Optical fiber monitoring early warning optimization method and device

Abstract

The invention relates to an optical fiber monitoring early warning optimization method and device, and the method comprises the steps: constructing a pipeline soil texture database of an oil and gas pipeline, and determining the parameter optimization direction of each soil texture type; determining optical fiber monitoring parameters of each monitoring unit according to the parameter optimization direction and the soil optical fiber signal mapping model; based on the soil texture type of the to-be-monitored unit, determining a target parameter optimization direction, and collecting to-be-detected soil texture parameters and real-time signal characteristics of the to-be-monitored unit according to the target parameter optimization direction; performing optical fiber monitoring early warning on the real-time signal characteristics according to the optical fiber monitoring parameters to obtain an early warning result; when the early warning result is that the risk exists, obtaining an early warning report according to the to-be-detected soil parameters and the soil optical fiber signal mapping model; by constructing the pipeline soil property database and optimizing the monitoring parameters based on the soil property types, more accurate risk early warning is realized, and the method has the advantages of improving the optical fiber monitoring accuracy and reducing the false alarm rate and the omission ratio.

Description

Technical Field

[0001] This invention relates to the field of oil and gas pipeline safety monitoring technology, and in particular to an optimized method and device for fiber optic monitoring and early warning. Background Technology

[0002] In an era of rapid growth in energy demand, precise prevention and timely early warning via fiber optic cables have become crucial for pipeline safety management. Oil and gas pipelines, as core infrastructure for energy transportation, rely on efficient monitoring and early warning technologies for safe operation. Among existing technologies, distributed fiber optic sensing technology, with its advantages of real-time, long-distance, and all-weather monitoring, has been widely applied in risk warning scenarios such as pipeline leaks and third-party sabotage. However, existing fiber optic monitoring solutions generally ignore the impact of soil variations along the pipeline on monitoring effectiveness, leading to signal distortion and insufficient early warning accuracy.

[0003] Therefore, there is an urgent need to propose an optimized method and device for fiber optic monitoring and early warning to solve the technical problem that existing fiber optic monitoring schemes generally ignore the impact of soil differences along the pipeline on the monitoring effect, resulting in distorted monitoring signals and insufficient early warning accuracy. Summary of the Invention

[0004] In view of this, it is necessary to provide an optimized method and device for fiber optic monitoring and early warning, in order to solve the technical problem that existing fiber optic monitoring schemes generally ignore the impact of soil differences along the pipeline on the monitoring effect, resulting in distorted monitoring signals and insufficient early warning accuracy.

[0005] To address the aforementioned problems, in a first aspect, the present invention provides an optimized method for fiber optic monitoring and early warning, comprising: Construct a pipeline soil database containing soil parameters and fiber optic signal characteristics for all soil types in multiple monitoring units along the oil and gas pipeline, and determine the parameter optimization direction for each soil type. The fiber optic monitoring parameters of each monitoring unit are determined based on the parameter optimization direction and the soil fiber optic signal mapping model. Based on the soil type of the unit to be monitored, the optimization direction of the target parameters is determined, and the soil parameters to be detected and real-time signal characteristics of the unit to be monitored are collected according to the optimization direction of the target parameters. Based on the fiber optic monitoring parameters, fiber optic monitoring and early warning are performed on the real-time signal characteristics to obtain early warning results; When the warning result indicates that there is a risk, a warning report is obtained based on the soil parameters to be detected and the soil fiber optic signal mapping model.

[0006] In one possible implementation, the construction of the pipeline soil database, which includes soil parameters and fiber optic signal characteristics for all soil types, for oil and gas pipelines includes: Multiple monitoring units were delineated along the oil and gas pipeline, and multiple parallel soil samples were collected from the multiple monitoring units. Laboratory analysis was performed on the multiple parallel soil samples to determine the soil type and soil parameters of each monitoring unit; Based on the soil type and soil parameters, a pipeline soil database is constructed.

[0007] In one possible implementation, the step of defining multiple monitoring units along the oil and gas pipeline and collecting multiple parallel soil samples from the multiple monitoring units includes: Using the centerline of the oil and gas pipeline as a reference, the main areas of influence within the pipeline burial depth range are extended to both sides to obtain the sampling zone; Based on the pipeline design drawings and the topographic map along the route, the sampling zone is divided into monitoring units and soil types to obtain soil classification results. The soil classification results include monitoring units divided at a preset first interval in the regular section, monitoring units divided at a second interval smaller than the first interval in the special section, and monitoring units added at the boundaries of soil differences. The monitoring area corresponding to the monitoring unit includes areas with different soil types. The sampling zone is marked according to the pipeline design drawing, the topographic map along the route, and the soil classification results to obtain multiple sampling points; The latitude and longitude coordinates of multiple sampling points on all monitoring units are collected to construct a sampling table; the sampling table includes the latitude and longitude coordinates and the marker stake number of each sampling point; Multiple parallel soil samples were collected from each sampling point based on the latitude and longitude coordinates and the marker stake number.

[0008] In one possible implementation, the soil-fiber signal mapping model takes soil parameters as input and outputs fiber optic signal characteristics as output. The soil parameters include key soil parameters and auxiliary variable parameters. The key soil parameters include acoustic attenuation coefficient, strain transfer efficiency, natural moisture content, soil density, and elastic modulus. The auxiliary variable parameters include soil type. The fiber optic signal characteristics include acoustic sensing core characteristics and temperature sensing auxiliary characteristics. The acoustic sensing core characteristics include vibration signal amplitude, vibration signal dominant frequency, and signal-to-noise ratio. The temperature sensing auxiliary characteristics include temperature change gradient and temperature settling time.

[0009] In one possible implementation, the training process of the soil-fiber signal mapping model includes: The system collects the first optical fiber signal of each monitoring unit under normal operating conditions, and the second optical fiber signal of each monitoring unit under different risk conditions. Abnormal data is removed from the first and second optical fiber signals to obtain the target optical fiber signal; The target optical fiber signal at the same time period is matched with the soil parameters in the pipeline soil database to obtain a complete sample; The soil parameters in the complete sample are normalized to obtain the target complete sample; The target complete sample is divided into a training set, a validation set, and a test set; The soil-fiber signal mapping model is trained using the training set, the validation set, and the test set to obtain the trained soil-fiber signal mapping model.

[0010] In one possible implementation, determining the fiber optic monitoring parameters of each monitoring unit based on the parameter optimization direction and the soil fiber optic signal mapping model includes: The soil parameters of each monitoring unit in the pipeline soil database are input into the soil fiber optic signal mapping model to obtain the theoretical signal characteristic threshold corresponding to each monitoring unit. Based on the parameter optimization direction and the theoretical signal characteristic threshold, the fiber optic monitoring parameters of the corresponding monitoring unit are generated and initialized; the fiber optic monitoring parameters include the sampling frequency and the early warning signal threshold.

[0011] In one possible implementation, the step of performing fiber optic monitoring and early warning on the real-time signal characteristics based on the fiber optic monitoring parameters to obtain an early warning result includes: Set the triggering warning conditions for each soil type; Determine whether the fiber optic monitoring parameters and the real-time signal characteristics meet the triggering early warning conditions corresponding to the soil type of the unit to be monitored; If so, then the monitored unit is determined to be at risk; If not, then it is determined that the unit to be monitored does not pose a risk.

[0012] In one possible implementation, the method further includes: The deviation rate is obtained by comparing the optical fiber monitoring parameters with the real-time signal characteristics. When the deviation rate is greater than a preset threshold, the fiber optic monitoring parameters of the monitoring unit are adjusted according to the soil parameters to be detected.

