Infrasonic wave detection method based on oil and gas pipeline defects
By combining an infrasound sensor array and an adaptive denoising algorithm with multi-sensor overlapping coverage technology, the problem of accurate positioning of infrasound detection under complex working conditions was solved, and high-precision identification and positioning of defects in oil and gas pipelines were achieved.
Patent Information
- Application Number
- CN202610121138.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-29
- Publication Date
- 2026-03-03
AI Technical Summary
Existing infrasound detection technology is difficult to achieve accurate detection under complex working conditions, is easily affected by environmental noise, and has large positioning errors, which cannot meet the needs of on-site operation and maintenance.
Signals are collected using an infrasound sensor array. A specific risk threshold is set by combining oil and gas status and operating parameters. The signal is processed by an adaptive denoising algorithm. The overlapping coverage of signals from multiple sensors is used for fuzzy and precise positioning. A defect feature library is built by combining machine learning to identify and precisely locate defect types.
It significantly improves the accuracy of defect location and identification, and can capture early weak defect signals and shield environmental noise interference under complex working conditions, so as to accurately distinguish defects such as cracks, corrosion and leakage.
Smart Images
Figure CN121595708A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of infrasound detection technology for defects in oil and gas pipelines, and more specifically, to a method for detecting defects in oil and gas pipelines using infrasound. Background Technology
[0002] As a core infrastructure for energy transportation, the safe and stable operation of oil and gas pipelines is directly related to the continuity of energy supply and the safety of the surrounding environment and personnel.
[0003] Existing infrasound detection technologies are insufficient to meet the precise detection requirements under complex operating conditions. Firstly, risk threshold settings lack flexibility, often employing fixed thresholds or simply adjusting a single operating parameter without considering the interrelationship between oil / gas conditions, operating parameters, and the effective detection distance of the sensor. This leads to misjudgments due to environmental noise interference under complex conditions such as high pressure, high flow, or drastic temperature changes, or the omission of early, weak defect signals due to excessively high thresholds. Secondly, defect location accuracy is limited, relying heavily on calculations based on the signal arrival time or amplitude attenuation of a single sensor. It fails to employ a layered strategy of narrowing the range through fuzzy positioning and optimizing for precise positioning, and does not fully utilize the overlapping coverage signals of multiple sensors, resulting in significant location errors and failing to meet the precise location requirements of on-site maintenance. Therefore, a method for detecting defects in oil and gas pipelines based on infrasound is proposed. Summary of the Invention
[0004] The purpose of this invention is to provide a method for detecting defects in oil and gas pipelines using infrasound, so as to solve the problems mentioned in the background art.
[0005] To achieve the above objectives, a method for detecting defects in oil and gas pipelines based on infrasound is provided, comprising the following steps:
[0006] S1. Deploy an array of infrasound sensors to collect infrasound signals from oil and gas pipelines, and simultaneously acquire oil and gas status parameters, operating parameters, and the installation location of each infrasound sensor.
[0007] S2. Set an effective detection distance for the infrasound sensor based on oil and gas state parameters and operating parameters. Then, set a specific risk threshold based on the effective detection distance. Preprocess the infrasound signal using an adaptive noise reduction algorithm and compare the preprocessed infrasound signal with the corresponding specific risk threshold.
[0008] S3. For infrasound signals whose comparison results exceed the exclusive risk threshold, risk features are extracted. Based on the risk features, the location of fuzzy defects in the oil and gas pipeline is located. Based on the location of the fuzzy defects, the synchronous infrasound signals of other acoustic sensors with overlapping detection distances are extracted.
[0009] S4. Combine the synchronous infrasound signal obtained in S3 with the historical infrasound signal of the ambiguous defect location, extract the signal difference features and construct the difference feature dataset. Then, fuse the difference feature dataset with the risk features to form a comprehensive feature. At the same time, compare the preset defect feature library with the comprehensive feature to calculate the similarity of each defect type.
[0010] S5. Set weighting coefficients based on the distance between each infrasound sensor and the fuzzy defect location. Combine the weighting coefficients with similarity for weighted fusion, select the optimal matching defect type, and then combine the defect type, signal difference features, and infrasound sensor to accurately locate and update the fuzzy defect location through a multi-sensor joint localization model, outputting the defect type and accurate defect location.
