A method for detecting and classifying risks in a pipeline of a petroleum transport network

Abstract

The present application relates to the technical field of pipeline detection, and particularly relates to a pipeline risk detection grading method for oil transportation pipe network. The present application realizes accurate detection and grading of pipeline risks of the oil transportation pipe network through initial detection and re-detection of unmanned aerial vehicle cooperative operation, multi-source data fusion identification of sediments, abnormal section risk quantitative evaluation and dynamic grading of pipe network level risks. The present application can not only effectively identify local risks of the pipeline caused by sediments, but also can evaluate the conduction and cumulative effect of risks by combining the physical connection relationship and spatial layout between the pipeline and the valve and the operating state of the valve, and finally provide intuitive basis for subsequent operation and maintenance decisions through a visual risk level diagram.

Description

Technical Field

[0001] This invention relates to the field of pipeline inspection technology, specifically to a method for risk detection and classification of oil transportation pipeline networks. Background Technology

[0002] As a critical infrastructure for energy transportation, oil pipelines are prone to deposit formation on their inner walls due to the deposition of oil components and the accumulation of impurities during long-term operation. The formation of deposits reduces the effective flow area of ​​the pipeline, increases energy consumption, and can lead to localized blockages or even leaks. Therefore, timely and accurate detection and classification of pipeline deposit risks are essential; risk detection is a crucial step in ensuring the safe and efficient operation of the pipeline network.

[0003] Existing technologies, such as Chinese invention patent publication number CN118261412A, disclose a method for risk analysis, assessment and display of oil and gas field pipelines. This method calculates the probability of failure by comprehensively considering indicators such as service life, medium water content, and damage records, and marks high-risk points on a GIS map.

[0004] For example, Chinese invention patent CN114037226B discloses a method for analyzing the risk probability level of pipelines in the petrochemical industry, which calculates the risk probability level based on factors such as corrosion rate, continuous uninspected time, expected thinning rate of the next major overhaul, and cracking history.

[0005] However, after in-depth analysis, the above-mentioned existing technologies still have the following limitations: 1. Risk factors mainly come from periodic detection or manual entry, lacking active real-time perception of sediments, and the assessment process does not integrate the spatial geometric characteristics and thermodynamic response of sediments, resulting in insufficient accuracy of risk quantification; 2. Risk level determination is isolated from the pipeline network operation status, failing to associate abnormal locations with valve operation status, and cannot reflect the transmission and accumulation effects of risks in the pipeline network. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and to achieve comprehensive detection of risks in oil transportation pipeline networks by combining initial inspection with re-inspection, and by comprehensively utilizing three types of data: lidar point cloud, thermal infrared imaging, and hyperspectral data.

[0007] The technical solution adopted by the present invention to solve its technical problem is: a risk detection and classification method for oil transportation pipeline network, including the following steps: deploying preliminary inspection and re-inspection drones to fly alternately along the pipeline; during the flight, the preliminary inspection drone collects lidar point cloud, thermal infrared imaging and hyperspectral data, identifies abnormal points through data fusion and sends their coordinates to the re-inspection drone.

[0008] After the re-inspection drone flew to the anomaly point, it re-collected lidar point cloud, thermal infrared imaging, and hyperspectral data, and identified the sediments accordingly.

[0009] Spatial clustering of sediments generates continuous anomalous segments; the longitudinal length, maximum thickness, surface temperature difference, and friction distance from the center of the anomalous segments to adjacent upstream and downstream valves are obtained, and the initial sedimentation risk level is obtained by fusion analysis based on these.

[0010] The initial deposition risk level corresponding to the initial inspection and re-inspection is compared with the preset threshold to determine the approved deposition risk level.

[0011] The associated risk level is obtained by combining the approved deposition risk level of all abnormal segments associated with each valve with the current valve opening level.

[0012] Risk levels are determined based on all approved deposition risk levels and associated risk levels, and a pipeline risk level map is generated.

[0013] Compared to existing technologies, this invention offers the following advantages: By employing collaborative operations between initial and follow-up inspection drones, multi-source data fusion for sediment identification, quantitative risk assessment of abnormal sections, and dynamic risk grading at the pipeline network level, this invention achieves accurate detection and grading of pipeline risks in oil transportation networks. This invention not only effectively identifies localized pipeline risks caused by sediments but also assesses the transmission and cumulative effects of risks by combining the physical connections and spatial layout between pipelines and valves, as well as the valve operating status. Ultimately, a visualized risk level map provides an intuitive basis for subsequent operation and maintenance decisions. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a schematic diagram of the detection and grading method of the present invention.