[0013] In one possible implementation, the method further includes: The fiber optic monitoring parameters of the multiple monitoring units are evaluated to obtain evaluation results; When the evaluation results show that the optimization effect is not good, the poor soil type in the monitoring unit corresponding to the poor optimization effect is determined. The soil parameters and fiber optic signal characteristics corresponding to the poor soil type are re-collected and added to the pipeline soil database; The soil fiber optic signal mapping model is retrained and its parameters are optimized by supplementing the pipeline soil database, resulting in updated fiber optic monitoring parameters for each monitoring unit.

[0014] Secondly, the present invention also provides an optical fiber monitoring and early warning optimization device, comprising: The database construction module is used to build a pipeline soil database containing soil parameters and fiber optic signal characteristics for all soil types in multiple monitoring units on oil and gas pipelines, and to determine the parameter optimization direction for each soil type. The parameter determination module is used to determine the fiber optic monitoring parameters of each monitoring unit based on the parameter optimization direction and the soil fiber optic signal mapping model. The parameter optimization module is used to determine the direction of target parameter optimization based on the soil type of the unit to be monitored, and to collect the soil parameters to be detected and real-time signal characteristics of the unit to be monitored according to the target parameter optimization direction. The result determination module is used to perform fiber optic monitoring and early warning on the real-time signal characteristics based on the fiber optic monitoring parameters, and obtain the early warning result. The report determination module is used to generate an early warning report based on the soil parameters to be detected and the soil fiber optic signal mapping model when the early warning result indicates that there is a risk.

[0015] The beneficial effects of this invention are as follows: A pipeline soil database is constructed, containing soil parameters and fiber optic signal characteristics for all soil types in multiple monitoring units along an oil and gas pipeline, and the parameter optimization direction for each soil type is determined. Based on the parameter optimization direction and the soil-fiber optic signal mapping model, the fiber optic monitoring parameters for each monitoring unit are determined. Based on the soil type of the unit to be monitored, the target parameter optimization direction is determined, and the soil parameters to be detected and real-time signal characteristics of the unit to be monitored are collected according to the target parameter optimization direction. Based on the fiber optic monitoring parameters, fiber optic monitoring and early warning are performed on the real-time signal characteristics to obtain early warning results. When the early warning result indicates a risk, an early warning report is obtained based on the soil parameters to be detected and the soil-fiber optic signal mapping model. By constructing a pipeline soil database and optimizing monitoring parameters based on soil types, more accurate risk early warning is achieved, which has the advantages of improving the accuracy of fiber optic monitoring and reducing false alarm and missed detection rates. Attached Figure Description

[0016] Figure 1 A schematic flowchart of an embodiment of the fiber optic monitoring and early warning optimization method provided by the present invention; Figure 2 A schematic diagram of an embodiment of the collection of multiple parallel soil samples provided by the present invention; Figure 3 A schematic diagram illustrating an embodiment of the training process for the soil-fiber signal mapping model provided by the present invention; Figure 4 This is a schematic diagram of an embodiment of the fiber optic monitoring and early warning optimization device provided by the present invention. Detailed Implementation

[0017] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0018] like Figure 1 As shown in the figure, a specific embodiment of the present invention discloses an optimization method for fiber optic monitoring and early warning, comprising: S101. Construct a pipeline soil database containing soil parameters and fiber optic signal characteristics for all soil types in multiple monitoring units on the oil and gas pipeline, and determine the parameter optimization direction for each soil type.

[0019] The monitoring unit refers to a specific geographical area with unique soil characteristics along the oil and gas pipeline. It serves as the basic unit for soil parameter acquisition, fiber optic signal monitoring, and parameter optimization. Soil parameters describe the physical, chemical, and mechanical properties of the soil, such as density, moisture content, elastic modulus, and acoustic attenuation coefficient. These parameters affect the propagation and response of fiber optic signals. Fiber optic signal characteristics refer to the specific attributes exhibited by signals acquired through distributed fiber optic sensing technology, such as vibration signal amplitude, dominant frequency, signal-to-noise ratio, and temperature gradient and settling time. These characteristics are used to determine if there are any anomalies in the pipeline. The pipeline soil database is a structured dataset that stores the soil type, soil parameters, and corresponding fiber optic signal characteristics under different operating conditions for each monitoring unit along the oil and gas pipeline, providing a data foundation for subsequent parameter optimization and early warning.

[0020] Based on soil classification results, this invention determines the parameter optimization direction for each soil type according to the signal propagation characteristics of different soil types. The specific logic is as follows: Sandy soil unit: Due to high porosity and fast signal attenuation, the optimization direction is to enhance the signal capture capability: it is necessary to increase the sampling frequency to reduce signal loss, increase the vibration signal threshold to avoid missed detection due to attenuation, and add signal gain compensation parameters. Clay soil unit: Due to large fluctuations in water content and a lot of noise interference, the optimization direction is to suppress noise and false alarms: it is necessary to reduce the sampling frequency to reduce noise acquisition, reduce the vibration signal threshold to capture weak effective signals, and enable the cross-validation threshold of temperature signal and vibration signal. Silty soil unit: Due to its uniform particle size distribution and signal characteristics between sandy soil and clay soil, the optimization direction is to balance capture and suppression: the sampling frequency and signal threshold are taken as the middle value, no additional gain or cross-validation is required, but the frequency of parameter stability verification needs to be increased. Frozen soil unit: Due to temperature changes, soil hardness can change drastically. The optimization direction is to dynamically adapt to different seasons: in winter, configure a high sampling frequency as "similar to sandy soil" and in spring, configure a low sampling frequency as "similar to clay soil". At the same time, add a temperature-triggered parameter switching threshold. Saline soil unit: Due to the high salt content, electrochemical interference is easily generated. The optimization direction is multi-signal fusion filtering: the sampling frequency is set to a medium value, the vibration threshold is slightly higher than that of silty soil, and an auxiliary threshold for electrochemical signal is added. When the electrochemical signal exceeds a certain value, the vibration threshold is automatically increased to filter the interference.

[0021] S102. Determine the fiber optic monitoring parameters for each monitoring unit based on the parameter optimization direction and the soil fiber optic signal mapping model.

[0022] The parameter optimization direction refers to parameter adjustment strategies or objectives set for different soil types to improve the effectiveness of fiber optic monitoring and early warning. For example, for a specific soil type, it may be necessary to increase the sampling frequency or adjust the early warning signal threshold. The soil-fiber optic signal mapping model can be a BP neural network model, which can capture the complex interaction between soil parameters and signal characteristics through nonlinear mapping of multiple layers of neurons. Based on a deep learning framework, the network layer is set to 3 layers (input layer + hidden layer × 2 + output layer) to avoid gradient vanishing caused by deep networks; the number of neurons in the input layer is consistent with the number of input variables; the number of neurons in the hidden layer is selected to balance fitting ability and computational cost, and the activation function is selected to solve the gradient vanishing problem; the number of neurons in the output layer is consistent with the number of output variables, and the activation function is selected to be suitable for continuous value output; the loss function is the mean squared error; the optimizer is the Adam optimizer, and the learning rate is selected with an initial reasonable value, which is then fine-tuned through the validation set. Fiber optic monitoring parameters refer to the specific settings used to guide the fiber optic monitoring system in data acquisition and early warning judgment, such as the sampling frequency of the fiber optic sensor and the trigger threshold of the early warning signal. These parameters will be adjusted according to the soil type and monitoring requirements.

[0023] S103. Based on the soil type of the unit to be monitored, determine the optimization direction of the target parameters, and collect the soil parameters to be monitored and real-time signal characteristics of the unit to be monitored according to the optimization direction of the target parameters.