[0011] As a further improvement to this technical solution, in step S1, multiple infrasound sensors are arranged axially along the oil and gas pipeline, and the detection range of the infrasound sensors covers the entire oil and gas pipeline. At the same time, a communication module is established to establish a data connection between all the infrasound sensors and the communication module, thereby collecting the infrasound signals fed back by the infrasound sensors.
[0012] As a further improvement to this technical solution, in step S1, a data connection is established between the communication module and the oil and gas management terminal, and the real-time oil and gas status parameters, operating parameters, and installation positions of each acoustic sensor in the oil and gas pipeline are synchronously acquired in the oil and gas management terminal.
[0013] As a further improvement to this technical solution, step S2 is as follows:
[0014] S2.1. Set the baseline effective detection distance according to the equipment parameters of the infrasound sensor, and then make personalized adjustments to the baseline effective detection distance according to the oil and gas state parameters and operating parameters to obtain the effective detection distance of each infrasound sensor in the oil and gas pipeline with the installation position as the starting point.
[0015] S2.2. Set a specific risk threshold by combining the effective detection distance of the infrasound sensor, oil and gas state parameters, and operating parameters;
[0016] S2.3. The infrasound signals fed back by each infrasound sensor are preprocessed by an adaptive wavelet threshold denoising algorithm. Then, in the same infrasound sensor, the preprocessed infrasound signal is compared with the specific risk threshold.
[0017] If the infrasound signal exceeds the exclusive risk threshold, it is determined to be a risky infrasound signal and S3 is performed.
[0018] If the infrasound signal does not exceed the designated risk threshold, monitoring will continue.
[0019] As a further improvement to this technical solution, the reference effective detection distance corresponds to standard operating parameters, which include standard pressure, standard temperature, and standard delivery flow rate.
[0020] Oil and gas condition parameters and operating parameters include actual pressure, actual temperature, and actual transport flow rate;
[0021] The higher the actual pressure is compared to the standard pressure, the farther the effective detection distance and the higher the risk threshold.
[0022] The greater the difference between the actual temperature and the standard temperature, the shorter the effective detection distance and the higher the risk threshold.
[0023] The larger the actual flow rate is compared to the standard flow rate, the shorter the effective detection distance and the higher the risk threshold.
[0024] The shorter the effective detection distance, the higher the risk threshold.
[0025] As a further improvement to this technical solution, step S3 is as follows:
[0026] S3.1 Extract risk features from the risk infrasound signal, obtain the risk features corresponding to the risk infrasound signal, and then calculate the time difference of the infrasound signal arriving at the infrasound sensor based on the arrival time of the risk features. At the same time, locate the fuzzy defect position by combining the time difference with the propagation speed of infrasound in the oil and gas pipeline.
[0027] S3.2. Combine the effective detection distance of each acoustic sensor with the location of the fuzzy defect to perform an overlap and coverage analysis;
[0028] When the effective detection distance overlaps with and covers the location of the fuzzy defect, the infrasound sensor is extracted.
[0029] If the effective detection distance does not overlap with the location of the fuzzy defect, the infrasound sensor will not be extracted.
[0030] S3.3 Based on the infrasound sensor extracted in S3.2, extract the synchronous infrasound signal during the period when the risky infrasound signal occurs.
[0031] As a further improvement to this technical solution, in S4, in the communication module, the historical infrasound signal of the fuzzy defect location is obtained from the infrasound sensor extracted in S3.2, and then the historical infrasound signal is combined with the synchronous infrasound signal for difference analysis to obtain signal difference features. Then, the signal difference features are summarized to construct a difference feature dataset.
[0032] The differential feature dataset is fused with the risk features of the risk infrasound signal to form a comprehensive feature.
[0033] As a further improvement to this technical solution, in step S4, historical defect features and corresponding defect types are collected at the oil and gas management end, and a defect identification model is established by combining historical defect features and corresponding defect types through machine learning, and a defect feature library is preset in the defect identification model.
[0034] Then, the comprehensive features are input into the defect identification model for identification, and the model outputs the similarity between the comprehensive features and the corresponding defect features for each defect type.
[0035] As a further improvement to this technical solution, in S5, the straight-line distance between the infrasound sensor extracted in S3.2, the infrasound sensor corresponding to the risk infrasound signal, and the fuzzy defect location is calculated, and then the weight coefficient of each sensor is set by using the reciprocal of the distance.
[0036] Calculate the similarity of various defects corresponding to the infrasound signals of each infrasound sensor. Multiply the similarity of each infrasound sensor of the same defect type by the corresponding weight coefficient, and then sum all the products to obtain the weighted average similarity of the defect type. Then select the defect type with the highest similarity as the optimal matching defect type and output the defect type result.
