresponding to the defect point, and the value interpolated according to the mileage of each defect point obtained by formula (8) is taken as the latitude corresponding to the defect point.

[0123] In the above example, the defect data and the fusion positioning data are first aligned to facilitate the interpolation of the mileage data of the defect points. Then, the present disclosure segments the curve between the data points (referred to as data points for short) corresponding to each time point in the fusion positioning data into a group of cubic polynomials in the form of cubic spline interpolation, and requires that these cubic polynomials have the same first and second derivatives at adjacent data points to ensure the smoothness of the interpolation curve. Thus, a smooth data point curve is obtained, which can better fit the data points of the fusion positioning data and the mileage data of the defect points, and provide a high-precision interpolation result. Using the high-precision interpolation result as the longitude and latitude of the defect point can further improve the accuracy of defect point detection.

[0124] In one possible design, the method includes:

[0125] The longitude and latitude of each defect point are taken as the observation input into the second Kalman filter, i.e., Z = [S_lon_d(m), S_lat_d(m)] (Z represents a matrix composed of the longitude and latitude of each defect point fitted by cubic spline interpolation), and the fusion positioning data are taken as the predicted value of the system state input into the second Kalman filter.

[0126] The second Kalman filter is used to fit the longitude, latitude and fusion positioning data of each defect point to obtain the longitude and latitude of each defect point after fitting.

[0127] Since the fusion positioning data are derived from the IMU sensor, they may be affected by the IMU sensor noise and external environmental factors, and the longitude and latitude of the defect point obtained by spline interpolation may be limited by the density and quality of data sampling. In addition, the second Kalman filter can estimate the state of a dynamic system from a series of incomplete and noisy measurements. Therefore, the present disclosure uses the second Kalman filter to process and fuse the longitude and latitude data of the defect point obtained by spline interpolation and the fusion positioning data to optimize the longitude estimate value, the latitude estimate value and the fusion positioning data of the defect point. In addition, since the present disclosure can estimate the pipeline direction information according to the fusion positioning data to realize trajectory tracking, it can also improve the accuracy of the estimated pipeline direction information.

[0128] In one possible design, the method includes:

[0129] The longitude and latitude of each defect point after fitting are taken as observation values, and the fusion positioning data are input into the second Kalman filter again as the predicted values of the system state.

[0130] The longitude and latitude of each defect point after fitting are corrected again by using the second Kalman filter to fit the longitude and latitude of each defect point after fitting and the fusion positioning data.

[0131] The above examples use the second Kalman filter to correct the longitude and latitude of the defect points after fitting again in combination with the fusion positioning data, so as to further improve the positioning accuracy of the defect points.

[0132] Exemplarily, the method comprises:

[0133] determining the direction information of the pipeline according to the fusion positioning data;

[0134] displaying the pipeline direction map according to the direction information.

[0135] Exemplarily, the elevation information map is displayed according to the elevation information of the fusion positioning data.

[0136] Exemplarily, the pipeline direction and the specific positioning information of the defect points are displayed on the map, and the positioning information comprises longitude information, latitude information, mileage information and elevation information.

[0137] The original data output by the three-way mileage wheel of the magnetic flux leakage detection sensor is greatly affected by the pipeline environment, which affects the accuracy of the first positioning data output by the inertial measurement unit sensor, and therefore the accuracy of determining the position of the defect point according to the first positioning data is low. In order to improve the accuracy of positioning the pipeline defects, the second positioning data recorded by the ground marker positioning sensor is obtained, time synchronization processing and space inference positioning processing are adopted, the first positioning data is fused and corrected by using the second positioning data, the accuracy of the first positioning data is improved by the way of multi-parameter fusion from each sensor, and the fused and corrected first positioning data is the fusion positioning data. Therefore, fitting and comparing the defect data and the fusion positioning data can obtain the accurate longitude and latitude of each defect point, so as to timely carry out repair work.

