A system for detecting defects on the inner wall of oil pipelines

Abstract

The present invention belongs to the technical field of oil pipeline inner wall defect detection, and specifically relates to an oil pipeline inner wall defect detection system, comprising a pipe section division module, a wall thickness analysis module, a straight pipe section detection module, and a curved pipe section detection module. The oil pipeline is divided into straight pipe sections and curved pipe sections, and the straight pipe sections are divided into different areas according to wall thickness. Ultrasonic incident detection and eddy current scanning are then simultaneously performed in the different wall thickness areas. Eddy current and ultrasonic dual-modal detection are fused to generate a three-dimensional map containing defect boundary and depth information. The curved pipe section is divided into deposition risk areas based on stress distribution simulated by flow field simulation. Adaptive defect detection methods are then adopted in different areas, thereby achieving targeted detection of oil pipeline inner wall defects, largely avoiding missed identification, and greatly improving the integrity and reliability of defect detection.

Description

Technical Field

[0001] The invention belongs to the technical field of oil pipeline inner wall defect detection, and in particular relates to an oil pipeline inner wall defect detection system. Background Art

[0002] As energy demand grows, oil, as a critical resource, plays a vital role in industry and society. To ensure its safe transportation, long-distance pipeline systems are widely used. However, due to high pressure, corrosive media, and environmental factors, pipeline inner walls are prone to defects such as cracks and corrosion, which can lead to leaks and even accidents. Therefore, inner wall defect detection is a crucial means to ensure safe pipeline operation.

[0003] Because inspections during operation can disrupt fluid stability and pose safety risks, they are currently mostly performed while pipelines are shut down. Traditional manual inspections are inefficient and unsafe, making them inadequate for modern pipeline inspections. To address this, inspection robots equipped with scanning equipment are now widely used to enter pipelines and automatically identify and analyze defects through image capture.

[0004] Numerous existing solutions employ inspection robots to identify defects on the inner walls of oil pipelines. For example, Chinese invention patent publication number CN119444722A proposes an image processing-based defect recognition method for pipeline inspection robots. This method uses an image acquisition device to continuously acquire images of the inner wall of the pipeline, and combines image preprocessing, defect area segmentation, feature extraction, and classification to achieve automated defect recognition.

[0005] Another example is the Chinese invention patent with publication number CN118329932A, which proposes a combined optical and magnetized eddy current detection method for pipeline inner wall defects. It proposes using an inspection robot equipped with an optical surface inspection mechanism and a magnetized eddy current inspection mechanism to achieve composite inspection of inner wall surface defects and buried defects during industrial pipeline shutdowns, thereby improving the comprehensiveness and accuracy of inspections.

[0006] However, these technical solutions still have certain limitations in practical application: 1. During oil transportation, due to factors such as flow rate fluctuations and pipe bends, components such as wax, asphalt, and sand in the oil are prone to depositing on the pipe walls, especially in bends. These deposits can obscure actual defects, and direct use of image acquisition or eddy current testing can result in missed defects or misidentification, seriously affecting the reliability of the inspection results.

[0007] 2. Due to welded joints during pipeline manufacturing and installation, pipe wall thickness distribution is uneven. In particular, the wall thickness in welded or thickened areas is significantly greater than the standard area. Relying solely on optical imaging or surface eddy current testing makes it difficult to obtain defect depth information and accurately assess defect severity, which in turn affects subsequent repair decisions.

[0008] In summary, the existing unified pipeline defect detection method has the problem of insufficient adaptability when facing complex working conditions such as deposition interference and wall thickness variation, which may lead to incomplete defect identification and inaccurate quantification in some areas, ultimately affecting the overall detection effect and the accuracy of safety assessment. Summary of the Invention

[0009] The purpose of the present invention is to improve the deficiencies in the prior art and provide a system for detecting inner wall defects of oil pipelines. The system can detect inner wall defects of oil pipelines by dividing the oil pipelines into straight sections and curved sections, and implementing differentiated detection strategies in straight sections based on wall thickness distribution and in curved sections based on deposition risk, respectively.

[0010] The purpose of the present invention can be achieved through the following technical solutions: A system for detecting inner wall defects of oil pipelines, comprising the following modules: a pipe segment division module: an inertial navigation unit is used to capture the axial inclination change rate of the pipeline in real time, and the pipeline is divided into curved pipe segments and straight pipe segments according to a preset inclination change threshold.

[0011] Wall thickness analysis module: An ultrasonic array is deployed in the straight pipe section to obtain wall thickness data through continuous axial scanning to generate a two-dimensional wall thickness map. Different wall thickness areas are then divided based on the layer-by-layer expansion of the wall thickness differences of neighboring pixels.

[0012] Straight pipe section inspection module: In different wall thickness areas, the ultrasonic wave is adjusted to the optimized incident angle based on the wall thickness-ultrasonic incident angle mapping relationship constructed through experimental calibration to perform focused scanning to generate ultrasonic echoes. At the same time, the annular electromagnetic eddy current probe is driven to scan and obtain the inner wall surface profile, and then the eddy current and ultrasonic dual-modal data are integrated to generate a three-dimensional map of surface defects.