[0037] Combining defect type, signal difference characteristics, and infrasound sensor, a multi-sensor joint localization model is constructed using the least squares method. With the precise coordinates of the defect as the optimization objective, the optimal three-dimensional coordinates of the defect are obtained and used as the precise defect location.
[0038] Finally, the output shows the defect type and precise defect location.
[0039] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0040] 1. In this infrasound detection method for defects in oil and gas pipelines, a layered positioning strategy is adopted, which combines fuzzy positioning to narrow down the range with multi-sensor joint precise positioning. First, the fuzzy defect location is quickly locked by the signal time difference and amplitude attenuation law. Then, combined with the defect type, signal difference characteristics and sensor spatial information, the least squares method is used to construct an optimization model, which greatly improves the accuracy of defect positioning and controls the positioning error within the acceptable range of actual operation and maintenance, providing accurate location reference for on-site defect handling.
[0041] 2. In this infrasound detection method for oil and gas pipeline defects, by establishing a linkage relationship between oil and gas status, operating parameters and the effective detection distance of the sensor, a personalized risk threshold is dynamically set. This effectively solves the problems of misjudgment and missed judgment caused by fixed thresholds or simple threshold adjustments in the existing technology, significantly improves the accuracy of risk infrasound signal screening, and ensures that early weak defect signals can be captured under complex operating conditions, while effectively shielding environmental noise interference.
[0042] 3. In this infrasound detection method for oil and gas pipeline defects, a comprehensive feature vector is constructed by fusing the core features of risky infrasound signals with the differences in historical and contemporary signals. A distance-weighted mechanism is introduced to optimize similarity calculation, which enables defect type identification to fully combine the effectiveness differences of multi-sensor signals, significantly improving the reliability and accuracy of defect identification. It effectively avoids misjudgment caused by single features or single sensor signals, and achieves accurate differentiation of common defects such as cracks, corrosion, leakage, and deformation. Attached Figure Description
[0043] Figure 1 This is a schematic flowchart of an infrasound detection method for defects in oil and gas pipelines according to the present invention.
[0044] Figure 2 This is a flowchart of S1 of the present invention;
[0045] Figure 3 This is a flowchart of S2 of the present invention;
[0046] Figure 4 This is a flowchart of S3 of the present invention;
[0047] Figure 5 This is a flowchart of S4 of the present invention;
[0048] Figure 6 This is a flowchart of S5 of the present invention. Detailed Implementation
[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0050] Please see Figures 1-6 As shown, the purpose of this embodiment is to provide a method for detecting defects in oil and gas pipelines using infrasound, including the following steps:
[0051] S1. Deploy an array of infrasound sensors to collect infrasound signals from oil and gas pipelines, and simultaneously acquire oil and gas state parameters, operating parameters, and installation locations of each infrasound sensor; complete the hardware deployment of the detection system and the spatiotemporal synchronous acquisition of basic data to provide data support and spatial reference for subsequent threshold setting and location analysis.
[0052] In S1, multiple infrasound sensors are arranged axially along the oil and gas pipeline. The detection range of the infrasound sensors covers the entire oil and gas pipeline. At the same time, a communication module is established to connect all the infrasound sensors with the communication module, thereby collecting the infrasound signals fed back by the infrasound sensors.
[0053] Calculate the axial spacing of the sensors to ensure that the detection ranges of adjacent sensors overlap (overlap rate ≥ 10%), avoid detection blind spots, and ensure that the detection ranges of all sensors completely cover the total length of the pipeline.
[0054] Install infrasound sensors along the axial direction of the oil and gas pipeline at the calculated spacing. The distance between the sensor and the outer wall of the pipeline is controlled between 5-20cm to ensure the stability of the sensor's signal acquisition. At the same time, build an industrial-grade communication module and establish a two-way data connection between all infrasound sensors and the communication module using wired or wireless transmission methods. Set the data transmission protocol to ensure that the infrasound signals acquired by the sensors are transmitted to the communication module in real time without loss.
[0055] S1 establishes a data connection between the communication module and the oil and gas management terminal, and synchronously obtains real-time oil and gas status parameters, operating parameters, and installation positions of various acoustic sensors in the oil and gas pipeline in the oil and gas management terminal.
