Oil field pipeline clustering leakage monitoring system

Abstract

The invention relates to the technical field of oil field pipeline leakage monitoring, and discloses an oil field pipeline clustering leakage monitoring system, which comprises a central platform for executing global resource allocation, task decision and visual management and control, and optimizing a task allocation path through a gradient descent algorithm; the redundant block chain is formed by interconnection of adjacent regional monitoring platforms through a smart contract, and is divided into a core chain and a standby chain; the regional monitoring platform comprises a multi-stage detection unit, and is used for collecting pipeline operation data in real time and triggering a multi-stage verification mechanism; the dynamic permission distribution module is used for dynamically migrating tasks to low-load nodes; and the self-learning module is used for constructing a decision result database to record a false alarm rate and a positioning error, and dynamically optimizing a flow decreasing threshold and a vibration spectrum matching weight by adopting a reinforcement learning model. According to the invention, the real-time performance, the positioning precision and the system fault tolerance of leakage monitoring are improved through a centralized-distributed hybrid architecture, multi-stage cooperative positioning and anti-interference design, and technical support is provided for safe operation and maintenance of oil field pipelines.

Description

Technical Field

[0001] The present invention relates to the technical field of oilfield pipeline leakage monitoring, in particular to a clustered oilfield pipeline leakage monitoring system. Background Art

[0002] With the development of the petroleum industry, the safe operation of oilfield pipelines is crucial. However, existing pipeline leakage monitoring technologies suffer from numerous drawbacks. Traditional leak monitoring technologies often rely on manual inspections or fixed sensor networks, which present significant drawbacks. Manual inspections are limited by environmental conditions and labor costs, making it difficult to detect even minor leaks in a timely manner. To improve automation, sensor monitoring solutions are gradually being introduced. For example, single-point sensors or segmented pressure and temperature sensors are installed on the pipeline, detecting leaks based on data anomalies. However, these methods generally suffer from insufficient positioning accuracy. Some existing systems attempt to improve monitoring performance through multi-parameter fusion (such as pressure, flow, and temperature) or artificial intelligence algorithms. However, existing technologies often rely on centralized management. Centralized sensor networks rely on fixed detection cycles and are unable to dynamically respond to changes in pipeline operating conditions, resulting in wasted resources and insufficient monitoring of critical areas. Despite attempts to incorporate distributed monitoring or IoT architectures, existing technologies still face challenges such as data silos, insufficient redundancy, and a lack of dynamic adjustment mechanisms. Therefore, a clustered oilfield pipeline leakage monitoring method that combines global collaboration and regional autonomy is urgently needed. Summary of the Invention

[0003] This paper proposes a clustered oilfield pipeline leakage monitoring system. Through regional management, dynamic adjustment mechanism, and global collaborative management, it avoids the defects of existing clustered oilfield pipeline leakage monitoring methods that cannot dynamically respond to changes in pipeline operation status, resulting in waste of resources or insufficient monitoring of key areas. The specific solution is as follows: A clustered leakage monitoring system for oilfield pipelines, comprising: Central platform: performs global resource allocation, task decision-making, and visual management and control, generates load heat maps in real time, and optimizes task allocation paths using a gradient descent algorithm; Redundant blockchain: formed by interconnecting adjacent regional monitoring platforms through smart contracts, divided into a core chain and a backup chain. The core chain takes priority over monitoring tasks in areas with a high incidence of historical leaks, cross-border pipelines, or high-pressure oil pipelines. The backup chain is normally in low-power monitoring mode and is activated through a consensus algorithm when the core chain load exceeds the safety threshold. Regional monitoring platform: Deployed in independent monitoring areas, it includes multi-level detection units, collects pipeline operation data in real time and triggers a multi-level verification mechanism; Dynamic authority allocation module: dynamically migrate tasks to low-load nodes based on the proportion of regional platform tasks and blockchain tasks; Self-learning module: Build a decision result database to record false alarm rates and positioning errors, and use a reinforcement learning model to dynamically optimize the flow reduction threshold and vibration spectrum matching weight.