[0024] Among them, real-time signal characteristics refer to the fiber optic signal attributes that are collected and extracted in real time by the fiber optic sensing system during the actual monitoring process. These attributes are compared with preset warning parameters to determine whether there are any abnormalities.

[0025] S104. Based on the fiber optic monitoring parameters, perform fiber optic monitoring and early warning of real-time signal characteristics to obtain early warning results.

[0026] Among them, the early warning result refers to the judgment made by the fiber optic monitoring system based on the comparison of real-time signal characteristics with fiber optic monitoring parameters, indicating whether there are potential risks or abnormalities in the unit to be monitored.

[0027] S105. When the warning result indicates that there is a risk, a warning report is obtained based on the soil parameters to be tested and the soil fiber optic signal mapping model.

[0028] Among them, the early warning report: When the early warning result indicates that there is a risk, the system generates a detailed report based on the soil parameters to be detected and the soil fiber optic signal mapping model, which may include information such as the risk type, location, possible causes and recommended disposal measures.

[0029] This invention constructs a pipeline soil database containing soil parameters and fiber optic signal characteristics for all soil types in multiple monitoring units along an oil and gas pipeline, and determines the parameter optimization direction for each soil type. In one implementation, the pipeline can be manually surveyed, and monitoring units can be divided based on experience. The soil in each monitoring unit can then be simply classified, and approximate soil parameters can be recorded, such as a rough soil type (e.g., sandy soil, clay) and visually estimated moisture content. Simultaneously, a preliminary parameter optimization direction can be set for each soil type based on past monitoring experience or industry standards; for example, for sandy soil areas, increasing the sensitivity of vibration signals might be preferable. In another implementation, existing Geographic Information System (GIS) data, combined with publicly available soil type maps, can be used to initially establish a database containing soil types and general fiber optic signal characteristics. The parameter optimization direction can then be based on an expert system, providing suggested monitoring parameter adjustment strategies according to different soil types.

[0030] The fiber optic monitoring parameters for each monitoring unit are determined based on the parameter optimization direction and the soil-fiber optic signal mapping model. Specifically, a rule-based mapping model can be used. This model directly retrieves the corresponding fiber optic monitoring parameters from a predefined lookup table based on the input soil type and the preset parameter optimization direction. For example, if the soil type of a monitoring unit is "wet clay" and the parameter optimization direction indicates "improving temperature sensitivity," the model will output a set of preset sampling frequencies and temperature warning thresholds. Alternatively, a simple statistical model, such as a linear regression model, can be constructed, using soil parameters as input and fiber optic signal characteristics as output, and trained using historical data. When determining the fiber optic monitoring parameters, this model can predict the theoretical signal characteristics based on the soil parameters, and then, combined with the parameter optimization direction, manually or through a simple algorithm, adjust the sampling frequency and warning signal thresholds to adapt to the predicted signal characteristics.

[0031] Based on the soil type of the unit to be monitored, the optimization direction of the target parameters is determined, and the soil parameters and real-time signal characteristics of the unit to be monitored are collected according to the optimization direction. Specifically, the operator can consult existing soil information based on the geographical location of the unit to be monitored, manually input the soil type of the unit, and determine a target parameter optimization direction based on experience or preset rules. Subsequently, conventional fiber optic sensing equipment can be used to collect the real-time fiber optic signal characteristics of the unit according to the sampling frequency and data acquisition mode indicated by the target parameter optimization direction, and preliminary soil parameter detection can be performed, such as measuring soil moisture using a portable sensor.

[0032] Based on the fiber optic monitoring parameters, the system performs fiber optic monitoring and early warning of the real-time signal characteristics, obtaining an early warning result. Specifically, the real-time acquired fiber optic signal characteristics (e.g., vibration signal amplitude) can be directly compared with the early warning signal threshold in the fiber optic monitoring parameters set by the current monitoring unit. If the real-time signal characteristics exceed the threshold, the system determines that there is a risk and generates an early warning result. When the early warning result indicates a risk, an early warning report is generated based on the soil parameters to be detected and the soil fiber optic signal mapping model. Specifically, after the system issues an early warning, the soil parameters to be detected acquired at the time of the warning can be input into the soil fiber optic signal mapping model. The model will generate an early warning report containing soil impact analysis based on these parameters and its internal logic. For example, the report may indicate that, under the current soil conditions, the abnormality of the fiber optic signal may be caused by soil subsidence, and provide corresponding signal characteristic data as support.

[0033] Compared with existing technologies, this embodiment provides a pipeline soil database that constructs soil parameters and fiber optic signal characteristics for all soil types in multiple monitoring units on an oil and gas pipeline, and determines the parameter optimization direction for each soil type; it determines the fiber optic monitoring parameters for each monitoring unit based on the parameter optimization direction and the soil fiber optic signal mapping model; it determines the target parameter optimization direction based on the soil type of the unit to be monitored, and collects the soil parameters to be detected and real-time signal characteristics of the unit to be monitored based on the target parameter optimization direction; it performs fiber optic monitoring and early warning on the real-time signal characteristics based on the fiber optic monitoring parameters, and obtains the early warning result; when the early warning result indicates that there is a risk, it obtains an early warning report based on the soil parameters to be detected and the soil fiber optic signal mapping model; by constructing a pipeline soil database and optimizing monitoring parameters based on soil types, more accurate risk early warning is achieved, which has the advantages of improving the accuracy of fiber optic monitoring and reducing the false alarm rate and missed detection rate.

[0034] In some embodiments of the present invention, step S101 includes: Multiple monitoring units were delineated along the oil and gas pipeline, and multiple parallel soil samples were collected from these monitoring units.

[0035] This step in the embodiment of the invention aims to systematically divide the area along the oil and gas pipeline and obtain representative soil samples. Delineating monitoring units is to break down long-distance oil and gas pipelines into manageable and analyzable segments, facilitating targeted data collection and management. Collecting parallel soil samples ensures the representativeness of the samples and the reliability of the analysis results, reducing the random errors of single samples. Within each monitoring unit, multiple representative points are selected for soil sampling, with multiple parallel soil samples collected at each point; for example, 2-3 soil samples are collected at the same or different depths. Within each monitoring unit, professional soil sampling tools (such as augers, soil samplers, etc.) are used to collect multiple parallel soil samples at preset sampling points, ensuring that the soil samples cover the main soil types within the monitoring unit.

[0036] Laboratory analysis was conducted on multiple parallel soil samples to determine the soil type and soil parameters for each monitoring unit.

[0037] This step in this embodiment of the invention involves conducting scientific and precise physical and chemical property tests on the collected soil samples to obtain detailed soil information. Through laboratory analysis, the soil classification (e.g., sandy soil, clay, loam) and its key parameters can be accurately identified. These parameters form the basis for constructing the pipeline soil database and are crucial for establishing a mapping relationship between subsequent fiber optic signal characteristics and soil parameters. As one possible implementation method, laboratory analysis may include, but is not limited to, particle size distribution analysis (determining the content of sand, silt, and clay particles), moisture content determination, density determination (dry density, wet density), void ratio determination, shear strength testing, compression testing, and permeability coefficient determination. Through these tests, the soil type (e.g., according to the Unified Soil Classification System (USCS)) and specific soil parameters of each monitoring unit can be determined. Simultaneously, by combining soil mechanics and geotechnical testing standards, the test results are processed and analyzed to ultimately determine the soil type and corresponding parameters of each monitoring unit.

[0038] A pipeline soil database is constructed based on soil type and soil parameters.

[0039] In this embodiment of the invention, this step involves integrating, storing, and managing the previously acquired structured data to form a database that can be queried, analyzed, and utilized. This database is the core data source of the entire fiber optic monitoring and early warning optimization method, providing fundamental data support for subsequent parameter optimization direction determination, fiber optic monitoring parameter setting, and early warning report generation.