[0016] Figure 2 This is a schematic diagram of the process for identifying sediments according to the present invention.

[0017] Figure 3 This is a schematic diagram of the process for determining the approved deposition risk level in this invention. Detailed Implementation

[0018] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention. Furthermore, it should be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale.

[0019] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the invention or its application or use. Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification.

[0020] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0022] The following description, in conjunction with the accompanying drawings, details the specific scheme of the risk detection and classification method for oil transportation pipelines provided by this invention.

[0023] The core of this invention lies in: through a collaborative strategy of division of labor between initial inspection and re-inspection, comprehensively utilizing three types of data—LiDAR point cloud, thermal infrared imaging, and hyperspectral data—to achieve comprehensive detection of risks in oil transportation pipeline networks.

[0024] Those skilled in the art should understand that differences in vegetation cover may exist in pipeline laying environments. These differences directly affect the validity of data collected during the initial inspection phase. For example, in areas with high vegetation cover, analyzing the normalized difference in vegetation index based on hyperspectral data is an effective means of identifying component anomalies; while in areas with virtually no vegetation cover, the effectiveness of hyperspectral data will be reduced, and the initial inspection phase should focus on thermal infrared imaging data to identify thermal anomalies.

[0025] Therefore, based on the different pipeline laying environments, those skilled in the art can adaptively adjust the focus of utilization and data processing flow of the above three types of data according to the specific pipeline laying environment. Regardless of the initial inspection process, the ultimate goal is to efficiently locate anomalies and trigger subsequent re-inspection and risk assessment processes based on lidar point cloud, thermal infrared imaging, and hyperspectral data fusion.

[0026] Please see Figure 1The invention illustrates a risk detection and classification method for oil pipeline networks, which includes the following steps: Step S1, deploying preliminary inspection and re-inspection drones to fly alternately along the pipeline; thereby achieving a division of labor and cooperation between preliminary screening and accurate verification, ensuring that the re-inspection drones can respond promptly to anomalies found in the preliminary inspection stage, and improving detection efficiency.

[0027] Specifically, the initial inspection drone is responsible for rapidly scanning the pipeline area, collecting lidar point cloud data, thermal infrared imaging, and hyperspectral data, while the follow-up inspection drone follows at a certain distance. In this invention, this distance can be in the range of [500, 1000] meters.

[0028] Among them, lidar point cloud data is used to provide three-dimensional geometric information of pipelines and surrounding vegetation; thermal infrared imaging is used to reflect the surface temperature distribution, which helps to identify local thermal anomalies caused by pipeline deposition; and hyperspectral data can identify differences in surface material composition by analyzing reflectance in different bands.

[0029] In practice, anomalies are identified by fusing and analyzing the initial thermal infrared imaging and hyperspectral data. The specific identification process involves geometric registration and radiometric calibration of the initial thermal infrared imaging and hyperspectral data, converting them into surface reflectance and surface temperature data. This ensures spatial and radiometric consistency between data from different sensors.

[0030] Geometric registration refers to unifying the spatial coordinate system of data from different sensors to eliminate spatial deviations caused by differences in sensor position and orientation.

[0031] Radiometric calibration converts the raw digital values ​​recorded by the sensor into radiance, reflectivity, or temperature values ​​with physical units.

[0032] For pipelines laid in environments with vegetation cover, healthy vegetation has high reflectivity in the near-infrared band and low reflectivity in the red band. Pipeline leaks or sediment infiltration may alter soil composition, thereby affecting vegetation growth. If the health of vegetation declines, abnormalities can be indirectly detected by monitoring the Normalized Difference Vegetation Index (NDVI).

[0033] For example, when vegetation withers, it will show a decrease in near-infrared reflectance, an increase in red light reflectance, and a decrease in the normalized vegetation index (NDVI).

[0034] Therefore, after obtaining the surface reflectance data, the near-infrared reflectance is extracted from it. Red light band reflectivity And based on the center line of the pipeline, a certain distance is extended to both sides to construct a strip-shaped pre-set area.

[0035] Typically, the preset area width can be set to 20-50 meters. The specific setting needs to be determined based on the geological conditions of the pipeline network location. For example, in sandy soil areas with high permeability, pollutants are more likely to spread laterally, so the preset area width can be set to 50 meters; in clay areas with low permeability, the spread range of pollutants is limited, so the preset area width can be set to 20 meters.