[0004] Furthermore, the multi-stage detection unit includes a pipeline detection unit, an edge detection unit, and an area detection unit; The pipeline detection units are arranged along the extension direction of the pipeline; the edge detection units are arranged at both ends of the pipeline; the edge detection units located at the nodes of the monitoring area constitute the area detection units; the edge detection units and the area detection units are used to locate each area and each pipeline and monitor the operating status of each area and each pipeline; the pipeline detection unit is used to monitor and determine the leaking pipeline and the leakage point of the leaking pipeline; The pipeline detection unit includes a positioning layer, a monitoring layer, and a protective layer, which are arranged in sequence from the inside to the outside along the radial direction of the pipeline; the positioning layer, the monitoring layer, and the protective layer form a nested module; the protective layer ensures that the vibration energy of the pipeline is transmitted to the monitoring layer; the monitoring layer captures the vibration of the pipeline in real time through the piezoelectric effect, obtains monitoring data, and decides whether to issue a detection warning by comparing the monitoring data with a first threshold; after the positioning layer issues a detection warning, the monitoring layer is activated and enters a positioning mode. The monitoring layer amplifies the defect location on the one hand, and on the other hand, achieves dual verification of the leakage mode and accurately locates the leakage location by temporal and spatial matching of the light intensity attenuation gradient with the vibration spectrum of the positioning layer on the other hand; The positioning layer includes an OTDR module, a color development layer located outside the pipeline, and a detection optical fiber; the color development layer covers the entire outer wall of the pipeline; the detection optical fiber is arranged outside the color development layer; the OTDR module is integrated at the end of the detection optical fiber; The monitoring layer includes a microprocessor located at the end of the pipeline, as well as a PVDF film and a conductive grid outside the monitoring layer. The PVDF film is tightly attached to the inner surface of the protective layer, with its polarization direction perpendicular to the pipeline surface. The conductive grid covers the side of the PVDF film closest to the positioning layer, and the microprocessor is embedded in the edge of the PVDF film. The monitoring layer amplifies the defect area and, through spatiotemporal matching of the light intensity attenuation gradient with the vibration spectrum of the positioning layer, achieves dual verification of the leakage mode and accurately locates the leak. The protective layer is an anti-corrosion polyurethane layer, which is evenly wrapped around the entire monitoring layer to protect the monitoring layer and the positioning layer from soil erosion. The inner surface of the protective layer is embedded with copper foil electrodes, which are evenly distributed along the circumference of the pipeline and are used to connect the conductive grid of the positioning layer.

[0005] Furthermore, the coordination mode of the positioning layer and the monitoring layer is: After the PVDF film captures the vibration signal, the microprocessor extracts the covariance features, including the training set and the target set; the training set is the CIR tap autocovariance matrix of the normal state vibration spectrum; the target set is the cross-covariance vector of the abnormal state vibration spectrum and the training set; The vibration spectrum residual is calculated through a linear prediction program, and OTDR positioning is triggered when the residual is greater than the second threshold; The copper foil electrode of the protective layer eliminates electromagnetic interference and improves the signal-to-noise ratio to ≥30dB.

[0006] Furthermore, the monitoring layer performs: The pipeline vibration is captured in real time through the piezoelectric effect, and a first-level warning is triggered when the vibration energy exceeds the first threshold. After entering the positioning mode, the OTDR light intensity attenuation gradient and vibration spectrum are analyzed synchronously, and dual leakage verification is achieved through time and space matching; The calculation method of the leakage point coordinates is: Step 1: Use OTDR to determine the distance L1 between the leakage point and the starting point of the optical fiber. Step 2: Compare the time delay Δt of the vibration signals at both ends of the pipeline and calculate the distance between the leakage point and the near end, L2 = (Δt × speed of sound) / 2; Step 3: Output coordinates (L1, L2) to the GIS system.

[0007] Furthermore, the dynamic rights allocation module includes: When the task volume of a regional platform exceeds 80% or the CPU utilization exceeds the safety threshold, the smart contract will migrate the excess tasks to an adjacent low-load platform. Based on the global load heat map, the central platform dynamically adjusts the jurisdiction boundaries of redundant blockchains as follows: when adding new monitoring areas, they are assigned to low-load blockchains based on distance priority; when migrating tasks, the migration volume is satisfied: the migration volume = original task volume × (1-remaining load rate of the target blockchain); The dynamic authority allocation module further includes: allocating monitoring authority, including: When the task load of a regional monitoring platform exceeds a preset threshold or the operating status is abnormal, the redundant blockchain will migrate some tasks to an adjacent low-load platform through a smart contract; The central platform divides redundant blockchains into core chains and backup chains based on the global task volume distribution. The core chain takes over high-priority tasks first, while the backup chain responds to sudden loads. The central platform also adjusts the boundaries of the redundant blockchain's jurisdiction in real time based on its task volume ratio, operating status, and regional feedback data. The calculation method of the task volume ratio is: regional monitoring platform task volume ratio = (current task processing time × task priority coefficient) / regional platform maximum load capacity; The proportion of redundant blockchain tasks = ∑ (the proportion of tasks in the platform in the corresponding area) × blockchain weight coefficient; wherein, the task priority coefficient is dynamically calculated based on the pipeline leakage risk level, task type weight and environmental sensitivity, and the blockchain weight coefficient is dynamically adjusted based on the historical task success rate, current load rate and node stability score.