[0013] Bend Detection Module: This module constructs a petroleum flow model based on historical operating parameters to simulate the flow field in a bend. During the simulation, it outputs an equivalent stress distribution cloud map, thereby dividing areas into high- and low-risk deposition areas. In high-risk deposition areas, dual-frequency ultrasonic scanning is triggered to identify defects after time-frequency separation of sediments. In low-risk deposition areas, single-frequency ultrasonic scanning is triggered to identify defects based on echo anomalies.

[0014] Combining all the above technical solutions, the positive effects of the present invention are as follows: 1. The present invention divides the oil pipeline into straight sections and curved sections. The straight sections are divided into different areas according to the wall thickness. Ultrasonic and eddy current dual-modal detection are fused to generate a three-dimensional map containing defect boundaries and depth information. The curved sections are divided into deposition risk areas based on the stress distribution of flow field simulation, and a differentiated detection strategy is implemented. This method can achieve accurate identification of defects on the inner wall of the pipeline, improve detection integrity, effectively avoid missed detection, and significantly improve detection reliability and accuracy.

[0015] 2. When performing ultrasonic testing on straight pipe sections with different wall thicknesses, the present invention establishes a mapping relationship between wall thickness and ultrasonic incident angle based on experimental calibration, and adjusts the probe incident angle accordingly. This method enables ultrasonic waves to achieve optimal coupling and penetration under different wall thicknesses, improves the quality of defect echo signals, enhances detection sensitivity and positioning accuracy, and thus improves the accuracy and reliability of defect identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention is further described with reference to the accompanying drawings. However, the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative effort.

[0017] Figure 1 It is a schematic diagram of the system module structure of the present invention.

[0018] Figure 2 Schematic diagram of the implementation of the inner wall defect detection of the oil pipeline in the present invention.

[0019] Figure 3 This is a flowchart for implementing the layer-by-layer expansion and division of regions with different wall thicknesses based on the difference in wall thickness of neighboring pixels in the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0021] The present invention provides a system for detecting inner wall defects of a petroleum pipeline, which comprises a pipe section division module, a wall thickness analysis module, a straight pipe section detection module and a curved pipe section detection module.

[0022] See Figure 1 As shown, the pipe segment division module is connected to the wall thickness analysis module and the curved pipe segment detection module respectively, and the wall thickness analysis module is connected to the straight pipe segment detection module.

[0023] The pipe segment division module captures the pipeline axial inclination change rate in real time through an inertial navigation unit, and divides the pipeline into curved pipe segments and straight pipe segments according to a preset inclination change threshold.

[0024] See also Figure 2 FIG. 1 is a schematic diagram of an implementation of the inner wall defect detection of a petroleum pipeline according to the present invention.

[0025] Preferably, the specific implementation content of the above module is as follows: linear acceleration data of the pipeline in three-dimensional space is continuously collected along the pipeline axis by an inertial navigation unit composed of an accelerometer.

[0026] The pipeline inclination at each measuring point is calculated based on the collected three-dimensional linear acceleration data, and the inclination change rate between adjacent measuring points is further calculated.

[0027] It should be noted that the accelerometer is used to measure linear acceleration along three orthogonal axes. The coordinate system is set to assume that the pipeline axis is the Z axis, and the directions perpendicular to the axis are the X axis and the Y axis respectively. When the inertial navigation unit moves along the pipeline in the pipeline, the accelerometer will measure the gravity components in various directions under the action of the gravity field. If the pipeline is completely horizontal, the accelerometer reading on the Z axis should be close to 0g, and the readings on the X and Y axes correspond to the components of gravity acceleration, respectively. The pipeline inclination at each measuring point can be calculated through the trigonometric function relationship based on the readings on the X, Y, and Z axes. The specific calculation operation belongs to the scope of the existing technology and will not be repeated here.

[0028] To supplement the above explanation, when using an inertial navigation unit to measure pipeline attitude, it is necessary to ensure the operational stability of the detection equipment during movement to avoid dynamic disturbances caused by mechanical vibration or shock that may interfere with the inclination angle calculation. This is because under non-ideal static conditions, the accelerometer responds not only to gravity but also to non-gravitational accelerations generated by the movement of the detection equipment, which can lead to errors in the inclination angle calculation.

[0029] In an example implementation of the above operation, the inclination change rate between adjacent measurement points can be obtained by calculating the ratio of the difference between the inclination angles of the two points to the distance between the two points. The ratio reflects the angle change per unit length, that is, the local curvature characteristic of the pipeline axis in space.

[0030] The inclination change rate of adjacent measurement points is compared with the preset inclination change threshold. When the inclination change rate of adjacent points exceeds the threshold, the current area is determined to be a curved pipe section, and the arc length range of the curved pipe section is recorded. Conversely, if the inclination change rate does not exceed the threshold, the corresponding section is classified as a straight pipe section.

[0031] It should be understood that the above-mentioned division of straight pipe sections and curved pipe sections by the rate of change of the inclination angle of adjacent measuring points is based on the consideration of the significant differences in the geometric structures of straight pipe sections and curved pipe sections in oil pipelines. The axis of the straight pipe section is basically straight, and the inclination angle of each point along the axis changes little, usually only causing slight fluctuations due to terrain undulations or installation errors. The axis of the curved pipe section has an obvious directional turning, resulting in a large change in the inclination angle between adjacent measuring points, which is manifested as a significant increase in the rate of change of the inclination angle. When the rate of change of the inclination angle of adjacent measuring points exceeds the threshold, it indicates that the pipeline has made a significant turn in this area and is determined to be a curved pipe section. If the rate of change of the inclination angle does not exceed the threshold, it is considered that the pipeline in this section has a flat direction and is classified as a straight pipe section.