[0138] As shown in FIG. 2, the present disclosure also provides a pipeline defect positioning system based on sensors, comprising:

[0139] The basic information import module is configured to input the original data of the three-way mileage wheel of the magnetic flux leakage detection sensor into the inertial measurement unit sensor after the magnetic flux leakage detection sensor finishes the magnetic flux leakage detection on the pipeline;

[0140] The basic information import module is configured to use an inertial measurement unit sensor to perform data processing on the raw data to obtain first positioning data. The first positioning data includes at least one first time point and the longitude and latitude corresponding to each first time point.

[0141] The basic information import module is configured to acquire second positioning data recorded by at least one ground marker positioning sensor after the magnetic flux leakage detection sensor completes the magnetic flux leakage detection of the pipeline. The at least one ground marker positioning sensor is buried along the pipeline direction. The second positioning data includes at least one second time point and the longitude and latitude corresponding to each second time point. The second time point is the time point when the magnetic flux leakage detection sensor passes the ground marker positioning sensor while running in the pipeline.

[0142] The first fusion algorithm module is configured to use time synchronization processing and spatial inference positioning processing to fuse the first positioning data and the second positioning data to obtain fused positioning data;

[0143] The display module is configured to use internal detection data analysis software to analyze and visualize the raw data;

[0144] The defect data determination module is configured to determine pipeline defect data based on the content visualized by the internal inspection data analysis software. The defect data includes the inspection time information and mileage information corresponding to each defect point.

[0145] The second fusion algorithm module is configured to fit and compare the defect data and the fused positioning data to obtain the longitude and latitude of each defect point.

[0146] In one feasible design, the system includes a project processing module configured to perform project initialization, project data saving, and loading of historical projects. A project refers to a single magnetic flux leakage detection (MFL) operation performed on a pipeline by a magnetic flux leakage sensor.

[0147] For example, the project processing module includes a project display interface for displaying newly created project data and historical project data. The project data includes first positioning data, second positioning data, fused positioning data, original defect data, and other data obtained in any process according to the embodiments of this disclosure. The data can be viewed in real time through the project display interface.

[0148] For example, the system includes an image display module configured to display a pipeline route diagram and an elevation information diagram.

[0149] For example, the system includes a map display module configured to display pipeline routing and defect location information.

[0150] For example, the map display module is configured to display and record in real time the location information of defects in the pipeline system and other relevant information. By inputting longitude and latitude information into the map display module, users can accurately locate the problem area in the pipeline system and take necessary maintenance measures in a timely manner. Furthermore, when any point on the pipeline route is clicked on the map, the system will quickly display the longitude and latitude information of that point (i.e., latitude and longitude information). This disclosure, by integrating map display and latitude and longitude information display functions, not only makes pipeline system monitoring more intuitive and efficient, but also provides users with more comprehensive data support.

[0151] In a feasible design, the first fusion algorithm module is implemented by using time synchronization processing and spatial inference positioning processing to fuse the first positioning data and the second positioning data to obtain fused positioning data:

[0152] Input the first positioning data and the second positioning data into the first neural network model;

[0153] Using a first neural network model, by identifying the time features in the first and second positioning data, time synchronization processing is performed on the first and second positioning data to align the time points in the second positioning data with the time points in the first positioning data.

[0154] Using a second neural network model, longitude and latitude functions are predicted based on the first positioning data. The longitude function is used to represent the correspondence between a time point sequence and longitude, and the latitude function is used to represent the correspondence between a time point sequence and latitude.

[0155] The second neural network model is used to correct the first positioning data based on longitude and latitude functions;

[0156] The corrected first positioning data is linearly interpolated using the second positioning data to obtain fused positioning data.

[0157] In a feasible design, the first fusion algorithm module optimizes the fused positioning data in the following way:

[0158] The longitude and latitude in the fused positioning data are used as one observation input to the first Kalman filter, and the corrected first positioning data is used as another observation input to the first Kalman filter.

[0159] The first Kalman filter is used to fit the fused positioning data and the first positioning data to obtain the fitted fused positioning data.

[0160] In a feasible design, the second fusion algorithm module is implemented by fitting and comparing the defect data and the fused localization data to obtain the longitude and latitude of each defect point, including:

[0161] Align the defect data and the fused location data using their timestamps;

[0162] Using fused positioning data and mileage information for each defect point, cubic spline interpolation is performed on the mileage interval where each defect point is located to obtain the longitude and latitude of each defect point.