[0040] This invention first divides the oil and gas pipeline route into multiple monitoring units. This detailed spatial division effectively addresses the diverse soil conditions along the pipeline. Subsequently, multiple parallel soil samples are collected within each monitoring unit, ensuring sample representativeness and data acquisition reliability, and avoiding errors that might arise from a single sampling point. Next, rigorous laboratory analysis is performed on these soil samples to accurately obtain the soil type and various soil parameters for each monitoring unit, transforming the actual soil physicochemical properties into quantifiable data. Finally, based on these accurate soil types and parameters, a pipeline soil database is constructed. This series of systematic steps ensures the comprehensiveness, accuracy, and reliability of the data in the database, providing a solid data foundation for subsequent fiber optic monitoring and early warning optimization methods. This allows for the determination of parameter optimization directions, the setting of fiber optic monitoring parameters, and the generation of early warning reports based on reliable soil information, thereby significantly improving the accuracy and effectiveness of the entire monitoring and early warning system.

[0041] In some embodiments of the present invention, such as Figure 2 As shown, multiple monitoring units were delineated along the oil and gas pipeline, and multiple parallel soil samples were collected from these monitoring units, including: S201. Using the centerline of the oil and gas pipeline as a reference, extend the sampling zone to both sides of the main influence area within the pipeline burial depth range.

[0042] This step aims to determine the effective width range for soil sampling. The main area of ​​influence can be determined through geological surveys, mechanical analysis, or empirical rules. For example, taking 5m as an example, a sampling zone 10m wide × the total length of the pipeline can be formed. If there are special sections in the pipeline, the sampling zone width can be further expanded to avoid missing special soil types.

[0043] S202. Based on the pipeline design drawings and the topographic map along the route, the sampling zone is divided into monitoring units and soil types to obtain soil classification results. The soil classification results include monitoring units divided at a preset first interval in the regular section, monitoring units divided at a second interval less than the first interval in the special section, and monitoring units added at the boundary of soil difference. The monitoring area corresponding to the monitoring unit includes areas with different soil types.

[0044] This step utilizes existing engineering data to perform preliminary, macroscopic regional division of the sampling zone to guide subsequent refined sampling. Pipeline design drawings provide information on pipeline burial depth, diameter, and corrosion protection type, while topographic maps along the route provide information on valve chambers, pumping stations, bends, and points of depth change. Combining this information, areas of potential soil variation can be preliminarily identified. This division strategy balances efficiency and accuracy, paying particular attention to areas with drastic soil changes. "Regular sections" refer to areas with relatively uniform soil and minimal variation, which can be divided using a larger "first spacing," such as 50m per sampling point. Each sampling point corresponds to one monitoring unit, with each unit covering a range of 25m before and after it; that is, one unit corresponds to one sampling point. "Special sections" refer to areas with frequent soil changes and complex geological conditions (such as landslides, soft soil areas, and permafrost areas). These require a smaller "second spacing," such as reducing it to 30m per sampling point, to ensure the capture of detailed soil changes. If significant differences in soil color and texture are observed on-site, additional sampling points should be added at the boundary of these differences to avoid missing abrupt changes. "Monitoring units added at soil difference boundaries" are to ensure accurate reflection of soil transitions at the interface between two different soil types; for example, additional monitoring units are set up at the interface between sand and clay. "The monitoring unit corresponds to a survey area including different soil types" means that a monitoring unit may not be a single soil type but rather a mixed area containing multiple soil types, requiring subsequent sampling to fully consider this complexity.

[0045] S203. Based on the pipeline design drawings, topographic maps along the route, and soil classification results, the sampling zone is marked to obtain multiple sampling points.

[0046] This step aims to refine the macro-level regional division into specific sampling locations. Geographic Information System (GIS) software can be used to overlay pipeline design drawings, topographic maps, and soil classification results. Within each monitoring unit, sampling points can be evenly or selectively placed based on the soil complexity and area size. Alternatively, on-site reconnaissance can be combined with manual selection of representative locations within each monitoring unit as sampling points, followed by GPS positioning.

[0047] S204. Collect the latitude and longitude coordinates of multiple sampling points on all monitoring units and construct a sampling table; the sampling table includes the latitude and longitude coordinates and the marker post number of each sampling point.

[0048] This step aims to record the precise location information of the sampling points to facilitate subsequent sampling and data management. Latitude and longitude coordinates can be collected on-site using a GPS locator. The sampling form can be a spreadsheet or database, containing not only latitude and longitude coordinates but also information such as the sampling point number, the monitoring unit to which it belongs, and the estimated soil type. The marker stake numbers can follow pre-defined coding rules, such as combining pipeline mileage markers with sampling point serial numbers. The marker stakes can be physical markers placed on-site, such as numbered wooden stakes or metal plaques, to help on-site personnel accurately locate the sampling positions.

[0049] S205. Collect multiple parallel soil samples for each sampling point based on latitude and longitude coordinates and marker stake numbers.

[0050] This step aims to ensure the representativeness and reliability of the sampling, providing sufficient and statistically significant samples for subsequent laboratory analysis. "Multiple parallel soil samples" refers to multiple soil samples collected at the same sampling point, at the same or similar depth, used to verify the consistency of analytical results or to conduct repeatability experiments. If the pipeline is buried deeper, a test pit with the same depth as the pipeline must be excavated first, and soil samples should be collected from the sidewall of the test pit at the same level as the pipeline to avoid interference from surface soil. Three sets of parallel soil samples are collected from each sampling point, used for soil classification, physical parameter testing, and as backup samples for retesting; impurities such as stones and plant roots must be removed from the soil samples during collection. Each set of soil samples is placed in a sealed bag, the bag opening is tied tightly with a rope, and the sampling point number, collection date, and collector are labeled; backup soil samples need to be placed in an additional waterproof bag to prevent moisture damage during transportation.

[0051] This invention first delineates sampling zones closely related to oil and gas pipeline safety, ensuring comprehensive monitoring coverage. Based on this, preliminary soil classification is performed using pipeline design drawings and topographic maps along the pipeline route, improving the scientific rigor and efficiency of the classification. Furthermore, a segmented, graded, and boundary-additional monitoring unit classification strategy effectively captures the complexity and variability of the soil along the oil and gas pipeline, avoiding resource waste and data distortion caused by blindly uniform sampling. This intelligent classification method allows for denser monitoring in areas with drastic soil variations, while allowing for more intensive monitoring in relatively homogeneous areas, thus optimizing sampling efficiency while ensuring data accuracy. Subsequently, by precisely marking sampling points and constructing a sampling table containing latitude and longitude coordinates and marker stake numbers, the sampling process is standardized and traceable. Finally, multiple parallel soil samples are collected at each sampling point, further ensuring the representativeness and reliability of the soil sample data. This series of systematic steps, through intelligent regional division and refined sampling management, ensured the accuracy and comprehensiveness of the constructed pipeline soil database, laying a solid foundation for the subsequent optimization of fiber optic monitoring parameters and the reliability of early warning results.

[0052] In some embodiments of the present invention, the input of the soil-fiber signal mapping model is soil parameters, and the output is fiber optic signal characteristics; the soil parameters include key soil parameters and auxiliary variable parameters; the key soil parameters include acoustic attenuation coefficient, strain transfer efficiency, natural moisture content, soil density, and elastic modulus; the auxiliary variable parameters include soil type; the fiber optic signal characteristics include acoustic sensing core characteristics and temperature sensing auxiliary characteristics; the acoustic sensing core characteristics include vibration signal amplitude, vibration signal dominant frequency, and signal-to-noise ratio; the temperature sensing auxiliary characteristics include temperature change gradient and temperature settling time.