[0032] The above-mentioned inclination change threshold can be set, for example, by collecting and analyzing measured data of multiple pipelines with known structural parameters, and calculating the maximum natural inclination change rate of their straight pipe sections caused by terrain undulations or installation deviations under actual laying conditions, which serves as the engineering basis for setting the threshold.

[0033] As another example, based on the requirements for the minimum bending radius of the curved pipe section in the pipeline design specifications, the corresponding theoretical inclination angle change per unit length can be derived through geometric relationships, which serves as a theoretical reference basis for setting the threshold value.

[0034] The fundamental reason for this segmentation of oil pipelines is that pipelines, influenced by topography and terrain during their actual laying process, are not ideally straight structures but rather exhibit geometric variations such as bends and undulations. To achieve more accurate and efficient inner wall defect detection, it is necessary to divide the entire pipeline into straight and curved sections based on these geometric features, enabling the implementation of targeted inspection strategies.

[0035] The wall thickness analysis module is used to deploy ultrasonic array sensors in the straight pipe section to obtain wall thickness data through continuous axial scanning to generate a two-dimensional wall thickness map, and divide different wall thickness areas based on the layer-by-layer expansion of the wall thickness differences of neighboring pixels.

[0036] As a way to implement the above solution, deploying ultrasonic array sensors in the straight pipe section is implemented as follows: multiple ultrasonic phased array probes are set in the straight pipe section and are evenly spaced along the circumference of the pipe to form a ring-shaped detection array.

[0037] It's important to note that an ultrasonic phased array probe consists of multiple independently controlled piezoelectric chips. Each chip can independently transmit and receive ultrasonic signals. The phased array probe can generate and control ultrasonic beams at multiple angles, enabling wide-area scanning of the pipe's inner wall without mechanical movement. Multiple probes are evenly spaced around the circumference, forming a complete 360° circumferential inspection coverage structure, ensuring synchronized scanning of the entire circumference and avoiding blind spots.

[0038] After the inspection robot enters the pipeline with a circular detection array, it sets the corresponding scanning step according to the effective detection coverage range of a single scan during the axial uniform speed advancement.

[0039] The detection robot triggers a full-circle synchronous scan each time it moves forward a set step, and each probe transmits and receives ultrasonic signals in turn to collect raw data.

[0040] The collected original signal is converted into digital form and processed digitally to extract the time difference information of the echo signal. The wall thickness is calculated based on the material sound velocity to obtain the wall thickness value of each circumferential sampling point at the current axial position.

[0041] By continuously moving and splicing multi-position data, a two-dimensional wall thickness map of the straight pipe section along the axial and circumferential directions of the pipe is constructed.

[0042] It should be understood that the principle of using ultrasonic pulses to detect the wall thickness of straight pipe sections is based on the property that ultrasonic waves can penetrate the inner wall of the pipe and be reflected by the bottom surface. Specifically, the ultrasonic probe transmits high-frequency sound waves to the inner wall of the pipe. After propagating to the opposite wall of the pipe, the sound waves are reflected and returned to the probe to be received. By recording the time difference between ultrasonic emission and reception and combining it with the sound velocity of the internal material under ultrasonic conditions, half of the time difference is multiplied by the sound velocity to calculate the pipe wall thickness value at the probe location.

[0043] As another way to implement the above solution, see Figure 3 As shown in FIG, the process of dividing different wall thickness areas by layer-by-layer expansion based on the difference in wall thickness of neighboring pixels is as follows: a neighborhood range is selected with each pixel point as the center in the two-dimensional wall thickness map of the straight pipe section.

[0044] In the specific implementation of the above solution, the neighborhood range of each pixel point is the neighborhood range of the pixel point formed by extending the directly adjacent neighboring pixel points around the pixel point as the center.

[0045] The wall thickness values ​​are compared within the neighborhood of each pixel. The wall thickness difference between the central pixel and other pixels in the neighborhood is calculated and compared with the configured limited wall thickness difference. If the wall thickness difference between a neighboring pixel and the central pixel is less than or equal to the limited wall thickness difference, the neighboring pixel is used as the new center point, and the above comparison process is repeated within its neighborhood, expanding outward in sequence until a pixel with a wall thickness difference exceeding the limited wall thickness difference is encountered. Thus, a set of pixels with continuous and consistent wall thickness characteristics is formed and defined as a wall thickness area.

[0046] If the wall thickness difference between a certain adjacent pixel point and the central pixel point exceeds the specified wall thickness difference, the neighborhood range is reconstructed with this pixel point as the new starting point and expanded outward, eventually dividing the entire straight pipe section into multiple areas with different wall thicknesses.

[0047] In the above implementation, the defined wall thickness difference serves as the key criterion for determining whether adjacent pixels belong to the same wall thickness region. This is determined by referring to the maximum wall thickness deviation allowed by the pipe inner wall material manufacturing standard and combining it with the measurement accuracy of the detection equipment. Specifically, a safety margin corresponding to the detection system error is added to the maximum wall thickness deviation to determine the final defined wall thickness difference. This ensures that the resulting region division not only conforms to the actual wall thickness distribution characteristics but also possesses sufficient robustness and reliability.