[0036] Furthermore, the Normalized Difference Vegetation Index (NDVI) corresponding to each pixel within the preset area is calculated using the following formula: .

[0037] The NDVI value ranges from [-1, 1], and the NDVI of healthy vegetation is usually greater than 0.3.

[0038] Next, a clustering algorithm is used to divide all pixels in the preset area into at least two clusters; the pixels contained in the cluster with the smallest normalized vegetation index are aggregated into spectral anomaly areas, which correspond to the locations where vegetation growth is hindered or the surface material composition is abnormal; it should be noted that the clustering algorithm can be K-means or DBSCAN algorithm.

[0039] Meanwhile, along the pipeline's centerline, the pipeline can be evenly divided into multiple continuous segments at 15-meter intervals. Next, the arithmetic mean of the temperature values ​​of all thermal infrared pixels within each segment is calculated to obtain the average surface temperature. Then, according to the spatial order of the pipe segments, a temperature difference sequence is constructed based on the deviation between each average surface temperature and the average pipeline temperature.

[0040] Furthermore, anomaly detection algorithms are used to identify outlier pipe segments in the temperature difference sequence, and all pixels within the outlier pipe segments are aggregated into thermal anomaly zones.

[0041] Among these, the Z-score algorithm can be preferentially used for anomaly detection. Specifically, it involves calculating the mean μ and standard deviation σ of the temperature difference sequence for all pipe segments throughout the entire pipeline, and identifying pipe segments where |temperature difference - μ| > 2σ as outliers. The Z-score algorithm can then be used to identify these outliers, effectively identifying persistent thermal anomalies and eliminating transient interference.

[0042] When pipelines leak or experience severe sediment buildup, they typically cause anomalies in both surface vegetation composition and thermal radiation. Spectral anomalies may originate from hydrocarbons inhibiting plant growth or polluting the soil, while thermal anomalies may stem from sediments altering thermal conductivity. After identifying the spectral and thermal anomalies, dual verification using both is necessary to improve the accuracy of anomaly identification.

[0043] The specific method involves performing a Boolean intersection operation on the spectral anomaly region and the thermal anomaly region under a unified geographic coordinate system, and selecting the geographically overlapping area as the anomaly intersection region. The specific process of the Boolean intersection operation is already existing technology and will not be elaborated here.

[0044] Finally, the arithmetic mean of the geographic coordinates of all pixels within each intersection region of anomalies is calculated as the coordinates of the anomaly points.

[0045] It should be noted that this applies to pipeline installations in environments where there is virtually no vegetation cover.

[0046] In the initial inspection phase, the NDVI-based spectral anomaly identification step can be omitted. Thermal anomaly areas identified directly from thermal infrared imaging data can be used as the basis for locating anomaly points. In this case, the thermal anomaly area may directly originate from abnormal pipe wall temperature in the installed pipeline, or abnormal surface heat conduction due to sediment / leakage in buried pipelines. The re-inspection and subsequent procedures remain unchanged.

[0047] Step S2: After identifying the anomaly point in step S1, the initial inspection drone will send the coordinates of the anomaly point to the re-inspection drone. The re-inspection drone receives the coordinates, flies to the anomaly point, and re-collects lidar point cloud, thermal infrared imaging, and hyperspectral data.

[0048] It should be added that after the re-inspection drone arrives at the anomaly point, its flight altitude usually needs to be reduced to 30-50 meters and its flight speed reduced to improve data resolution and ensure the accuracy of subsequent identification.

[0049] After reacquiring lidar point cloud, thermal infrared imaging, and hyperspectral data, these three types of data need to be geographically aligned to ensure they correspond to the same object. Then, sediment identification can proceed.

[0050] Please see Figure 2 The specific identification process is as follows: Step S20: Since sediments often form local uplifts, causing elevation anomalies, a spherical neighborhood with a radius of 0.5 meters is constructed centered on each point based on the elevation coordinates of each point in the lidar point cloud data. However, it should be noted that the radius of this neighborhood is not a fixed value and can be adjusted according to the point cloud density; for example, the radius of the spherical neighborhood can be appropriately reduced in dense point clouds.

[0051] Then, calculate the difference between the elevation of each point and the average elevation of all points in its neighborhood. Sort the points by difference from largest to smallest, and aggregate the points with the highest difference to form a geometric anomaly region.