[0056] S2. Set an effective detection distance for the infrasound sensor based on oil and gas state parameters and operating parameters. Then, set a specific risk threshold based on the effective detection distance. Preprocess the infrasound signal using an adaptive denoising algorithm and compare the preprocessed infrasound signal with the corresponding specific risk threshold. Dynamically adjust the sensor's detection capability and judgment criteria based on operating parameters. Combined with the denoising algorithm to purify the signal, achieve accurate screening of risk signals and avoid misjudgments and missed judgments caused by fixed thresholds.
[0057] The steps for S2 are as follows:
[0058] S2.1. Set the baseline effective detection distance according to the equipment parameters of the infrasound sensor, and then adjust the baseline effective detection distance according to the oil and gas state parameters and operating conditions. Obtain the effective detection distance of each infrasound sensor in the oil and gas pipeline with the installation position as the starting point. The steps are as follows:
[0059] Based on the equipment parameters (sensitivity, operating frequency, signal-to-noise ratio) of the infrasound sensor, the benchmark effective detection distance under standard operating conditions is calibrated, and the standard operating condition parameters, including standard pressure, standard temperature, and standard delivery flow rate, are defined as the benchmark for subsequent personalized adjustments.
[0060] Real-time oil and gas status parameters (actual pressure, actual temperature) and operating parameters (actual delivery flow rate) are obtained from the oil and gas management terminal. Simultaneously, the baseline effective detection distance is corrected according to the rule that higher actual pressure results in greater distance and greater temperature difference / flow rate results in shorter distance, to obtain the actual effective detection distance of each sensor. The formula is as follows: ;
[0061] in, This represents the actual effective detection distance of the sensor. This is the baseline effective detection distance under standard operating conditions. / This is the ratio of actual pressure to standard pressure. This represents the absolute difference between the actual temperature and the standard temperature. / This is the ratio of the actual flow rate to the standard flow rate. , , The correction factor was calibrated through on-site experiments.
[0062] S2.2. Set a specific risk threshold by combining the effective detection distance of the infrasound sensor, oil and gas state parameters, and operating parameters;
[0063] Based on the background infrasound signal under standard operating conditions, a baseline risk threshold is set. Then, combined with the actual effective detection distance and real-time operating parameters, the threshold is adjusted to a higher level for each sensor, following the principle that the shorter the effective detection distance and the more complex the operating conditions. The specific risk threshold is then bound to the corresponding sensor and stored in the management system, as shown in the following formula:
[0064] ;
[0065] in, For sensor-specific risk thresholds, This is the baseline risk threshold under standard operating conditions. / This is the ratio of the baseline detection distance to the actual detection distance. , , This is the threshold correction coefficient, calibrated on-site.
[0066] S2.3. The infrasound signals fed back by each infrasound sensor are preprocessed by an adaptive wavelet threshold denoising algorithm. Then, in the same infrasound sensor, the preprocessed infrasound signal is compared with the specific risk threshold.
[0067] The adaptive wavelet threshold denoising algorithm consists of the following steps: performing wavelet decomposition on the original signal to obtain low-frequency approximation coefficients and high-frequency detail coefficients; adaptively calculating the threshold based on the signal-to-noise ratio; processing the high-frequency coefficients using an improved soft threshold function; and reconstructing the signal using inverse wavelet transform to output the purified infrasound signal.
[0068] If the infrasound signal exceeds the exclusive risk threshold, it is determined to be a risky infrasound signal and S3 is performed.
[0069] If the infrasound signal does not exceed the designated risk threshold, monitoring will continue.
[0070] The benchmark effective detection distance corresponds to standard operating parameters, which include standard pressure, standard temperature, and standard delivery flow rate.
[0071] Oil and gas condition parameters and operating parameters include actual pressure, actual temperature, and actual transport flow rate;
[0072] The higher the actual pressure is compared to the standard pressure, the farther the effective detection distance and the higher the risk threshold.
[0073] The greater the difference between the actual temperature and the standard temperature, the shorter the effective detection distance and the higher the risk threshold.
[0074] The larger the actual flow rate is compared to the standard flow rate, the shorter the effective detection distance and the higher the risk threshold.
[0075] The shorter the effective detection distance, the higher the risk threshold.