[0048] Understandably, in actual engineering applications, variations in pipe wall thickness can arise from structural reinforcements designed into the design or uneven corrosion and wear during long-term operation. Such variations in wall thickness typically exhibit a gradual distribution. By analyzing the thickness differences between neighboring pixels and performing a layer-by-layer expansion, we adaptively segment and classify regions of varying wall thickness within the 2D wall thickness map of straight pipe sections, thereby enabling automated identification of wall thickness distribution characteristics.

[0049] The straight pipe section detection module is used to adjust the ultrasound to the optimized incident angle in different wall thickness areas according to the wall thickness-ultrasonic incident angle mapping relationship constructed by experimental calibration, perform focused scanning to generate ultrasonic echoes, and simultaneously drive the annular electromagnetic eddy current probe to scan and obtain the inner wall surface contour, and then fuse the eddy current and ultrasonic dual-modal data to generate a three-dimensional map of surface defects.

[0050] It's important to note that even when the wall thickness is known, it's still necessary to match the ultrasonic incident angle to the specific wall thickness region for scanning. This is primarily aimed at improving the identifiability and detection accuracy of defect echo signals. While pipe wall thickness can be measured using the ultrasonic pulse-echo method, where the probe transmits ultrasonic waves and receives reflected signals from the bottom surface, calculating the total wall thickness based on the propagation time and material sound velocity, in the presence of localized defects, some of the ultrasonic waves will be prematurely reflected at the defect site, forming a defect echo. Because defect echo signals are typically weak and easily masked by noise under non-optimal incident conditions, failing to adjust the ultrasonic incident angle based on the wall thickness of the current region and perform focused scanning may result in ineffective excitation or capture of clear defect echoes, leading to missed detections or misjudgments. Conventional wall thickness measurement doesn't adjust the incident angle because it primarily focuses on bottom surface echo information and is insensitive to defect characteristics. Therefore, matching the incident angle and focusing control for different wall thickness regions is a key step in improving defect recognition and obtaining accurate defect depth information.

[0051] Furthermore, by dynamically adjusting the ultrasonic incident angle, optimal control of the detection depth can be achieved, thereby improving the detection resolution and defect detection rate in the target area. Generally speaking, for areas with thicker walls, using a larger incident angle helps to enhance the penetration of the sound beam and improve the longitudinal resolution of defects; while for areas with thinner walls, using a smaller incident angle helps to improve the lateral resolution and enhance the ability to identify subtle defects.

[0052] Specifically, the mapping relationship between the wall thickness and the ultrasonic incident angle can be established through the following experimental calibration process: first, a group of standard samples with different known wall thicknesses are selected and divided into a training set and a test set.

[0053] During the calibration process, key parameters such as the operating frequency and bandwidth of the phased array ultrasonic probe are uniformly set, and environmental conditions such as temperature and humidity are kept constant to ensure data consistency.

[0054] Phased array ultrasonic testing equipment is used to scan each standard sample in the training set at multiple preset incident angles, and the corresponding echo signal quality indicators, such as signal-to-noise ratio and resolution, are collected and recorded.

[0055] Based on the obtained signal quality evaluation results, the optimal incident angle was screened out under each standard wall thickness condition, and then the functional relationship model between wall thickness and optimal incident angle was fitted by regression analysis method.

[0056] Finally, the model was validated using a test set to evaluate its prediction accuracy and generalization ability, and the model parameters were optimized accordingly, ultimately establishing a wall thickness-ultrasonic incident angle mapping relationship suitable for actual inspection scenarios.

[0057] In the specific implementation of the above scheme, in different wall thickness areas, the ultrasound is adjusted to the optimized incident angle according to the wall thickness-ultrasonic incident angle mapping relationship constructed by experimental calibration to perform focused scanning to generate ultrasonic echoes. Please refer to the following process: in the straight pipe section, the average wall thickness is calculated for the wall thickness values ​​of all pixel points in different wall thickness areas, and the optimized incident angle of the corresponding area is determined in combination with the wall thickness-ultrasonic incident angle mapping relationship established by experimental calibration.

[0058] After the inspection robot enters the area with a phased array probe, it adjusts the emission angle according to the optimized incident angle and performs focused scanning to obtain ultrasonic echoes in the corresponding wall thickness area.

[0059] In a further implementation of the above scheme, the annular electromagnetic eddy current probe is driven to scan and obtain the inner wall surface profile as follows: while performing ultrasonic focused scanning on different wall thickness areas of the straight pipe section, the annular electromagnetic eddy current probe array is driven to continuously apply an alternating electromagnetic field.

[0060] It should be noted that electromagnetic eddy current testing is based on the principle of electromagnetic induction and uses alternating current to excite eddy currents in metal pipes, the distribution of which is affected by the inner wall structure. When defects such as corrosion, cracks, and dents are present, the local conductive path changes, causing eddy current density disturbances, which are reflected as changes in the probe impedance or voltage signal. Through signal acquisition and inversion processing, the eddy current response can be converted into height or depth information and mapped to a three-dimensional spatial coordinate system to reconstruct the surface profile of the inner wall of the pipeline. The inner wall surface profile can intuitively reflect its geometric integrity and identify the spatial distribution and morphological characteristics of surface defects such as corrosion, wear, and pits. Since most oil pipelines are made of metal and have good electrical conductivity, they are particularly suitable for electromagnetic eddy current testing methods.