[0052] The first preset proportion can be determined as follows: in an area with known sediments, the distribution of outliers relative to the total number of points in the area is statistically analyzed, and the 95th percentile of this distribution can be taken as a reference value for the first preset proportion. This reference value is typically between 8% and 12%. In a preferred embodiment of the present invention, this reference value can be set to 10%.

[0053] Step S21: Since the thermal conductivity of sediments differs from that of soil and pipes, their surface temperature exhibits abnormalities during diurnal or seasonal variations. Therefore, based on the temperature value of each pixel in the thermal infrared imaging data, the difference between the value and the average temperature of each pixel in the neighborhood is calculated. The pixels are then sorted from largest to smallest difference, and the pixels with the highest sorting ratio are aggregated into a thermal anomaly region.

[0054] In this invention, the second preset ratio can be exemplarily set at 13%. When the ambient thermal noise is high, the second preset ratio can be increased to 15%-18% to reduce the risk of missed detection.

[0055] Step S22: Based on the reflectance of each pixel in the hyperspectral data in a specific characteristic band, calculate the spectral angle between the pixel and the average reflectance of each pixel in the neighborhood, sort the pixels by spectral angle from largest to smallest, and aggregate the pixels with the third preset ratio at the top of the sort into a component anomalous region.

[0056] Among them, the specific characteristic bands refer to the CH bond absorption characteristic bands unique to petroleum hydrocarbons, mainly including 1720nm±20nm and 2310nm±20nm. Hyperspectral imaging can distinguish petroleum hydrocarbons from soil and vegetation.

[0057] The neighborhoods mentioned in steps S22 and S21 can both be constructed using the same method as in step S20. However, the difference lies in the fact that the neighborhood radius in step S21 can be dynamically adjusted according to the resolution of the thermal infrared image, increasing as the resolution decreases. Similarly, the neighborhood radius in step S22 can be dynamically adjusted according to the resolution of the hyperspectral image, also increasing as the resolution decreases.

[0058] In this invention, a third preset ratio can be set in conjunction with a clustering algorithm. Specifically, the spectral angles of all pixels and their neighboring average reflectance are calculated to form a spectral angle distribution. Subsequently, the Otsu algorithm can be used to automatically segment the spectral angle distribution, dividing the pixels into two categories: background and potential anomalies. The percentage of potential anomaly pixels out of the total number of pixels is the third preset ratio. In this invention, for example, the third preset ratio can be 10%.

[0059] Step S23: Using the same Boolean intersection operation, the geometric anomaly region, thermal anomaly region, and compositional anomaly region are spatially superimposed, and the geographically overlapping area of ​​the three is selected as the sediment.

[0060] Step S3: Spatial clustering of the identified sediments to generate continuous anomalous segments.

[0061] This is done because sediments are often not isolated points, but rather distributed continuously or in clusters. Clustering can combine adjacent sediments into anomalous segments, facilitating an accurate assessment of the overall risk of the sediments.

[0062] The specific procedure is as follows: First, the three-dimensional spatial coordinates of each identified sediment are vertically projected onto the centerline of the pipeline to obtain the projected coordinates. Then, the curvilinear distance along the centerline between any two sediment projected coordinates is calculated.

[0063] Then, sediments whose curve distances to each other are less than the preset clustering radius are grouped into the same sediment cluster; the pipe segment between the minimum and maximum values ​​of all projected coordinates within each sediment cluster is defined as a continuous anomalous segment.

[0064] The determination of the preset cluster radius needs to consider the continuity of sediment distribution in the pipe. In this invention, the preset cluster radius can be exemplarily set to 10 meters. When the pipe diameter is large or the sediment distribution is relatively dispersed, the preset cluster radius can be larger, for example, 15 meters; when the pipe diameter is small or the sediment distribution is concentrated, the preset cluster radius can be smaller, for example, 5 meters.

[0065] Next, obtain the longitudinal length, maximum thickness, surface temperature difference, and the distance from the center of the abnormal section to the adjacent upstream and downstream valves: Step S30, based on the projection coordinates of all the sediments contained in the abnormal section, determine the starting point and ending point along the center line of the pipeline, and take the path length between the two as the longitudinal length.

[0066] Step S31: Based on the lidar point cloud data corresponding to the abnormal segment, calculate the vertical distance from each point in the sediment point cloud to the standard cylindrical surface of the pipe; take the maximum value among all vertical distances as the maximum thickness.

[0067] The calculation formula is: .