[0076] S3. For infrasound signals whose comparison results exceed the exclusive risk threshold, risk features are extracted. Based on the risk features, the location of fuzzy defects in the oil and gas pipeline is located. Based on the location of the fuzzy defects, the synchronous infrasound signals of other acoustic sensors with overlapping detection distances are extracted. Key features are extracted from the risk signals to initially lock the fuzzy range where the defect is located. Effective sensors covering this range are selected to define the target range for subsequent feature analysis and precise positioning.
[0077] The steps for S3 are as follows:
[0078] S3.1 Extract risk features from the risk infrasound signal, obtain the risk features corresponding to the risk infrasound signal, and then calculate the time difference of the infrasound signal arriving at the infrasound sensor based on the arrival time of the risk features. At the same time, locate the fuzzy defect position by combining the time difference with the propagation speed of infrasound in the oil and gas pipeline.
[0079] Extract infrasound data segments identified as risk signals, extract core risk characteristic parameters, including signal peak value, peak frequency, duration, and timestamp of signal arrival at the corresponding sensor. Classify and store the extracted risk characteristic parameters according to sensor number to establish a sensor-risk characteristic association dataset.
[0080] Select the sensor that triggers the risk signal and its two adjacent valid sensors. Extract the timestamps of the three sensors receiving the same risk signal. Then, calculate the signal arrival time difference between any two sensors. Combined with the propagation speed of infrasound in the pipe wall (corrected for operating conditions), preliminarily calculate the distance range between the defect and each sensor. Finally, combine the distance ranges of multiple sensor sets to determine the axial and circumferential interval boundaries and output the fuzzy defect location. The formula is as follows:
[0081] ;
[0082] in, The speed at which infrasound propagates within the pipe wall after working condition correction. The propagation speed under standard operating conditions, such as that of a steel pipeline. =5000m / s, / This is the ratio of actual pressure to standard pressure. This represents the absolute difference between the actual temperature and the standard temperature. , The correction factor was calibrated in the field, with a value range of 0 < , <0.05;
[0083] S3.2. Combine the effective detection distance of each acoustic sensor with the location of the fuzzy defect to perform an overlap and coverage analysis;
[0084] Retrieve the actual effective detection distance and three-dimensional installation coordinates of each acoustic wave sensor, calculate the straight-line distance from each sensor to the center of the fuzzy defect location, and compare the straight-line distance with the effective detection distance of the sensor. If the straight-line distance is less than the effective detection distance, it is determined to overlap; otherwise, it is not overlapping. The formula is as follows:
[0085] ;
[0086] ;
[0087] in, Let be the distance difference from the center of the fuzzy defect location to sensor i and sensor j. , For the timestamps of the risk signals received by sensors i and j, The axial mounting distance between sensor i and sensor j is... The circumferential angle for sensor placement. The straight-line distance from the center of the fuzzy defect location to sensor i;
[0088] When the effective detection distance overlaps with and covers the location of the fuzzy defect, the infrasound sensor is extracted.
[0089] If the effective detection distance does not overlap with the location of the fuzzy defect, the infrasound sensor will not be extracted.
[0090] S3.3. Based on the infrasound sensor data extracted in S3.2, extract the synchronous infrasound signal during the period when the risk infrasound signal occurs. Determine the occurrence time window of the risk signal, extract the infrasound signal data within the above time window from all marked target sensors, classify and store them according to sensor number, and form a target sensor-synchronous signal dataset;
[0091] S4. Combine the synchronous infrasound signal obtained in S3 with the historical infrasound signal at the ambiguous defect location to extract signal difference features and construct a difference feature dataset. Then, fuse the difference feature dataset with the risk features to form a comprehensive feature. At the same time, compare the pre-set defect feature library with the comprehensive feature to calculate the similarity of each defect type. Extract defect-related features through the difference analysis of synchronous signal and historical signal, construct a comprehensive feature vector by combining the risk features, and use a machine learning model to achieve a preliminary determination of the defect type.
[0092] S4, in the communication module, the historical infrasound signal of the fuzzy defect location is obtained from the infrasound sensor extracted in S3.2. Then, the historical infrasound signal is combined with the synchronous infrasound signal for difference analysis to obtain signal difference features. Finally, the signal difference features are summarized to construct a difference feature dataset.
[0093] From the historical database of the communication module, retrieve the infrasound signals of the pipeline section corresponding to the fuzzy defect location, filter the pressure, temperature, and flow rate of the historical operating conditions with the current operating conditions with a similarity of ≥90%, and ensure that there are no risk signal records in the historical period to ensure the consistency of the background signals. Then, classify and store the filtered historical infrasound signals according to the sensor number to form a sensor-historical signal benchmark dataset.