[0061] In particular, the annular electromagnetic eddy current probe array is an annular probe array composed of multiple independent electromagnetic eddy current sensor units, which are arranged at equal intervals along the circumference of the pipeline to form 360° detection coverage without blind spots, similar to the annular detection array composed of multiple ultrasonic phased array probes.

[0062] The eddy current response signal is converted into height or depth related to the surface topography through signal inversion, and then mapped to a three-dimensional spatial coordinate system to generate the inner wall surface profile.

[0063] It should be pointed out that the conversion of electromagnetic eddy current response signals into height or depth information associated with the surface morphology through signal inversion is usually based on existing technical means such as electromagnetic field numerical modeling, impedance analysis and inverse problem solving. It is a conventional signal processing method in electromagnetic non-destructive testing and will not be elaborated here.

[0064] In a further implementation of the above scheme, the eddy current and ultrasonic dual-modal data are integrated to generate a three-dimensional map of surface defects as follows: the inner wall surface profile and ultrasonic echo obtained by scanning different wall thickness areas are aligned in time and space.

[0065] The purpose of the above-mentioned time-space alignment is to ensure that the data of the two modes - namely the three-dimensional contour of the inner wall and the corresponding ultrasonic echo signal - are consistent in the time point and spatial position of acquisition, so that data from different detection modes can be analyzed in the same reference frame, thereby improving the accuracy of defect identification and quantitative analysis.

[0066] The defect area is identified and located from the aligned inner wall surface contour, and the defect opening boundary is extracted. At the same time, the depth distribution profile of the defect along the wall thickness direction is constructed in combination with the diffraction time difference data from the ultrasonic echo.

[0067] When identifying defects in the surface contour of the inner wall of a pipeline, the image gradient analysis method can be used to achieve automated detection. The specific implementation process is as follows: First, the acquired three-dimensional contour image of the inner wall is filtered and preprocessed to suppress noise interference and improve image quality.

[0068] Then, the gradient components of the image in the horizontal and vertical directions are calculated respectively, the gradient maps in the two directions are constructed, and the overall gradient intensity map is obtained by vector synthesis.

[0069] A critical threshold is set based on the distribution characteristics of the gradient image, and areas with gradient amplitudes higher than the threshold are marked as potential abnormal areas. This threshold is used to distinguish normal surfaces from areas with possible structural abnormalities such as dents and cracks. Its setting can be determined by statistically analyzing the gradient histograms of a large number of defect-free areas and determining the maximum gradient upper limit of the normal surface as the critical threshold.

[0070] Morphological operations are further performed on the potential abnormal areas, such as opening operations to remove isolated noise points and closing operations to fill internal holes. Finally, abnormal areas with continuous features are identified through connected domain analysis and defined as defect areas.

[0071] It's important to note that image gradients reflect the spatial rate of change in image brightness or intensity. Normal regions typically have low gradients due to their smooth and uniform surfaces. Defective regions, on the other hand, exhibit high local contrast and gradient response due to localized geometric abrupt changes. Therefore, they can be used to effectively distinguish between normal and abnormal structures.

[0072] The diffraction time difference of ultrasonic echo refers to the fact that when ultrasonic waves encounter defects in the pipeline, part of the energy will be reflected at the defect, forming a defect echo. By recording the time difference between these echoes reaching the probe, the distance of the defect relative to the inner wall is calculated using the material sound velocity combined with the time difference between the transmitted pulse and the received defect echo.

[0073] The 2D defect boundary of the surface contour and the depth distribution profile are fused in 3D to generate a 3D surface defect map with dimension annotations, where the dimensions include the defect area and defect depth.

[0074] The bend detection module is used to construct an oil flow model based on historical operating parameters to simulate the flow field of the bend section. During the simulation, it outputs an equivalent stress distribution cloud map, thereby dividing high- and low-risk deposition areas. In the high-risk deposition area, dual-frequency ultrasonic scanning is triggered to identify defects after separating the sediments based on time and frequency. In the low-risk deposition area, single-frequency ultrasonic scanning is triggered to identify defects based on echo anomalies.

[0075] Optionally, an oil flow model is constructed based on historical operating parameters to simulate the flow field of the bend section, and an equivalent stress distribution cloud map is output during the simulation process, thereby dividing the high and low risk deposition areas. The implementation is as follows: a three-dimensional digital twin model containing the structural characteristics of the bend section is constructed based on the pipeline geometric parameters and material properties, and a distributed strain sensor network is virtually deployed along the circumferential and axial directions of the bend section in the model.

[0076] The three-dimensional digital twin model constructed by the above-mentioned pipeline geometric parameters such as pipe diameter, wall thickness, bending radius, and material properties such as elastic modulus and Poisson's ratio has high fidelity and can truly reflect the flow behavior of oil in the bend section under actual transportation conditions. In addition, the stress sensor network deployed in the bend section of the model is used to simulate and collect the structural stress response caused by the fluid passing through the bend.

[0077] The petroleum physical properties and transportation condition data collected during historical operation are used as boundary conditions and input into the digital twin model to simulate petroleum transportation.