[0068] Among them, the standard cylindrical surface of the pipeline refers to the ideal outer surface constructed from the pipeline design GIS data, and its radius is equal to the nominal radius; that is, R is the nominal radius.

[0069] P is a point in the lidar point cloud data; Q is the nearest point from point P to the centerline of the pipeline; d is the vertical distance.

[0070] If d > 0, it means that point P is outside the pipe; if d < 0, it means that point P is inside the pipe.

[0071] Step S32: Based on the thermal infrared imaging data corresponding to the abnormal section, calculate the difference between its surface temperature and the average surface temperature of the adjacent upstream normal pipe section, and use it as the surface temperature difference.

[0072] The normal pipe section refers to a pipe section that is 50-100 meters away from the starting point of the current abnormal section along the pipeline centerline, and that was not identified as a thermal anomaly or spectral anomaly zone in the initial inspection. The surface temperature of this pipe section is taken as the arithmetic mean of the temperature values ​​of all pixels within it.

[0073] Step S33: Based on the start and end points of the abnormal segment, calculate the midpoint between them on the pipeline centerline, and use it as the center of the abnormal segment; using the center of the abnormal segment as the reference point, search upstream and downstream along the pipeline centerline until the adjacent upstream valve and downstream valve are located.

[0074] Obtain the positions of the upstream and downstream valves, and calculate the total path length along the pipeline centerline between each valve and the center of the abnormal section to obtain the corresponding friction distance.

[0075] Furthermore, considering that longitudinal length reflects the deposition range, maximum thickness reflects the severity of deposition, surface temperature difference reflects the intensity of thermal anomalies, and distance along the path reflects that: the closer to the valve, the relatively controllable risk; the farther from the valve, the larger the potential impact range and the higher the risk potential. This invention obtains the initial deposition risk level based on the integrated analysis of these four factors, thereby achieving a comprehensive assessment of deposition risk and improving the accuracy of the assessment.

[0076] The specific analysis process is as follows: select the minimum value of the friction distance between the upstream valve and the downstream valve corresponding to the abnormal segment as the critical impact distance.

[0077] The key influence distance, longitudinal length, maximum thickness and surface temperature difference are normalized to the same numerical range [0,1].

[0078] The difference between the upper limit of the numerical range and the normalized critical impact distance is calculated as the near-valve impact index. This transforms the perception that a smaller critical impact distance indicates higher risk into a larger near-valve impact index indicating higher risk; in other words, it converts distance into a positive impact indicator.

[0079] Further impact on the near valve index The initial deposition risk level is obtained by weighting and fusing the normalized longitudinal length L, maximum thickness T, and surface temperature difference Δt. .

[0080] The specific formula is as follows: .

[0081] in, , , The weight coefficients are all greater than zero and satisfy the following conditions: + + =1. Given that L×T directly relates to the total loss of the flow cross-section and the difficulty of pipeline cleaning, and is a major cause of outages, abnormal pressure drops, and even pipeline deformation, it needs to be given a high weight. ≥0.3.

[0082] In this invention, by example, =0.5、 =0.2、 =0.3.

[0083] +L can comprehensively characterize the macroscopic impact of the anomaly segment in terms of both spatial location and spatial scale. Even if the deposition is very long, if it is close to the valve, i.e. If the deposit is large, the overall risk is controllable; conversely, if the deposit is short but far from the valve, the overall risk remains high.

[0084] Δt reflects the degree of thermal anomaly caused by differences in the thermal conductivity of sediments. T+Δt can comprehensively characterize the influence of the physical and thermodynamic states of the anomaly segment itself. For example, an anomaly segment that is both thick and has a large surface temperature difference is inherently very risky. If only thickness is considered, developing thin layers of sediment may be overlooked; if only surface temperature difference is considered, transient disturbances may be misjudged.

[0085] When both L and T are large, it signifies a significant loss of flow cross-section in the pipeline, leading to blockages, jams, and abnormal pressure. To comprehensively reflect the combined effects of L and T, therefore... The product term L×T between the two is introduced in the calculation.

[0086] Please see Figure 3 Step S4: Compare the initial deposition risk level corresponding to the initial inspection and re-inspection with the preset threshold to determine the approved deposition risk level.

[0087] The purpose of initial screening is to screen for potential risks with high sensitivity, but false anomalies may occur during this process. To balance detection sensitivity and result reliability, this invention employs a collaborative approach of initial screening and re-screening. Re-screening is used to accurately confirm anomalies identified in the initial screening, thereby improving the accuracy of the overall risk assessment.