[0163] In a feasible design, the second fusion algorithm module is implemented by fitting the longitude and latitude of each defect point to the fused positioning data in the following way:

[0164] The longitude and latitude of each defect point are used as observations and input into the second Kalman filter, and the fused positioning data is used as the predicted value of the system state and input into the second Kalman filter.

[0165] Using a second Kalman filter, the longitude, latitude, and fused positioning data of each defect point are fitted to obtain the fitted longitude and latitude of each defect point.

[0166] In a feasible design, the second fusion algorithm module is implemented by correcting the longitude and latitude of each defect point after fitting in the following way:

[0167] The longitude and latitude of each defect point after fitting are used as observations and then input into the second Kalman filter. The fused positioning data is used as the predicted value of the system state and input into the second Kalman filter.

[0168] The longitude and latitude of each defect point are fitted again using a second Kalman filter, and the fused positioning data are then fitted together. The longitude and latitude of each defect point are then corrected.

[0169] In one feasible design, the system includes a data preprocessing module configured to perform a first preprocessing on first positioning data and / or second positioning data, the first preprocessing comprising one or more of the following:

[0170] Data cleaning, time calibration and synchronization, and format conversion.

[0171] In one feasible design, the data preprocessing module is configured to perform a second preprocessing on the defective data, the second preprocessing comprising one or more of the following methods:

[0172] Format conversion processing, standardization, and normalization processing.

[0173] In one feasible design, the system includes a three-way odometer wheel synchronization module, which is configured to apply the odometer wheel synchronization processing mechanism on the internal detection data analysis software to synchronize the three odometer wheels before the raw data is analyzed and visualized using the internal detection data analysis software.

[0174] Other implementation methods and effects of the above system can be found in the description of the embodiment of the sensor-based pipeline defect location method, and will not be repeated here.

[0175] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0176] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0177] The block diagrams of devices, apparatuses, devices, and systems disclosed herein are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0178] It should also be noted that in the apparatus, devices, and methods of this disclosure, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions to this disclosure.

[0179] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0180] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.

Claims

1. A sensor-based pipeline defect positioning method, comprising: inputting original data of a three-way odometer of a magnetic flux leakage detection sensor to an inertial measurement unit sensor after the magnetic flux leakage detection sensor finishes detecting a pipeline; performing data resolution on the original data using the inertial measurement unit sensor to obtain first positioning data, the first positioning data comprising at least one first time point and longitude and latitude corresponding to each first time point; obtaining second positioning data recorded by at least one ground marker positioning sensor after the magnetic flux leakage detection sensor finishes detecting the pipeline, wherein the at least one ground marker positioning sensor is buried along a pipeline direction of the pipeline, and the second positioning data comprises at least one second time point and longitude and latitude corresponding to each second time point, the second time point being a time point when the magnetic flux leakage detection sensor passes through the ground marker positioning sensor while running in the pipeline; performing fusion on the first positioning data and the second positioning data using time synchronization processing and spatial inference positioning processing to obtain fused positioning data; analyzing and visually displaying the original data using internal detection data analysis software; determining defect data of the pipeline according to the visual display of the internal detection data analysis software, the defect data comprising detection time information and mileage information corresponding to each defect point; fitting and comparing the defect data and the fused positioning data to obtain longitude and latitude of each defect point.

2. The method of claim 1, wherein, The fusion of the first positioning data and the second positioning data using time synchronization processing and spatial inference positioning processing to obtain fused positioning data comprises: inputting the first positioning data and the second positioning data to a first neural network model; performing time synchronization processing on the first positioning data and the second positioning data by identifying time characteristics in the first positioning data and the second positioning data using the first neural network model to align the time points in the second positioning data with the time points in the first positioning data; predicting longitude and latitude functions according to the first positioning data using a second neural network model, the longitude function being used to represent a corresponding relationship between a time point sequence and longitude, and the latitude function being used to represent a corresponding relationship between a time point sequence and latitude; correcting the first positioning data according to the longitude function and the latitude function using the second neural network model; performing linear interpolation on the corrected first positioning data using the second positioning data to obtain fused positioning data.