[0078] The above-mentioned CNPC physical properties such as viscosity, density, and transportation conditions data such as flow rate, pressure, temperature, etc.

[0079] During the simulation process, the virtual distributed strain sensor network is used to collect the equivalent stress distribution cloud diagram under the stress formation time series at each node of the bend section in real time.

[0080] It is important to understand that curved pipe sections, due to geometric changes, are prone to causing flow rate drops, pressure fluctuations, and eddy currents, making them key areas where sediments tend to accumulate. The scouring and shearing effects of the fluid on the pipe wall will change the local stress distribution. High stress concentration often reflects flow instability and can be used as an indirect assessment indicator of sedimentation tendency, in line with the basic principles of engineering fluid mechanics and structural mechanics. While it is difficult to obtain real-time fluid parameters during an outage, digital twin models driven by historical data can accurately reproduce flow field characteristics. Combined with a virtually deployed distributed strain sensor network, they can effectively replace physical sensors, reduce hardware costs, and improve detection flexibility.

[0081] The equivalent stress distribution cloud map in the time series is clustered and analyzed frame by frame to divide it into several stress cluster areas.

[0082] The ratio of the stress value of each stress cluster area to the average stress value of the adjacent areas is defined as the stress concentration index of the area.

[0083] The stress concentration index is classified and judged according to the set stress concentration threshold:

[0084] If the stress concentration in a cluster exceeds the warning threshold, it is directly marked as a high-risk area for deposits. Areas with local stress significantly higher than the average often correspond to structural weaknesses or areas of severe flow disturbance, and are also more prone to impurity deposits.

[0085] The stress concentration warning thresholds described above reflect the critical mechanical state at which deposition risk occurs. Specifically, laboratory flow simulations or numerical analysis can be used to identify the critical stress variation range that is prone to impurity deposition. Based on this, an empirical correlation model between stress concentration and deposition probability is constructed, and a response curve is plotted to determine the threshold that matches the deposition tendency.

[0086] For areas where stress concentration is lower than the warning threshold, the sliding time window analysis is further combined to statistically analyze the stress concentration change trend in multiple consecutive time windows. If the stress concentration in the area is found to show a continuous upward trend, it is marked as a high-risk deposition area. Otherwise, it is classified as a low-risk deposition area.

[0087] As a specific implementation of the above scheme, when the stress concentration in a certain stress clustering area is lower than the set warning threshold, it is further determined whether there is an increase by comparing the stress concentration changes at the current moment with the previous moment. If an increase in stress concentration is detected, the duration window of the growth state is recorded. If growth occurs in three consecutive time windows, it is determined that the stress concentration in the area shows a continuous upward trend.

[0088] The present invention combines cluster analysis with sliding window trend judgment when dividing sedimentation risk areas based on the flow simulation model of the bend section. It takes into account both the stress differences in space and the evolution characteristics in the time dimension. It can not only identify the current high-risk areas, but also predict the development trend of potential risk areas, and is forward-looking.

[0089] Further optionally, dual-frequency ultrasonic scanning is triggered in the high-risk deposition area, and defect identification is performed based on time-frequency separation of the sediments as follows: the detection robot is controlled to carry the dual-frequency ultrasonic probe to the high-risk deposition area, and the probe synchronously transmits two groups of high-frequency and low-frequency ultrasonic pulse signals to the area to be tested, and receives the composite reflection echo signal generated by the combined action of the pipeline structure and the sediment layer.

[0090] Time-frequency processing is performed on the acquired composite echo signal, and the thickness of the deposited layer is estimated by comparing the time delay of the high-frequency and low-frequency signals in the propagation path and combining the material sound velocity.

[0091] It should be understood that high-frequency ultrasound has a shorter wavelength, while low-frequency ultrasound has a longer wavelength. For areas with sediment, low-frequency ultrasound can better penetrate the sediment layer, reach the pipeline substrate and obtain its internal structure information. By simultaneously emitting high-frequency and low-frequency ultrasonic pulse signals, the high-frequency signal is more easily affected by the sediment layer, resulting in a significant time delay in its echo signal. Therefore, by analyzing the time delay of the high-frequency and low-frequency echo signals, the thickness of the sediment layer can be estimated.

[0092] Based on the estimated sediment thickness information, the composite echo signal is processed with sediment interference separation and signal compensation to remove the influence of the sediment layer on the low-frequency echo signal and focus on the low-frequency effective echo component.

[0093] The above steps are intended to suppress the interference of the sediment layer on the low-frequency echo signal and extract the effective low-frequency echo component after penetrating the sediment layer.

[0094] The processed low-frequency effective echo component is compared with the normal echo curve of the deposition-free area, where the normal echo curve is the echo curve without defects. By identifying the differences in echo amplitude, waveform, and propagation time, and combining the diffraction time difference, the defect location is performed, and the identified defect area is mapped to the three-dimensional geometric model of the bent pipe section.

[0095] It is important to note that defects can cause abnormal echo characteristics such as amplitude attenuation, waveform distortion, or propagation time offset. By comparing the amplitude, waveform, and propagation time of the low-frequency effective echo component with the normal echo curve of the deposition-free area, once any of the above anomalies is detected, the presence of a defect can be determined. The defect depth can then be calculated in combination with the ultrasonic diffraction time difference method, thereby achieving accurate defect positioning and depth assessment.