[0088] Therefore, after analyzing the data collected during the re-inspection to obtain the corresponding initial deposition risk level, based on the initial inspection data, the initial deposition risk level corresponding to the initial inspection of the same abnormal segment can be obtained in the same way, according to the process of obtaining the initial deposition risk level.

[0089] Then, the initial deposition risk level corresponding to the initial inspection and the re-inspection is compared with the preset threshold: if the initial deposition risk level of the re-inspection is greater than or equal to the preset threshold, the initial deposition risk level calculated based on the higher precision data source collected in the re-inspection has higher credibility, so it is used as the approved deposition risk level.

[0090] If the initial deposition risk level of the re-inspection is less than the preset threshold, but the initial deposition risk level of the initial inspection is greater than or equal to the preset threshold, then, in order to avoid the existence of potential risks that cannot be ignored, the smaller of the two is selected as the approved deposition risk level.

[0091] If the initial deposition risk levels of both the initial inspection and the re-inspection are less than the preset threshold, then both the initial inspection and the re-inspection indicate low risk, and the more accurate initial deposition risk level of the re-inspection is adopted as the approved deposition risk level.

[0092] It should be noted that the initial deposition risk level can be determined from step S33. It is obtained by weighted fusion of normalized variables; its ideal minimum value is 0, when When =L=T=Δt=1, assume =0.5、 =0.2、 =0.3, at which point the ideal maximum value is 1.5; therefore The value range is [0, 1.5]. Within this range, a reasonable value range for the preset threshold is [0.5, 1.1].

[0093] Step S5: Based on the approved deposition risk level of all abnormal segments associated with each valve, and combined with the current valve opening, obtain the associated risk level.

[0094] Given that oil pipeline networks are hydraulic systems, risks have a transmission and cumulative effect. Risk assessment of a single abnormal segment is localized and insufficient to reflect the overall condition of pipeline segments between adjacent valves.

[0095] Meanwhile, as control elements of the pipeline network, the opening degree of valves directly determines the flow rate and pressure of the associated pipeline sections. Pipeline sections with larger opening degrees typically carry larger loads, and under the same risk of sediment accumulation, their operational risks and potential impact on downstream areas are also higher.

[0096] Therefore, this invention introduces a correlation risk level, which combines static sediment physical characteristics with dynamic pipeline operation status to achieve more comprehensive pipeline risk detection.

[0097] First, identify all abnormal segments on the pipe section between any valve and its downstream adjacent valve, and sum the approved deposition risk of all abnormal segments to obtain the cumulative deposition risk of the valve.

[0098] Meanwhile, considering that valve opening affects flow rate, pressure, and sediment migration, and that a larger valve opening increases the likelihood of risk transmission, it is necessary to obtain the current opening of all valves on the monitored pipeline. Based on the maximum and minimum values, the opening of the valve to be evaluated is normalized, and the normalization result is used as the opening influence coefficient.

[0099] The specific process is as follows: Assuming the current valve opening is 0, the maximum value among all valve openings on the current detection pipeline is... The minimum value is Then the opening degree influence coefficient The calculation formula is: .

[0100] The opening influence coefficient reflects the relative activity of the valve in the pipeline network control. The larger the value, the higher the operating load of the pipeline section corresponding to the valve may be, and the greater the possibility of risk transmission.

[0101] Finally, the cumulative deposition risk level is multiplied by the corresponding aperture influence coefficient to obtain the associated risk level.

[0102] Step S6: Based on all approved deposition risk levels and associated risk levels, classify the risk levels and generate a pipeline risk level map.

[0103] A risk matrix is ​​constructed with the approved deposition risk level as the horizontal axis and the associated risk level as the vertical axis; the risk matrix is ​​then divided into several risk level regions; for example, it is divided into four risk level regions: low, medium, high, and extremely high.

[0104] The approved deposition risk level of each abnormal segment and the associated risk level of each valve are located in the risk matrix. Based on the located risk level area, the final risk level of each abnormal segment and each valve is determined.

[0105] The location of each abnormal section and valve, along with its corresponding final risk level, is marked on the pipeline network map to generate a pipeline network risk level map. This risk level map provides direct and comprehensive visual information for subsequent maintenance, repair, and scheduling decisions.

[0106] Among them, abnormal segments can be drawn along the center line of the pipeline in the form of colored line segments, with the color corresponding to its final risk level. For example, green represents low risk, yellow represents medium risk, orange represents high risk, and red represents extremely high risk.