[0053] The soil-fiber optic signal mapping model is a mathematical or algorithmic model used to establish the correlation between soil parameters and fiber optic signal characteristics. Its function is to transform complex soil environmental information into signal characteristics that can be used for fiber optic monitoring and early warning. This model can be based on predictive models built using machine learning algorithms, such as support vector machines, neural networks, and decision trees, or statistical regression models built based on physical principles and empirical formulas. Soil parameters refer to various indicators used to describe the physical, mechanical, and hydraulic properties of the soil surrounding oil and gas pipelines. These parameters are key factors in assessing soil stability, corrosivity, and their impact on pipelines. Fiber optic signal characteristics refer to the signal attributes collected by fiber optic sensors during monitoring that reflect changes in the environment surrounding the pipeline. These characteristics are the core data source of the fiber optic monitoring and early warning system. They can be achieved through distributed fiber optic sensing technology.

[0054] Key soil parameters refer to soil properties that have a decisive impact on the safety of oil and gas pipelines and the response of fiber optic signals. Changes in these parameters are often directly related to potential risk events. These parameters can be obtained through standard geotechnical testing methods. For example, the acoustic attenuation coefficient describes the degree of energy loss when sound waves propagate in soil, reflecting soil density, moisture content, and particle composition. It can be calculated from the attenuation curves of ultrasonic or shock waves propagating in soil, or from soil physical properties. Natural moisture content refers to the ratio of water mass to dry soil mass, and is an important indicator affecting soil strength, deformation, and freeze-thaw characteristics. Soil density refers to the mass of soil per unit volume, reflecting the soil's compactness and bearing capacity. The elastic modulus describes the soil's ability to resist elastic deformation and is an important mechanical parameter for assessing soil stiffness and deformation characteristics. Auxiliary variable parameters refer to soil properties that, in addition to key soil parameters, have a complementary impact on the soil-fiber optic signal mapping model. These parameters help to more comprehensively characterize the soil environment. Soil type refers to the classification of soil based on its physical, chemical, and biological properties. The core characteristics of acoustic sensing refer to the core signal attributes acquired through fiber optic acoustic sensing technology that directly reflect pipe vibration or acoustic events in the surrounding environment. These characteristics are essential for identifying abnormal events. Vibration signal amplitude describes the intensity of the fiber optic acoustic signal, reflecting the energy level of the vibration event. The dominant frequency of the vibration signal describes the most prominent frequency component in the fiber optic acoustic signal, reflecting the type and source of the vibration event. The signal-to-noise ratio (SNR) describes the ratio of effective signal to noise in the fiber optic acoustic signal, reflecting the signal quality and reliability.

[0055] Temperature sensing auxiliary features refer to signal attributes acquired through fiber optic temperature sensing technology that assist in judging pipeline conditions or environmental changes. These features can provide supplementary information for acoustic sensing, improving the accuracy of early warnings. Temperature change gradient describes the rate of temperature change in space or time, reflecting the intensity and range of heat sources or thermal anomalies. Temperature settling time describes the time required for temperature to change from one state to another and tend to stabilize, reflecting the characteristics of heat conduction or heat diffusion.

[0056] In the aforementioned optimized method for fiber optic monitoring and early warning, to ensure that the soil-fiber optic signal mapping model accurately reflects the intrinsic relationship between the soil environment surrounding the oil and gas pipeline and the fiber optic signal response, this application further clarifies the model's input and output. Specifically, the soil-fiber optic signal mapping model is designed to take soil parameters as input and fiber optic signal characteristics as output. This clear input-output definition allows the model to focus on establishing the quantitative relationship between these two types of key information. The soil parameters are further refined into key soil parameters and auxiliary variable parameters. Key soil parameters, such as acoustic attenuation coefficient, strain transfer efficiency, natural moisture content, soil density, and elastic modulus, directly reflect the physical and mechanical properties of the soil. These properties have a direct and significant impact on the pipeline's stress, deformation, and the response of the fiber optic sensor. For example, soil compaction (reflected by soil density) and moisture content affect the propagation speed and attenuation of sound waves in the soil, thus affecting the signal of the fiber optic acoustic sensor; the soil's elastic modulus and strain transfer efficiency directly determine how effectively pipeline or soil deformation is transmitted to the optical fiber, affecting the accuracy of the strain sensor. Auxiliary variable parameters, such as soil type, provide more macroscopic soil classification information, helping the model to make more accurate judgments and corrections under different geological backgrounds. Meanwhile, fiber optic signal features are refined into acoustic sensing core features (DAS) and temperature sensing auxiliary features (DTS). The extraction of acoustic sensing core features (DAS) is based on signal processing tools. The original vibration signal is low-pass filtered, and the peak-to-peak value of the filtered signal is calculated as the vibration amplitude. A Fourier transform is performed on the filtered time-domain signal, and the frequency value at the point of maximum power spectral density is taken as the dominant vibration frequency. The average power ratio of the effective frequency band to the noise frequency band is calculated and converted into the signal-to-noise ratio using a formula. The extraction of temperature sensing auxiliary features (DTS) is based on a data processing library. DTS temperature data within a specific range around the measuring point is selected, and the ratio of the temperature difference between adjacent points to their distance is calculated. The absolute value is taken as the temperature change gradient. The time from the temperature deviating from the baseline to temperature stabilization is recorded as the temperature stabilization time.

[0057] By defining soil parameters and fiber optic signal characteristics in such detail and with such specificity, the soil-fiber optic signal mapping model can receive comprehensive and representative input data and output signal characteristics highly correlated with the warning target. This refined input-output structure enables the model to more accurately learn and capture the complex nonlinear relationship between soil conditions and fiber optic response, thereby improving the model's ability to identify potential risk events and the accuracy of warnings. For example, when soil moisture content or density changes abnormally, the model can predict possible changes in fiber optic signal characteristics (such as vibration signal amplitude and signal-to-noise ratio) by inputting key soil parameters and considering their impact on acoustic attenuation coefficient and strain transfer efficiency. Therefore, in actual monitoring, once these abnormal fiber optic signal characteristics are detected, the model can more accurately trace back to specific soil problems, achieving more reliable early warnings.

[0058] In some embodiments of the present invention, such as Figure 3 As shown, the training process of the soil-fiber signal mapping model includes: S301. Collect the first optical fiber signal of each monitoring unit under normal operating conditions, and the second optical fiber signal of each monitoring unit under different risk conditions.

[0059] The first fiber optic signal consists of baseline data from the oil and gas pipeline under normal conditions. The second fiber optic signal comprises abnormal signal characteristics obtained by simulating various potential risks (such as leakage, third-party sabotage, and geological disasters) for each monitoring unit under different risk conditions. These signals form the basis for constructing a mapping model that can distinguish between normal and abnormal states. Data acquisition can be achieved by deploying distributed fiber optic sensors along the pipeline and continuously recording fiber optic signal data as the first fiber optic signal during normal pipeline operation. Simultaneously, fiber optic signal data under these risk scenarios can be obtained by simulating or actually triggering different types of risk events, or by utilizing historical event data, as the second fiber optic signal.

[0060] S302. Abnormal data is removed from the first and second fiber optic signals to obtain the target fiber optic signal.

[0061] The process involves outlier removal to purify the raw data. This includes removing signals with excessively low signal-to-noise ratios, extreme values, and soil parameters outside reasonable ranges. For field-measured data, the source of the outlier must be verified against environmental factors or sampling errors; erroneous data is directly removed to obtain the target fiber optic signal. Outlier data can severely impact model training and generalization capabilities, leading to incorrect patterns. Removing outlier data ensures the model learns from high-quality, representative data. Outlier removal methods can employ statistical approaches, such as using the 3σ criterion or box plots to identify and remove outliers. Another approach is to use machine learning algorithms, such as isolated forests or local anomaly detection, to automatically detect and remove outliers from the data.