[0096] Further optionally, a single-frequency ultrasonic scan is triggered in a low-risk deposition area, and defects are identified based on echo anomalies. The specific implementation is as follows: the detection robot is controlled to carry a dual-frequency ultrasonic probe to move to a low-risk deposition area, and the probe transmits a low-frequency ultrasonic pulse signal to the area to be tested and receives a reflected echo signal.

[0097] After converting the reflected echo signal into a time-domain waveform diagram and comparing it with the normal echo curve of the deposition-free area, the defects are located by identifying the differences in the amplitude, waveform, and propagation time of the echo, and combining it with the diffraction time difference method. The identified defect area is then mapped to the three-dimensional geometric model of the bent pipe section.

[0098] It should be noted that due to the low likelihood of deposits occurring in low-risk areas, there is no need to employ complex dual-frequency ultrasonic scanning strategies when there is no significant deposit interference. In this case, only low-frequency ultrasonic testing is required, which allows for efficient defect identification and accurate judgment while maintaining a high signal-to-noise ratio, thereby improving inspection efficiency and reliability.

[0099] It should be noted that in curved sections, due to their geometric complexity and fluid dynamics, stress concentration and deposition are of primary concern. Therefore, dual-frequency ultrasonic scanning is employed to address the complex deposition interference and stress distribution. In straight sections, while the geometry is relatively simple and the stress distribution is relatively uniform, potential wall thickness inhomogeneities and their impact on ultrasonic propagation require wall thickness analysis and optimized incident angles to improve detection accuracy. Multimodal data fusion technology is also employed to generate detailed 3D defect maps.

[0100] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0101] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0102] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0103] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0104] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A system for detecting defects on the inner wall of a petroleum pipeline, characterized in that: Includes the following modules: Pipe segmentation module: This module uses an inertial navigation unit to capture the pipeline's axial inclination change rate in real time and divides the pipeline into curved and straight sections based on a preset inclination change threshold. Wall thickness analysis module: An ultrasonic array is deployed in the straight pipe section to obtain wall thickness data through continuous axial scanning to generate a two-dimensional wall thickness map. Different wall thickness areas are then divided layer by layer based on the thickness differences of neighboring pixels. Straight pipe section inspection module: In different wall thickness areas, the ultrasonic wave is adjusted to the optimized incident angle based on the wall thickness-ultrasonic incident angle mapping relationship established through experimental calibration to generate focused scanning and ultrasonic echoes. At the same time, the annular electromagnetic eddy current probe is driven to scan and obtain the inner wall surface profile, and then the eddy current and ultrasonic dual-modal data are integrated to generate a three-dimensional map of surface defects. Bend detection module: This module constructs an oil flow model based on historical operating parameters to simulate the flow field in a bend. During the simulation, it outputs an equivalent stress distribution cloud map, thereby dividing areas into high- and low-risk deposition areas. In high-risk deposition areas, dual-frequency ultrasonic scanning is triggered to identify defects after time-frequency separation of sediments. In low-risk deposition areas, single-frequency ultrasonic scanning is triggered to identify defects based on echo anomalies.

2. The oil pipeline inner wall defect detection system according to claim 1, characterized in that: The specific contents of the pipe segment division module are as follows: The inertial navigation unit composed of accelerometers continuously collects the linear acceleration data of the pipeline in three-dimensional space along the pipeline axis, thereby obtaining the pipeline inclination at each measuring point and calculating the inclination change rate of adjacent measuring points; The inclination change rate of adjacent measurement points is compared with the preset inclination change threshold. When the inclination change rate of adjacent points exceeds the threshold, the current area is determined to be a curved pipe section, and the arc length range of the curved pipe section is recorded. Conversely, if the inclination change rate does not exceed the threshold, the corresponding section is classified as a straight pipe section.

3. The oil pipeline inner wall defect detection system according to claim 1, characterized in that: The two-dimensional wall thickness map is generated as follows: Multiple ultrasonic phased array probes are arranged in a straight pipe section at equal intervals along the circumference to form a circular detection array; The detection robot is equipped with a circular detection array and moves forward at a constant speed. The scanning step is set according to the coverage range of a single scan, and a full-circle synchronous scan is triggered after each forward step. Each probe transmits and receives ultrasonic signals in turn to collect the original echo signal. Extract the echo time difference from the collected original echo signal, and calculate the wall thickness based on the material sound velocity to obtain the wall thickness value of each circumferential sampling point at the current axial position; By continuously moving and splicing multi-position data, a two-dimensional wall thickness map of the straight pipe section along the axial and circumferential directions of the pipe is constructed.

4. The oil pipeline inner wall defect detection system according to claim 1, characterized in that: The process of dividing the different wall thickness areas layer by layer based on the difference in wall thickness of neighboring pixels is as follows: In the two-dimensional wall thickness map of the straight pipe section, a neighborhood range is selected with each pixel as the center, and the wall thickness difference between it and each pixel in the neighborhood is calculated, and then compared with the configured limited wall thickness difference; If the wall thickness difference between a certain neighboring pixel and the central pixel is less than or equal to the specified wall thickness difference, then the neighboring pixel is used as the new center point and the area is expanded outwards in sequence within its neighborhood until a pixel whose wall thickness difference exceeds the specified wall thickness difference is encountered. Thus, a set of continuous pixels is formed and defined as a wall thickness area. If the wall thickness difference between a certain adjacent pixel point and the central pixel point exceeds the specified wall thickness difference, the neighborhood range is reconstructed with this pixel point as the new starting point and expanded outward, eventually dividing the entire straight pipe section into multiple areas with different wall thicknesses.