[0107] Valves can be represented by icons with borders. The border color of the icon should match its final risk level, and the color can be different from or the same as the risk color corresponding to the abnormal segment. The icon's current opening degree is displayed inside.

[0108] In summary, this invention achieves accurate detection and classification of pipeline risks in oil transportation networks through collaborative operations of initial and follow-up inspection drones, multi-source data fusion for sediment identification, quantitative risk assessment of abnormal sections, and dynamic risk grading at the pipeline network level. This invention not only effectively identifies localized pipeline risks caused by sediments but also assesses the transmission and cumulative effects of risks by combining the physical connections and spatial layout between pipelines and valves, as well as valve operating status. Finally, it provides an intuitive basis for subsequent operation and maintenance decisions through a visualized risk level map.

[0109] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0110] Those skilled in the art will recognize that the modules and algorithm steps of the various examples 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 implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0111] In addition, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0112] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0113] 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 within the protection scope of the present invention.

Claims

1. A method for risk detection and classification of oil transportation pipeline networks, characterized in that, Includes the following steps: Initial inspection and re-inspection drones are deployed to fly alternately along the pipeline; during the flight, the initial inspection drone collects lidar point cloud, thermal infrared imaging and hyperspectral data, identifies abnormal points through data fusion and sends their coordinates to the re-inspection drone; After the re-inspection drone flew to the anomaly point, it re-collected lidar point cloud, thermal infrared imaging and hyperspectral data, and identified the sediments accordingly; Spatial clustering of sediments generates continuous anomalous segments; the longitudinal length, maximum thickness, surface temperature difference, and friction distance from the center of the anomalous segments to adjacent upstream and downstream valves are obtained, and the initial sedimentation risk level is obtained by fusion analysis based on these data. If the initial deposition risk level of the re-inspection is greater than or equal to the preset threshold, or if the initial deposition risk levels of both the initial inspection and the re-inspection are less than the preset threshold, then the initial deposition risk level of the re-inspection will be used as the approved deposition risk level. If the initial deposition risk level of the re-inspection is less than the preset threshold, but the initial deposition risk level of the initial inspection is greater than or equal to the preset threshold, then the smaller of the two will be used as the approved deposition risk level. Accumulate the approved deposition risk of all abnormal segments associated with each valve, and normalize the opening of the valve to be evaluated based on the maximum and minimum values ​​of the current opening of all valves on the current detection pipeline. The normalized result is multiplied by the sum of the approved depositional risk levels to obtain the associated risk level; Risk levels are determined based on all approved deposition risk levels and associated risk levels, and a pipeline risk level map is generated.

2. The method for risk detection and classification of oil transportation pipelines according to claim 1, characterized in that, The process for identifying the anomalies is as follows: Geometric registration and radiometric calibration were performed on the initial thermal infrared imaging and hyperspectral data, which were then converted into surface reflectance and surface temperature data. Based on surface reflectance data, the reflectance of near-infrared band and red band is extracted, and the normalized vegetation index corresponding to each pixel in the preset area on both sides of the pipeline centerline is calculated. The clustering algorithm divides all pixels in the preset area into at least two clusters; the pixels contained in the cluster with the smallest normalized vegetation index are aggregated into a spectral anomaly area. The pipeline is divided into multiple continuous segments at equal intervals along its centerline. The average surface temperature of each segment is extracted, and a temperature difference sequence is generated according to the spatial location of the segments. Anomaly detection algorithms are used to identify outlier pipe segments in the temperature difference sequence, and all pixels within the outlier pipe segments are aggregated into thermal anomaly zones. Spatial superposition of spectral anomaly regions and thermal anomaly regions is performed, and the geographically overlapping area between the two is selected as the anomaly intersection region. Calculate the arithmetic mean of the geographic coordinates of all pixels within the intersection area of ​​each anomaly, and use it as the coordinates of the anomaly point.