[0008] Furthermore, the multi-level verification mechanism is: When the regional platform detects that the pipeline vibration energy increases by more than 20% for three consecutive hours, it is marked as an abnormal pipeline; Obtain the flow data of adjacent areas and the pressure fluctuation value of edge units, if the spatiotemporal matching conditions are met: Condition 1: Regional traffic decrease > dynamic threshold; Condition 2: The time difference between the flow valley value and the vibration peak value is less than 10 seconds; If there is a leak, the detection period is extended to 1 hour.

[0009] Furthermore, the self-learning module performs: Supervised learning model: Uses a random forest algorithm to analyze historical traffic drop characteristics and environmental factors, and outputs dynamic threshold adjustment recommendations. For every 1% increase in false alarm rate, the threshold is tightened by 0.8%; Reinforcement learning model: Generate task migration paths based on the DQN network, with a reward function R = number of successful detections × 1 + positioning error reduction × 0.2 - number of false alarms × 0.5; The model parameters are iteratively updated every 24 hours, and manual intervention is triggered when the positioning error is greater than 1 meter.

[0010] Furthermore, the central platform updates and displays in real time each reallocated monitoring area and the corresponding management area monitoring platform, as well as the area monitoring governed by the redundant blockchain.

[0011] Furthermore, the updating of the self-learning module includes: Build a decision result database to record the success rate, false alarm rate and positioning error of leak verification; A reinforcement learning model is used to dynamically adjust the flow reduction threshold, vibration spectrum matching weight and task allocation strategy with the goal of minimizing the false alarm rate and positioning error.

[0012] Furthermore, the hierarchical management of permissions of the redundant blockchain is specifically as follows: The core chain will prioritize monitoring tasks in areas with a high incidence of historical leaks, cross-border pipelines, or pipelines transporting high-pressure oil products; The backup chain is normally in low-power monitoring mode. When the core chain load exceeds the safety threshold or the regional platform fails, it is activated and takes over the task through the consensus algorithm; The redundant blockchain uses an atomic transaction log: Each blockchain node maintains a global transaction log, where entries store task descriptors and atomic TLE flags; Power failure recovery process: Step 1: Scan the global transaction log entries to identify TLE flag state discontinuities; Step 2: Determine the LTE adjacent to the discontinuity as a valid task; Step 3: weak programming flag execution is enhanced to a stable state; Step 4: Restore abnormal tasks from LTE to the migration queue; The atomic TLE flag enhancement mechanism includes: Weak clear mark: Apply +0.5V bias to the copper foil electrode for 10ms to reset the polarization direction of PVDF; Weak setting flag: injecting 5mA current through the conductive grid triggers the piezoelectric film to self-calibrate; After enhancement, the bit error rate is reduced to <10 6 ; The pressure boost verification process is as follows: When the flow difference ΔQ is greater than the flow drop threshold, the booster valve increases the pipeline pressure by 0.5 MPa; Re-test the flow difference ΔQ'. If the expansion rate = (ΔQ'-ΔQ) / ΔQ>15%, leakage is confirmed. Positioning accuracy is improved to ±0.5 meters, and verification takes less than 2 minutes.

[0013] Compared with the prior art, the present invention has the following advantages: This application achieves a synergistic effect between global coordination and regional autonomy by constructing a three-tier hybrid architecture: a central platform, redundant blockchains, and regional monitoring platforms. The central platform is responsible for global resource allocation and task decision-making. Regional monitoring platforms adaptively optimize detection cycles based on real-time pipeline operational data, reducing the central platform's load and improving the system's responsiveness to changes in pipeline operating status. Redundant blockchains connect to adjacent regional platforms through smart contracts, forming a complementary backup chain. When a regional monitoring platform experiences overload or operational anomalies, it can promptly migrate some tasks to a neighboring, less-loaded platform. The regional monitoring platform's jurisdictional boundaries are adjusted in real time, effectively mitigating the impact of single points of failure on the entire monitoring system and significantly enhancing system stability and reliability. The regional monitoring platform uses real-time pipeline operational data to trigger a multi-level verification mechanism to determine pipeline operating status, improving the accuracy of leak risk assessments and triggering dynamic monitoring mechanisms based on verification conclusions. Dynamically adjusting detection cycles based on pipeline risk levels ensures optimal allocation of monitoring resources and significantly improves resource utilization. The central platform, redundant blockchains, and regional monitoring platforms record the results of each leak verification decision and use machine learning to optimize flow reduction thresholds and task allocation algorithms. As monitoring data continues to accumulate, the system can continuously learn and optimize, continuously improving the accuracy of subsequent monitoring. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 The relationship between the regional monitoring platform and the redundant chain in the embodiment of this application, as well as the principle architecture diagram of the central platform; Figure 2 : is a schematic diagram of the layout of the multi-stage detection unit in the embodiment of the present application; Figure 3 This is the structural intention of the pipeline detection unit in the embodiment of this application; In the figure: 13, edge monitoring unit; 14, positioning layer; 15, monitoring layer; 16, protection layer; 17, pipeline detection unit; 18, area monitoring unit. DETAILED DESCRIPTION