5. The oil pipeline inner wall defect detection system according to claim 4, characterized in that: The process of adjusting the ultrasound to the optimized incident angle based on the wall thickness-ultrasonic incident angle mapping relationship constructed by experimental calibration in different wall thickness areas for focused scanning and generating ultrasonic echoes is as follows: In the straight pipe section, the average wall thickness is calculated for all pixel points in different wall thickness areas, and the optimized incident angle of the corresponding area is determined based on the wall thickness-ultrasonic incident angle mapping relationship established by experimental calibration. After the inspection robot enters the area with a phased array probe, it adjusts the emission angle according to the optimized incident angle and performs focused scanning to obtain ultrasonic echoes in the corresponding wall thickness area.

6. The oil pipeline inner wall defect detection system according to claim 1, characterized in that: The driving of the annular electromagnetic eddy current probe to scan and obtain the inner wall surface profile is as follows: While performing ultrasonic focused scanning on different wall thickness areas of the straight pipe section, the annular electromagnetic eddy current probe array is driven to continuously apply an alternating electromagnetic field; The eddy current response signal is converted into height or depth related to the surface topography through signal inversion, and then mapped to a three-dimensional spatial coordinate system to generate the inner wall surface profile.

7. The oil pipeline inner wall defect detection system according to claim 1, characterized in that: The fusion of eddy current and ultrasonic dual-modal data to generate a three-dimensional surface defect map is performed as follows: The inner wall surface profile and ultrasonic echo obtained by scanning different wall thickness areas are aligned in time and space; The defect area is identified and located from the aligned inner wall surface contour, and the defect opening boundary is extracted. At the same time, the depth distribution profile of the defect along the wall thickness direction is constructed by combining the diffraction time difference data from the ultrasonic echo; The two-dimensional defect boundary of the surface contour and the depth distribution profile are fused three-dimensionally to generate a three-dimensional surface defect map with dimension annotations.

8. The oil pipeline inner wall defect detection system according to claim 1, characterized in that: The high and low risk areas of sedimentation are divided into the following operations: A three-dimensional digital twin model containing the structural characteristics of the bend is constructed based on the pipeline's geometric parameters and material properties. A distributed strain sensor network is virtually deployed along the circumferential and axial directions of the bend in the model. Use the oil physical properties and transportation condition data collected during historical operations as boundary conditions and input them into the digital twin model to simulate oil transportation; During the simulation process, the virtual distributed strain sensor network is used to collect the equivalent stress distribution cloud diagram under the stress formation time series at each node of the bend section in real time; The equivalent stress distribution cloud map in the time series is clustered and analyzed frame by frame to divide it into several stress cluster areas; The ratio of the stress value of each stress cluster area to the average stress value of the adjacent areas is defined as the stress concentration index of the area; Classify and judge the stress concentration index according to the set warning threshold: If the stress concentration of a cluster area is higher than the warning threshold, it is directly marked as a high-risk deposition area; For areas where stress concentration is lower than the warning threshold, the sliding time window is used to analyze the stress concentration change trend within multiple consecutive time windows. If the stress concentration in the area is found to be increasing, it is marked as a high-risk deposition area. Otherwise, it is classified as a low-risk deposition area.

9. The oil pipeline inner wall defect detection system according to claim 1, characterized in that: The dual-frequency ultrasonic scanning is triggered in the high-risk deposition area, and the defect identification is performed based on the time-frequency separation of the deposition. The inspection robot is controlled to carry a dual-frequency ultrasonic probe and move to the high-risk area of ​​deposition. The probe synchronously transmits two sets of high-frequency and low-frequency ultrasonic pulse signals to the area to be tested, and receives the composite echo signal generated by the interaction of the pipeline structure and the deposition layer. The thickness of the sediment layer is estimated based on the time delay of high and low frequencies of the composite echo signal, and the composite echo signal is separated from the sediment interference and then the low-frequency effective echo component is focused; The low-frequency effective echo component is compared with the normal echo curve of the deposition-free area. By identifying the differences in echo amplitude, waveform, and propagation time, and combining the diffraction time difference, the defect location is performed, and the identified defect area is mapped to the three-dimensional geometric model of the bent pipe section.

10. The oil pipeline inner wall defect detection system according to claim 9, characterized in that: The method of triggering a single-frequency ultrasonic scan in a low-risk deposition area and identifying defects based on echo anomalies is implemented as follows: Control the inspection robot to carry the dual-frequency ultrasonic probe to the low-risk area of ​​deposition. The probe transmits low-frequency ultrasonic pulse signals to the area to be tested and receives reflected echo signals. The reflected echo signal is converted into a time-domain waveform and compared with the normal echo curve of the deposition-free area. By identifying the differences in echo amplitude, waveform, and propagation time, and combining it with the diffraction time difference, the defect is located. The identified defect area is then mapped to the 3D geometric model of the elbow section.

Images