[0062] S303. Match the target optical fiber signal of the same time period with the soil parameters in the pipeline soil database to obtain a complete sample.

[0063] Establishing the correlation between fiber optic signal characteristics and pipeline soil parameters to obtain a complete sample is crucial. The propagation and attenuation characteristics of fiber optic signals are closely related to the surrounding soil environment. Therefore, to train an effective soil-fiber optic signal mapping model, it is essential to accurately correlate fiber optic signal data within a specific time period with the soil parameters of the corresponding monitoring unit. This matching is key to constructing the model's input-output pairs. The matching process can be based on timestamps and monitoring unit identifiers. The complete sample can include data from each typical unit and each working condition, such as the correspondence between "5 soil parameters + 5 signal characteristics." The total sample size must meet the model training requirements to avoid overfitting.

[0064] S304. Normalize the soil parameters in the complete sample to obtain the target complete sample.

[0065] To eliminate differences in dimensions and numerical ranges among different soil parameters and prevent certain parameters with larger values ​​from dominating model training, thereby improving the model's convergence speed and training stability, it is necessary to normalize the soil parameters in the complete sample to obtain the target complete sample. This is particularly important for machine learning models based on optimization algorithms such as gradient descent. This avoids model weights being biased towards a particular parameter due to differences in parameter magnitudes; the output variables are not normalized initially to retain their actual physical meaning.

[0066] S305. Divide the target complete sample into training set, validation set and test set.

[0067] S306. The soil-fiber signal mapping model is trained using the training set, validation set, and test set to obtain the trained soil-fiber signal mapping model.

[0068] This step is a standard practice in machine learning model training, designed to evaluate the model's performance and generalization ability. The training set is used to learn the model parameters, the validation set is used to adjust the model's hyperparameters and prevent overfitting, and the test set is used to finally evaluate the model's performance on unseen data. A reasonable partitioning ensures the model has good predictive performance in practical applications. The partitioning method can be random partitioning or stratified sampling to ensure that data from different soil types or risk conditions are evenly distributed across the datasets. The soil-fiber optic signal mapping model is trained using the training, validation, and test sets to obtain the trained model. This step is the core logic of model learning. The training set is used for iterative optimization of the model parameters, the validation set is used to monitor the training process, prevent overfitting, and select the optimal model hyperparameters, and the test set is used for the final unbiased evaluation of the trained model. Through this process, the model learns the complex nonlinear relationship between soil parameters and fiber optic signal characteristics. Various machine learning algorithms can be used in the training process, such as support vector machines, random forests, gradient boosting trees, or deep learning neural networks.

[0069] In some embodiments of the present invention, step S102 includes: The soil parameters of each monitoring unit in the pipeline soil database are input into the soil fiber optic signal mapping model to obtain the theoretical signal characteristic threshold corresponding to each monitoring unit.

[0070] This step aims to leverage the established correlation between soil properties and fiber optic signals. The pipeline soil database stores detailed soil parameters for each monitoring unit along the oil and gas pipeline, such as acoustic attenuation coefficient, strain transfer efficiency, natural moisture content, soil density, and elastic modulus. Using these specific soil parameters as input allows the soil-fiber optic signal mapping model to predict the corresponding fiber optic signal performance based on the characteristics of different soil types. This input method ensures the soil-specificity of subsequent parameter determination. The theoretical signal characteristic thresholds for each monitoring unit are obtained. These thresholds refer to the critical values ​​of fiber optic signal characteristics predicted by the soil-fiber optic signal mapping model under specific soil conditions and associated with potential risk events (such as pipeline leaks and geological subsidence). These thresholds are derived from the model's understanding of the complex relationship between soil parameters and fiber optic signal characteristics, reflecting the theoretical limits of abnormal changes in fiber optic signals under different soil environments.

[0071] Based on the parameter optimization direction and theoretical signal characteristic thresholds, the fiber optic monitoring parameters of the corresponding monitoring unit are generated and initialized; the fiber optic monitoring parameters include the sampling frequency and the early warning signal threshold.

[0072] This method involves inputting soil parameters from each monitoring unit in the pipeline soil database into a soil-fiber optic signal mapping model to obtain theoretical signal characteristic thresholds closely related to the soil properties of that monitoring unit. These theoretical thresholds are learned by the model based on extensive historical data and soil characteristics, reflecting the performance boundaries of fiber optic signals under normal or abnormal conditions in specific soil environments. Based on this, and combined with pre-determined parameter optimization directions (e.g., for a certain soil type, higher early warning sensitivity or a lower false alarm rate may be required), the system can purposefully generate and initialize the fiber optic monitoring parameters for that monitoring unit, specifically including sampling frequency and early warning signal thresholds. This method ensures that the setting of fiber optic monitoring parameters is no longer uniform or empirical, but rather customized and optimized for the actual soil conditions and monitoring needs of each monitoring unit. In this way, the monitoring system can more accurately identify abnormal signals related to specific soil types, thereby improving the accuracy and reliability of early warnings and effectively avoiding false alarms or missed alarms caused by improper parameter settings.

[0073] In some embodiments of the present invention, step S104 includes: Set the trigger warning conditions for each soil type.

[0074] The soil environment along oil and gas pipelines is complex and diverse, and different soil types respond differently to external disturbances (such as geological subsidence, landslides, and third-party construction) and have different mechanisms of influence on fiber optic signals. Taking sandy soil and clay soil as examples: Sandy soil: Due to significant signal attenuation, an early warning is triggered when the signal characteristics reach 90% of the theoretical threshold, and a formal early warning is triggered when they reach 100%; Clay soil: Because the signal is easily affected by moisture content, it needs to be cross-verified with temperature signals, and an early warning is triggered when both signals meet the standards.

[0075] Determine whether the fiber optic monitoring parameters and real-time signal characteristics meet the triggering and early warning conditions corresponding to the soil type of the unit to be monitored; If so, then the unit to be monitored is determined to be at risk; If not, then it is determined that there is no risk in the unit to be monitored.

[0076] This invention compares real-time acquired fiber optic signal characteristics with pre-defined trigger warning conditions for the soil type of the unit to be monitored. This judgment mechanism ensures the targetedness and accuracy of the warning, avoiding the potential biases caused by a "one-size-fits-all" warning strategy. The judgment process may include, but is not limited to: comparing real-time signal characteristics (such as vibration signal amplitude, dominant frequency, signal-to-noise ratio, temperature change gradient, etc.) with preset thresholds; or inputting real-time signal characteristics into a classifier or regression model trained based on soil type to determine whether it belongs to an abnormal state. If so, it is determined that the unit to be monitored has a risk. It is clear that when the real-time signal characteristics meet the trigger warning conditions under a specific soil type, the system will determine that the unit to be monitored has a potential risk. This clear risk determination is the basis for subsequent countermeasures. Specifically, when a risk signal is detected, the system automatically calls the soil data and mapping model of the corresponding unit, outputs the risk level + soil influence coefficient, and assists maintenance personnel in judging the credibility of the warning. If not, it is determined that there is no risk to the monitored unit. This clarifies that when real-time signal characteristics do not meet the triggering warning conditions for a specific soil type, the system will determine that the monitored unit is currently in a normal state and there is no immediate risk. This helps reduce unnecessary intervention and improves system operating efficiency. For example, when the judgment result is "no," the system can continue routine monitoring without taking additional risk response measures.