[0094] For each target sensor, extract the core features in the time domain and frequency domain of the synchronous signal and the historical signal respectively, calculate the feature difference value of the two sets of signals, remove meaningless fluctuation features, retain the difference parameters that are strongly correlated with defects, and summarize all difference features according to sensor number to construct a difference feature dataset.
[0095] The differential feature dataset is fused with the risk features of the risk infrasound signal to form a comprehensive feature.
[0096] The constructed differential feature dataset and the previously extracted risk infrasound signal features (peak value, peak frequency, duration, etc.) are retrieved and fused using the feature vector concatenation method. The differential features and risk features are combined in dimensional order to form a one-dimensional comprehensive feature vector, ensuring that there is no redundancy or omission of features.
[0097] In S4, historical defect features and corresponding defect types are collected at the oil and gas management end. A defect identification model is established by combining historical defect features and corresponding defect types through machine learning, and a defect feature library is preset in the defect identification model.
[0098] Then, the comprehensive features are input into the defect identification model for identification. The defect identification model outputs the similarity between the comprehensive features and the corresponding defect features for each defect type. The steps are as follows:
[0099] Historical defect data is collected at the oil and gas management end, including infrasound features and corresponding defect type labels for four types of defects: cracks, corrosion, leakage, and deformation. An improved random forest algorithm is used to build a defect identification model. The number of decision trees (100-500 trees) and the maximum depth (5-20 layers) are optimized through grid search. The AdaBoost ensemble mechanism is introduced to improve the generalization ability. The standard feature parameters of the four types of defects are input into the model, and a pre-set defect feature library is used as the benchmark for similarity comparison.
[0100] Then, the fused comprehensive feature vector is input into the defect recognition model. The model uses the cosine similarity algorithm to calculate the matching degree between the comprehensive feature and the standard features of various defects in the defect feature library, and outputs the similarity value between the comprehensive feature and each defect type (the value ranges from 0 to 1). The closer the similarity is to 1, the higher the matching degree.
[0101] S5. Weighting coefficients are set based on the distance between each infrasound sensor and the ambiguous defect location. These weighting coefficients are then combined with similarity for weighted fusion to select the optimal matching defect type. Finally, combining the defect type, signal difference characteristics, and infrasound sensor data, a multi-sensor joint localization model is used to accurately locate and update the ambiguous defect location, outputting the defect type and precise defect location. The defect type determination result is optimized by distance weighting, and the final output of the defect type and precise location is achieved by combining the multi-sensor joint localization model.
[0102] In S5, the straight-line distance between the infrasound sensor extracted in S3.2, the infrasound sensor corresponding to the risk infrasound signal, and the fuzzy defect location is calculated, and then the weight coefficient of each sensor is set by the reciprocal of the distance.
[0103] Retrieve the three-dimensional installation coordinates of the target sensor and the sensor corresponding to the risk signal extracted in S3.2, and at the same time obtain the center three-dimensional coordinates of the fuzzy defect location. Using the three-dimensional spatial distance formula, calculate the straight-line distance between each sensor and the center of the fuzzy defect location, form a distance dataset and bind it to the sensor number, which is consistent with the calculation method in S3.2.
[0104] Based on the straight-line distance between each sensor and the location of the fuzzy defect, the initial weight is calculated using the inverse of the distance. The initial weights of all sensors are then normalized to ensure that the sum of all weight coefficients is 1. Finally, the unique weight coefficient for each sensor is obtained, as shown in the following formula:
[0105]
[0106]
[0107] in, Let be the initial weight of the i-th sensor. For the i-th sensor, the normalized weight coefficients are the specific weights. Let be the straight-line distance between the i-th sensor and the center of the fuzzy defect location. This is a correction factor (values range from 0.1 to 1.0m). The total number of sensors involved in the calculation;
[0108] Calculate the similarity of various defects corresponding to the infrasound signals of each infrasound sensor. Multiply the similarity of each infrasound sensor of the same defect type by the corresponding weight coefficient, and then sum all the products to obtain the weighted average similarity of the defect type. Then select the defect type with the highest similarity as the optimal matching defect type and output the defect type result.