3. The method for risk detection and classification of oil transportation pipelines according to claim 1, characterized in that, The process for identifying the sediments is as follows: The lidar, thermal infrared imaging, and hyperspectral data re-examined at the anomaly points are aligned with geographic coordinates. Based on the elevation coordinates of each point in the lidar point cloud data, calculate the difference between the elevation coordinates of each point and the average elevation of each point in the neighborhood, sort them from largest to smallest, and aggregate the points with the highest sorting ratio into a geometric anomaly region. Based on the temperature value of each pixel in the thermal infrared imaging data, the difference between the temperature value of each pixel and the average temperature value of each pixel in the neighborhood is calculated. The pixels are sorted from largest to smallest, and the pixels with the highest sorting ratio are aggregated into thermal anomaly regions. Based on the reflectance of each pixel in a specific characteristic band in the hyperspectral data, the spectral angle between the pixel and the average reflectance of each pixel in the neighborhood is calculated. The pixels are sorted from largest to smallest according to the spectral angle, and the pixels with the highest third-preset ratio are aggregated into a component anomaly region. Geometric anomaly regions, thermal anomaly regions, and compositional anomaly regions are spatially superimposed, and the geographically overlapping areas of the three are selected as sediments.

4. The method for risk detection and classification of oil transportation pipelines according to claim 1, characterized in that, The process of spatially clustering sediments to generate continuous anomalous segments is as follows: The three-dimensional spatial coordinates of each identified sediment are vertically projected onto the centerline of the pipe to obtain the projected coordinates; Calculate the curvilinear distance along the pipe centerline between any two sediment projection coordinates; Sediments whose curve distance is less than the preset clustering radius are grouped into the same sediment cluster; The pipe segment between the minimum and maximum values ​​of all projected coordinates within each sediment cluster is defined as a continuous anomalous segment.

5. The method for risk detection and classification of oil transportation pipelines according to claim 1, characterized in that, The process of obtaining the longitudinal length, maximum thickness, and surface temperature difference of the abnormal segment is as follows: Based on the projected coordinates of all sediments contained in the anomaly segment, determine its starting point and ending point along the pipeline centerline, and take the path length between the two as the longitudinal length. Based on the lidar point cloud data corresponding to the anomaly segment, the vertical distance from each point in the sediment point cloud to the standard cylindrical surface of the pipe is calculated; the maximum value among all vertical distances is taken as the maximum thickness. Based on the thermal infrared imaging data corresponding to the abnormal section, the difference between its surface temperature and the average surface temperature of the adjacent upstream normal pipe section is calculated as the surface temperature difference.

6. The method for risk detection and classification of oil transportation pipelines according to claim 1, characterized in that, Calculate the friction distance from the center of the abnormal section to the adjacent upstream and downstream valves, specifically: Based on the start and end points of the abnormal segment, calculate the midpoint between them on the pipeline centerline, and use it as the center of the abnormal segment. Using the center of the abnormal section as a reference point, search upstream and downstream along the pipeline centerline until the adjacent upstream and downstream valves are located. Obtain the positions of the upstream and downstream valves, and calculate the total path length along the pipeline centerline between each valve and the center of the abnormal section to obtain the corresponding friction distance.

7. The method for risk detection and classification of oil transportation pipelines according to claim 6, characterized in that, The process for obtaining the initial deposition risk level is as follows: The minimum distance between the upstream and downstream valves corresponding to the abnormal section is selected as the critical impact distance. The key influencing factors, such as distance, longitudinal length, maximum thickness, and surface temperature difference, are normalized to the same range of values. The difference between the upper limit of the numerical interval and the normalized critical influence distance is calculated as the near-valve influence index. The initial deposition risk level is obtained by weighting and fusing the near-valve influence index, normalized longitudinal length, maximum thickness, and surface temperature difference.

8. The method for risk detection and classification of oil transportation pipelines according to claim 1, characterized in that, Based on the process of obtaining the initial deposition risk level corresponding to the re-inspection, the initial deposition risk level corresponding to the initial inspection of the same abnormal segment can be obtained similarly based on the data collected in the initial inspection.

9. The method for risk detection and classification of oil transportation pipelines according to claim 4, characterized in that, The approved deposition risk of all abnormal segments associated with each valve is accumulated. Specifically, this involves identifying all abnormal segments on the pipe section between any valve and its downstream adjacent valve, and accumulating the approved deposition risk of all abnormal segments.

10. The method for risk detection and classification of oil transportation pipelines according to claim 1, characterized in that, The process of generating the pipeline risk level map is as follows: A risk matrix is ​​constructed with the approved deposition risk level as the horizontal axis and the associated risk level as the vertical axis; The risk matrix is ​​then divided into several risk level zones. The approved deposition risk level of each abnormal segment and the associated risk level of each valve are located in the risk matrix. Based on the located risk level area, the final risk level of each abnormal segment and each valve is determined respectively. Mark the location of each abnormal section and valve and its corresponding final risk level on the pipeline network map to generate a pipeline network risk level map.

Images