[0077] In some embodiments of the present invention, the method further includes: The deviation rate is obtained by comparing the fiber optic monitoring parameters with the real-time signal characteristics. When the deviation rate exceeds the preset threshold, the fiber optic monitoring parameters of the monitoring unit are adjusted according to the soil parameters to be tested.

[0078] Based on the working principle of fiber optic sensors, this invention selects the parameters that have the greatest impact on monitoring accuracy as optimization targets, avoiding meaningless parameter adjustments that increase system complexity: Sampling parameters: directly affect the density and integrity of signal acquisition; Sampling frequency: the number of times signals are acquired per unit time, determining whether rapidly changing risk signals can be captured; Sampling resolution: the accuracy of signal quantization, determining whether minute signal differences can be distinguished; Sampling duration: the length of a single continuous acquisition, balancing data volume and signal effectiveness. The core criteria for determining whether an early warning is triggered include two types of thresholds: Basic signal threshold: a single parameter threshold for DAS vibration amplitude and DTS temperature change gradient, serving as the core basis for early warning judgment; Auxiliary verification threshold: a supplementary threshold for special soil types, used to reduce false alarms. Initial parameters only adapt to static soil characteristics, but in reality, soil parameters change with the environment, requiring a real-time adjustment mechanism to ensure that parameters always fit the current soil condition: specifically, a fixed monitoring cycle is set, and the real-time signal characteristics of each monitoring unit are compared with the fiber optic monitoring parameters to determine whether parameter adjustments are needed. Within each cycle, real-time signal characteristics of the unit are extracted, and soil parameters of the unit are collected synchronously. Deviation calculation involves calculating the deviation rate of the fiber optic monitoring parameters from the real-time signal characteristics. If the deviation rate is within a preset threshold (i.e., less than or equal to a preset threshold), the current parameters are maintained. If the deviation rate exceeds the preset threshold (i.e., greater than a preset threshold), the unit is marked as needing adjustment. Preliminary determination of the cause of deviation involves analyzing the cause of deviation based on the soil parameters of the monitored unit. If the deviation is due to increased moisture content, parameters related to moisture content are adjusted. If the deviation is due to temperature changes, seasonal parameter switching is triggered, resulting in adjusted fiber optic monitoring parameters. This ensures that the adjusted parameters more accurately reflect the monitoring needs under the current environment. Through this feedback closed-loop mechanism, fiber optic monitoring parameters can be corrected in real time according to environmental changes, sensor aging, and other factors, thereby ensuring the accuracy and reliability of the early warning system during long-term operation and effectively compensating for the limitations that fixed parameters may bring.

[0079] In some embodiments of the present invention, the method further includes: The fiber optic monitoring parameters of multiple monitoring units were evaluated, and the evaluation results were obtained.

[0080] This step in the embodiments of the present invention aims to periodically or as needed check the effectiveness of the currently used fiber optic monitoring parameters. The optimization effect of parameters for all monitoring units can be quantitatively evaluated periodically. Core evaluation dimensions include: Early warning accuracy indicators: statistically analyzing the false alarm rate and false negative rate of each soil unit. If the false alarm rate of a certain soil unit is too high, then it is necessary to trace back whether there is a problem with excessively low thresholds in its parameters; Signal effectiveness indicators: analyzing the signal-to-noise ratio and effective signal ratio of each unit. If the effective signal ratio of a unit is too low, then it is necessary to check whether its sampling frequency or gain compensation parameters are insufficient; Equipment operation indicators: statistically analyzing the abnormal situations caused by parameter settings of the monitoring equipment. If the equipment in a certain area frequently malfunctions, then it is necessary to reduce the sampling frequency of that area or optimize the data compression strategy to obtain the evaluation results.

[0081] When the evaluation results show that the optimization effect is not good, the poor soil type in the monitoring unit corresponding to the poor optimization effect is determined.

[0082] In this embodiment of the invention, this step is used to accurately pinpoint the specific reasons for poor monitoring results. When the evaluation results show that the monitoring parameters in certain areas have failed to meet expectations, the system will further analyze the characteristics of these areas, especially their soil type.

[0083] Re-collect soil parameters and fiber optic signal characteristics corresponding to poor soil types and add them to the pipeline soil database.

[0084] By supplementing the pipeline soil database, the soil fiber optic signal mapping model is retrained and its parameters are optimized to obtain updated fiber optic monitoring parameters for each monitoring unit.

[0085] If the evaluation of this invention reveals that the parameter optimization effect remains unsatisfactory, it is necessary to backtrack the mapping model to investigate whether the parameter deviation is caused by insufficient model accuracy, and then update the model. Specifically, this can be done as follows: Supplementing sample collection: For soil types with poor performance, re-collect "real-time soil parameters - actual signal characteristics" samples for that type of soil and supplement them to the model training set; Model retraining: Retrain the soil-fiber optic signal mapping model using the supplemented training set and optimize the model parameters; Parameter reconfiguration: Use the theoretical threshold output by the updated model as the new basis for parameter calculation, re-optimize the parameters of all monitoring units, overwrite the original parameter configuration, obtain the updated fiber optic monitoring parameters, and record the effect comparison before and after the update.

[0086] This invention addresses the potential performance degradation of fiber optic monitoring and early warning systems during long-term operation by introducing a feedback mechanism of continuous evaluation and iterative optimization. First, the system periodically or as needed evaluates the effectiveness of established fiber optic monitoring parameters. This evaluation not only focuses on the accuracy of early warning results but may also involve the analysis of key indicators such as false alarms and missed alarms. When the evaluation results show that the optimization effect of certain monitoring units is poor, the system does not simply adjust the parameters but delves into the underlying causes, accurately pinpointing the specific soil type leading to the problem. This step is crucial because it avoids blind adjustments and ensures that subsequent improvements are targeted. Once the poor soil type is identified, the system proactively triggers a data supplementation process, re-collecting soil parameters and fiber optic signal characteristics for these specific soil types. This new, more accurate data is integrated into the pipeline soil database, enriching the model's training samples. Subsequently, using this updated database, the soil-fiber optic signal mapping model is retrained and its parameters optimized. This process enables the model to learn a more comprehensive and accurate understanding of the complex relationship between soil and fiber optic signals, thereby generating theoretical signal characteristic thresholds that better reflect actual conditions. Ultimately, based on this optimized model, the system can generate updated fiber optic monitoring parameters for each monitoring unit. These parameters are better adapted to current environmental conditions and soil characteristics, thereby significantly improving the accuracy and reliability of fiber optic monitoring and early warning. Through this closed-loop iterative optimization, the proposed solution enables the fiber optic monitoring and early warning system to learn and adapt itself, ensuring that it provides efficient and accurate risk warnings throughout the entire lifecycle of oil and gas pipelines.

[0087] To better implement the fiber optic monitoring and early warning optimization method in this embodiment of the invention, correspondingly, this embodiment of the invention also provides a fiber optic monitoring and early warning optimization device, such as... Figure 4 As shown, the fiber optic monitoring and early warning optimization device 400 includes: The database construction module 401 is used to construct a pipeline soil database containing soil parameters and fiber optic signal characteristics for all soil types in multiple monitoring units on the oil and gas pipeline, and to determine the parameter optimization direction for each soil type. The parameter determination module 402 is used to determine the fiber optic monitoring parameters of each monitoring unit based on the parameter optimization direction and the soil fiber optic signal mapping model. The parameter optimization module 403 is used to determine the direction of target parameter optimization based on the soil type of the unit to be monitored, and to collect the soil parameters to be detected and real-time signal characteristics of the unit to be monitored according to the direction of target parameter optimization. The result determination module 404 is used to perform fiber optic monitoring and early warning of real-time signal characteristics based on fiber optic monitoring parameters, and obtain the early warning result. The report determination module 405 is used to generate an early warning report based on the soil parameters to be tested and the soil fiber optic signal mapping model when the early warning result indicates that there is a risk.