3. The method of claim 2, wherein, The method comprises: inputting longitude and latitude in the fused positioning data as one observation into a first Kalman filter and inputting the corrected first positioning data as another observation into the first Kalman filter; performing fitting on the fused positioning data and the first positioning data using the first Kalman filter to obtain fitted fused positioning data.

4. The method of any one of claims 1-3, wherein, The fitting and comparison of the defect data and the fusion positioning data obtain the longitude and latitude of each defect point, and the method comprises: Aligning the defect data and the fusion positioning data by using the time stamps of the defect data and the fusion positioning data; Performing three times spline interpolation fitting on the mileage segment interval where each defect point is located by using the fusion positioning data and the mileage information of each defect point, to obtain the longitude and latitude of each defect point.

5. The method of claim 4, wherein, The method comprises: Inputting the longitude and latitude of each defect point as observation into a second Kalman filter, and inputting the fusion positioning data as the predicted value of the system state into the second Kalman filter; Using the second Kalman filter to fit the longitude, latitude of each defect point and the fusion positioning data, to obtain the longitude and latitude of each defect point after fitting.

6. The method of claim 5, wherein, The method comprises: Inputting the longitude and latitude of each defect point after fitting as observation into a second Kalman filter again, and inputting the fusion positioning data as the predicted value of the system state into the second Kalman filter; Using the second Kalman filter to fit the longitude and latitude of each defect point after fitting and the fusion positioning data again, and correcting the longitude and latitude of each defect point after fitting.

7. The method of any one of claims 1-3, wherein, Before the fusion of the first positioning data and the second positioning data by using time synchronization processing and space inference positioning processing, the method comprises: Performing first preprocessing on the first positioning data and / or the second positioning data, and the first preprocessing comprises one or more of the following modes: Data cleaning processing, time calibration synchronization processing and format conversion processing.

8. The method of claim 5 or 6, wherein, Before aligning the defect data and the fusion positioning data by using the time stamps of the defect data and the fusion positioning data, the method comprises: Performing second preprocessing on the defect data, and the second preprocessing comprises one or more of the following modes: Format conversion processing, standardization and normalization processing.

9. The method of any one of claims 1-3, wherein, Before analyzing and visually displaying the original data by using the internal detection data analysis software, the method comprises: Applying a synchronization processing mechanism of a mileage wheel on the internal detection data analysis software to perform synchronization processing on the three mileage wheels. 10.A sensor-based pipeline defect positioning system, comprising: A basic information import module configured to input original data of three mileage wheels of a magnetic flux leakage detection sensor to an inertial measurement unit sensor after the magnetic flux leakage detection sensor finishes magnetic flux leakage detection on a pipeline; The basic information import module is configured to use the inertial measurement unit sensor to perform data calculation on the original data to obtain first positioning data, and the first positioning data comprises at least one first time point and longitude and latitude corresponding to each first time point; The basic information import module is configured to acquire second positioning data recorded by at least one ground mark positioning sensor after the magnetic flux leakage detection sensor finishes the magnetic flux leakage detection on the pipeline, wherein the at least one ground mark positioning sensor is buried along a pipeline direction of the pipeline, and the second positioning data comprises at least one second time point and longitude and latitude corresponding to each second time point, the second time point being a time point when the magnetic flux leakage detection sensor passes through the ground mark positioning sensor while running in the pipeline; The first fusion algorithm module is configured to fuse the first positioning data and the second positioning data by using time synchronization processing and space inference positioning processing to obtain fused positioning data; The display module is configured to analyze and visually display the original data by using internal detection data analysis software; The defect data determination module is configured to determine defect data of the pipeline according to content visually displayed by the internal detection data analysis software, wherein the defect data comprises detection time information and mileage information corresponding to each defect point; The second fusion algorithm module is configured to fit and compare the defect data and the fused positioning data to obtain longitude and latitude of each defect point.

Images