[0109] Retrieve the similarity data of various defects corresponding to each sensor, and establish a sensor-defect type-similarity association table. For each type of defect, multiply the similarity of each corresponding sensor by its specific weight coefficient, and then sum all the products to obtain the weighted average similarity of the defect type. Compare the weighted average similarity of all defect types, and select the defect type with the highest value as the optimal matching result. If the maximum value is lower than 0.8, output a suspected defect that needs to be checked. The formula is as follows:
[0110] ;
[0111] in, Let be the weighted average similarity of the k-th type of defect. Let be the similarity of the k-th type of defect corresponding to the i-th sensor, where k is the defect type number. There are four types: crack defect, corrosion defect, leakage defect, and deformation defect.
[0112] Combining defect type, signal difference characteristics, and infrasound sensor, a multi-sensor joint localization model is constructed using the least squares method. With the precise coordinates of the defect as the optimization objective, the optimal three-dimensional coordinates of the defect are obtained by minimizing the weighted sum of squared differences between the distances from each sensor to the candidate defect location and the actual distances based on signal difference characteristics. These coordinates are then used as the precise defect location.
[0113] The defect type, signal difference characteristics of each sensor (such as amplitude attenuation and time difference), and sensor 3D coordinate information are integrated as input to the localization model. Using the precise 3D coordinates of the defect as the optimization objective, a least-squares optimization function is constructed. The objective is to minimize the weighted sum of squared differences between the distance from each sensor to the candidate defect location and the actual distance calculated based on signal difference characteristics. The optimization function is then solved to obtain the optimal 3D coordinates of the defect, i.e., the precise defect location. The formula is as follows:
[0114]
[0115] in, This is the weighted sum of squared differences (optimizing the objective function value). Let be the straight-line distance from the i-th sensor to the candidate defect location. The actual distance from the i-th sensor to the defect is calculated based on signal difference characteristics;
[0116]
[0117] in, The precise three-dimensional coordinates of the defect, i.e., the optimal solution. The three-dimensional coordinates corresponding to the minimum value of the objective function J;
[0118] Finally, the defect type and precise defect location are output; the best matching defect type and the calculated precise defect location (3D coordinates) are summarized, and a standardized inspection report is output, which includes information such as defect type, precise coordinates, weighted average similarity, and positioning error.
[0119] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.
Claims
1. A method for detecting defects in oil and gas pipelines using infrasound, characterized in that: Includes the following steps: S1. Deploy an array of infrasound sensors to collect infrasound signals from oil and gas pipelines, and simultaneously acquire oil and gas status parameters, operating parameters, and the installation location of each infrasound sensor. S2. Set an effective detection distance for the infrasound sensor based on oil and gas state parameters and operating parameters. Then, set a specific risk threshold based on the effective detection distance. Preprocess the infrasound signal using an adaptive noise reduction algorithm and compare the preprocessed infrasound signal with the corresponding specific risk threshold. S3. For infrasound signals whose comparison results exceed the exclusive risk threshold, risk features are extracted. Based on the risk features, the location of fuzzy defects in the oil and gas pipeline is located. Based on the location of the fuzzy defects, the synchronous infrasound signals of other acoustic sensors with overlapping detection distances are extracted. S4. Combine the synchronous infrasound signal obtained in S3 with the historical infrasound signal of the ambiguous defect location, extract the signal difference features and construct the difference feature dataset. Then, fuse the difference feature dataset with the risk features to form a comprehensive feature. At the same time, compare the preset defect feature library with the comprehensive feature to calculate the similarity of each defect type. S5. Set weighting coefficients based on the distance between each infrasound sensor and the fuzzy defect location. Combine the weighting coefficients with similarity for weighted fusion, select the optimal matching defect type, and then combine the defect type, signal difference features, and infrasound sensor to accurately locate and update the fuzzy defect location through a multi-sensor joint localization model, outputting the defect type and accurate defect location.
2. The infrasound detection method for defects in oil and gas pipelines according to claim 1, characterized in that: In step S1, multiple infrasound sensors are arranged axially along the oil and gas pipeline, and the detection range of the infrasound sensors covers the entire oil and gas pipeline. At the same time, a communication module is established to establish a data connection between all the infrasound sensors and the communication module, thereby collecting the infrasound signals fed back by the infrasound sensors.
3. The infrasound detection method for defects in oil and gas pipelines according to claim 2, characterized in that: S1 establishes a data connection between the communication module and the oil and gas management terminal, and synchronously acquires real-time oil and gas status parameters, operating parameters, and installation positions of various acoustic sensors in the oil and gas pipeline in the oil and gas management terminal.