[0088] The fiber optic monitoring and early warning optimization device 400 provided in the above embodiments can realize the technical solutions described in the above fiber optic monitoring and early warning optimization method embodiments. The specific implementation principles of each module or unit can be found in the corresponding content in the above fiber optic monitoring and early warning optimization method embodiments, and will not be repeated here.

[0089] The fiber optic monitoring and early warning optimization method and device provided by the present invention have been described in detail above. Specific examples have been used to illustrate the principle and implementation of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation and application scope based on the idea of ​​the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. An optimized method for fiber optic monitoring and early warning, characterized in that, include: Construct a pipeline soil database containing soil parameters and fiber optic signal characteristics for all soil types in multiple monitoring units along the oil and gas pipeline, and determine the parameter optimization direction for each soil type. The fiber optic monitoring parameters of each monitoring unit are determined based on the parameter optimization direction and the soil fiber optic signal mapping model. Based on the soil type of the unit to be monitored, the optimization direction of the target parameters is determined, and the soil parameters to be detected and real-time signal characteristics of the unit to be monitored are collected according to the optimization direction of the target parameters. Based on the fiber optic monitoring parameters, fiber optic monitoring and early warning are performed on the real-time signal characteristics to obtain early warning results; When the warning result indicates that there is a risk, a warning report is obtained based on the soil parameters to be detected and the soil fiber optic signal mapping model.

2. The fiber optic monitoring and early warning optimization method according to claim 1, characterized in that, The constructed oil and gas pipeline soil database includes soil parameters and fiber optic signal characteristics for all soil types, including: Multiple monitoring units were delineated along the oil and gas pipeline, and multiple parallel soil samples were collected from the multiple monitoring units. Laboratory analysis was performed on the multiple parallel soil samples to determine the soil type and soil parameters of each monitoring unit; Based on the soil type and soil parameters, a pipeline soil database is constructed.

3. The fiber optic monitoring and early warning optimization method according to claim 2, characterized in that, The process involves defining multiple monitoring units along the oil and gas pipeline and collecting multiple parallel soil samples from these monitoring units, including: Using the centerline of the oil and gas pipeline as a reference, the main areas of influence within the pipeline burial depth range are extended to both sides to obtain the sampling zone; Based on the pipeline design drawings and the topographic map along the route, the sampling zone is divided into monitoring units and soil types to obtain soil classification results. The soil classification results include monitoring units divided at a preset first interval in the regular section, monitoring units divided at a second interval smaller than the first interval in the special section, and monitoring units added at the boundaries of soil differences. The monitoring area corresponding to the monitoring unit includes areas with different soil types. The sampling zone is marked according to the pipeline design drawing, the topographic map along the route, and the soil classification results to obtain multiple sampling points; The latitude and longitude coordinates of multiple sampling points on all monitoring units are collected to construct a sampling table; the sampling table includes the latitude and longitude coordinates and the marker stake number of each sampling point; Multiple parallel soil samples were collected from each sampling point based on the latitude and longitude coordinates and the marker stake number.

4. The fiber optic monitoring and early warning optimization method according to claim 1, characterized in that, The input to the soil-fiber signal mapping model is soil parameters, and the output is fiber optic signal characteristics. The soil parameters include key soil parameters and auxiliary variable parameters. The key soil parameters include acoustic attenuation coefficient, strain transfer efficiency, natural moisture content, soil density, and elastic modulus. The auxiliary variable parameters include soil type. The fiber optic signal characteristics include acoustic sensing core characteristics and temperature sensing auxiliary characteristics. The acoustic sensing core characteristics include vibration signal amplitude, vibration signal dominant frequency, and signal-to-noise ratio. The temperature sensing auxiliary characteristics include temperature change gradient and temperature settling time.

5. The fiber optic monitoring and early warning optimization method according to claim 2, characterized in that, The training process of the soil-fiber signal mapping model includes: The system collects the first optical fiber signal of each monitoring unit under normal operating conditions, and the second optical fiber signal of each monitoring unit under different risk conditions. Abnormal data is removed from the first and second optical fiber signals to obtain the target optical fiber signal; The target optical fiber signal at the same time period is matched with the soil parameters in the pipeline soil database to obtain a complete sample; The soil parameters in the complete sample are normalized to obtain the target complete sample; The target complete sample is divided into a training set, a validation set, and a test set; The soil-fiber signal mapping model is trained using the training set, the validation set, and the test set to obtain the trained soil-fiber signal mapping model.

6. The fiber optic monitoring and early warning optimization method according to claim 1, characterized in that, The process of determining the fiber optic monitoring parameters for each monitoring unit based on the optimized direction of the parameters and the soil fiber optic signal mapping model includes: The soil parameters of each monitoring unit in the pipeline soil database are input into the soil fiber optic signal mapping model to obtain the theoretical signal characteristic threshold corresponding to each monitoring unit. Based on the parameter optimization direction and the theoretical signal characteristic threshold, the fiber optic monitoring parameters of the corresponding monitoring unit are generated and initialized; the fiber optic monitoring parameters include the sampling frequency and the early warning signal threshold.

7. The fiber optic monitoring and early warning optimization method according to claim 1, characterized in that, The step of performing fiber optic monitoring and early warning based on the fiber optic monitoring parameters to obtain early warning results includes: Set the triggering warning conditions for each soil type; Determine whether the fiber optic monitoring parameters and the real-time signal characteristics meet the triggering early warning conditions corresponding to the soil type of the unit to be monitored; If so, then the monitored unit is determined to be at risk; If not, then it is determined that the unit to be monitored does not pose a risk.

8. The fiber optic monitoring and early warning optimization method according to claim 1, characterized in that, The method further includes: The deviation rate is obtained by comparing the optical fiber monitoring parameters with the real-time signal characteristics. When the deviation rate is greater than a preset threshold, the fiber optic monitoring parameters of the monitoring unit are adjusted according to the soil parameters to be detected.

9. The fiber optic monitoring and early warning optimization method according to claim 1, characterized in that, The method further includes: The fiber optic monitoring parameters of the multiple monitoring units are evaluated to obtain evaluation results; When the evaluation results show that the optimization effect is not good, the poor soil type in the monitoring unit corresponding to the poor optimization effect is determined. The soil parameters and fiber optic signal characteristics corresponding to the poor soil type are re-collected and added to the pipeline soil database; The soil fiber optic signal mapping model is retrained and its parameters are optimized by supplementing the pipeline soil database, resulting in updated fiber optic monitoring parameters for each monitoring unit.

10. A fiber optic monitoring and early warning optimization device, characterized in that, include: The database construction module is used to build a pipeline soil database containing soil parameters and fiber optic signal characteristics for all soil types in multiple monitoring units on oil and gas pipelines, and to determine the parameter optimization direction for each soil type. The parameter determination module is used to determine the fiber optic monitoring parameters of each monitoring unit based on the parameter optimization direction and the soil fiber optic signal mapping model. The parameter optimization module is used to determine the direction of target parameter optimization based on the soil type of the unit to be monitored, and to collect the soil parameters to be detected and real-time signal characteristics of the unit to be monitored according to the target parameter optimization direction. The result determination module is used to perform fiber optic monitoring and early warning on the real-time signal characteristics based on the fiber optic monitoring parameters, and obtain the early warning result. The report determination module is used to generate an early warning report based on the soil parameters to be detected and the soil fiber optic signal mapping model when the early warning result indicates that there is a risk.

Images