4. The infrasound detection method for defects in oil and gas pipelines according to claim 1, characterized in that: The steps in S2 are as follows: S2.
1. Set the baseline effective detection distance according to the equipment parameters of the infrasound sensor, and then make personalized adjustments to the baseline effective detection distance according to the oil and gas state parameters and operating parameters to obtain the effective detection distance of each infrasound sensor in the oil and gas pipeline with the installation position as the starting point. S2.
2. Set a specific risk threshold by combining the effective detection distance of the infrasound sensor, oil and gas state parameters, and operating parameters; S2.
3. The infrasound signals fed back by each infrasound sensor are preprocessed by an adaptive wavelet threshold denoising algorithm. Then, in the same infrasound sensor, the preprocessed infrasound signal is compared with the specific risk threshold. If the infrasound signal exceeds the exclusive risk threshold, it is determined to be a risky infrasound signal and S3 is performed. If the infrasound signal does not exceed the designated risk threshold, monitoring will continue.
5. The infrasound detection method for defects in oil and gas pipelines according to claim 4, characterized in that: The reference effective detection distance corresponds to standard operating parameters, which include standard pressure, standard temperature, and standard delivery flow rate. Oil and gas condition parameters and operating parameters include actual pressure, actual temperature, and actual transport flow rate; The higher the actual pressure is compared to the standard pressure, the farther the effective detection distance and the higher the risk threshold. The greater the difference between the actual temperature and the standard temperature, the shorter the effective detection distance and the higher the risk threshold. The larger the actual flow rate is compared to the standard flow rate, the shorter the effective detection distance and the higher the risk threshold. The shorter the effective detection distance, the higher the risk threshold.
6. The infrasound detection method for defects in oil and gas pipelines according to claim 4, characterized in that: The steps in S3 are as follows: S3.1 Extract risk features from the risk infrasound signal, obtain the risk features corresponding to the risk infrasound signal, and then calculate the time difference of the infrasound signal arriving at the infrasound sensor based on the arrival time of the risk features. At the same time, locate the fuzzy defect position by combining the time difference with the propagation speed of infrasound in the oil and gas pipeline. S3.
2. Combine the effective detection distance of each acoustic sensor with the location of the fuzzy defect to perform an overlap and coverage analysis; When the effective detection distance overlaps with and covers the location of the fuzzy defect, the infrasound sensor is extracted. If the effective detection distance does not overlap with the location of the fuzzy defect, the infrasound sensor will not be extracted. S3.3 Based on the infrasound sensor extracted in S3.2, extract the synchronous infrasound signal during the period when the risky infrasound signal occurs.
7. The infrasound detection method for defects in oil and gas pipelines according to claim 6, characterized in that: In step S4, the communication module obtains historical infrasound signals of the ambiguous defect location from the infrasound sensor extracted in step S3.2, then performs difference analysis by combining the historical infrasound signals with the infrasound signals of the same period to obtain signal difference features, and then summarizes the signal difference features to construct a difference feature dataset. The differential feature dataset is fused with the risk features of the risk infrasound signal to form a comprehensive feature.
8. The infrasound detection method for defects in oil and gas pipelines according to claim 7, characterized in that: In S4, historical defect features and corresponding defect types are collected at the oil and gas management end. A defect identification model is established by combining historical defect features and corresponding defect types through machine learning, and a defect feature library is preset in the defect identification model. Then, the comprehensive features are input into the defect identification model for identification, and the model outputs the similarity between the comprehensive features and the corresponding defect features for each defect type.
9. The infrasound detection method for defects in oil and gas pipelines according to claim 1, characterized in that: In S5, the straight-line distance between the infrasound sensor extracted in S3.2, the infrasound sensor corresponding to the risk infrasound signal, and the fuzzy defect location is calculated, and then the weight coefficient of each sensor is set by using the reciprocal of the distance. Calculate the similarity of various defects corresponding to the infrasound signals of each infrasound sensor. Multiply the similarity of each infrasound sensor of the same defect type by the corresponding weight coefficient, and then sum all the products to obtain the weighted average similarity of the defect type. Then select the defect type with the highest similarity as the optimal matching defect type and output the defect type result. Combining defect type, signal difference characteristics, and infrasound sensor, a multi-sensor joint localization model is constructed using the least squares method. With the precise coordinates of the defect as the optimization objective, the optimal three-dimensional coordinates of the defect are obtained and used as the precise defect location. Finally, the output shows the defect type and precise defect location.