[0014] An oilfield pipeline cluster leakage monitoring system, comprising: System Architecture: Equipped with a central platform, it assumes core responsibilities for global resource allocation, task decision-making, and visual management and control. Based on the regional characteristics of oilfield pipelines, and taking into account environmental factors, geological factors, and historical leakage data, the pipeline cluster is divided into several independent monitoring areas, each equipped with a dedicated regional monitoring platform. Adjacent regional monitoring platforms are interconnected using smart contracts to form a redundant blockchain. This redundant blockchain works in conjunction with the central platform to build a three-tier hybrid architecture: "central platform - redundant blockchain - regional monitoring platform."

[0015] Pipeline operation status monitoring and processing: The regional monitoring platform is responsible for collecting the operation data of pipelines within its jurisdiction in real time to verify whether the pipelines are operating normally. Once a pipeline is found whose operation data exceeds the normal threshold range, it is determined to be an abnormal pipeline, and the multi-level verification mechanism is triggered immediately. In the multi-level verification mechanism, the regional monitoring platform obtains the overall operation data trend of the adjacent areas of the abnormal pipeline and the monitoring area under its jurisdiction. If the regional flow drop and the edge abnormal data match each other in time and space, the pipeline can be determined to be a faulty pipeline. Based on the conclusions of the multi-level verification mechanism, the dynamic monitoring mechanism is further triggered. For high-risk pipelines, the detection cycle is shortened to the preset threshold to allow more frequent monitoring and timely detection of potential leakage risks; for low-risk pipelines, the detection cycle is extended to avoid unnecessary waste of resources.

[0016] Dynamic Allocation of Monitoring Rights: Monitoring rights are dynamically allocated to regional monitoring platforms and redundant blockchains based on the workload share of the regional monitoring platforms and the operational status of the redundant blockchains. When the workload share of a regional monitoring platform exceeds a preset threshold or its operational status deviates, the redundant blockchain uses smart contracts to migrate some tasks to a neighboring, less-loaded platform, ensuring efficient overall system operation. Based on the global workload distribution, the central platform divides the redundant blockchains into core and backup chains. The core chain prioritizes high-priority monitoring tasks, such as those in areas with a high history of leaks, cross-border pipelines, or pipelines transporting high-pressure oil products. The backup chain normally operates in a low-power monitoring mode. When the core chain's load exceeds a safety threshold or a regional platform experiences a failure, it is activated and takes over the corresponding tasks through a consensus algorithm. Furthermore, the central platform adjusts the boundaries of its jurisdiction in real time based on the workload share, operational status, and regional feedback data of the redundant blockchains to optimize resource allocation.

[0017] Optimizing Monitoring Accuracy: The central platform, redundant blockchains, and regional monitoring platforms jointly record the decision results of each leak verification and utilize machine learning techniques to optimize the flow reduction threshold and task allocation algorithm. By continuously accumulating monitoring data, the system continuously learns and improves, thereby enhancing subsequent monitoring accuracy. Specifically, a decision result database is constructed to record key information such as the leak verification success rate, false alarm rate, and positioning error. A reinforcement learning model is employed to dynamically adjust the flow reduction threshold, vibration spectrum matching weight, and task allocation strategy with the goal of minimizing false alarm rate and positioning error.

[0018] Multi-level detection units work together: The regional detection unit is equipped with multi-level detection units, including pipeline detection units, edge detection units, and regional monitoring units. The pipeline detection unit is arranged along the extension direction of the pipeline to collect the operating status of the pipeline itself in real time. Once a data anomaly is detected, a first-level warning signal is triggered. The edge detection unit is set at both ends of the pipeline to synchronously analyze the operating status of both ends of the pipeline and generate a second-level verification signal when an anomaly is detected. The regional monitoring unit monitors the operating status of the monitoring area based on the GIS system. When the operating change trend exceeds the dynamic threshold and there is a spatiotemporal correlation with the first-level warning signal and the second-level verification signal, the leakage event is confirmed and the leakage location mode is activated. Through the coordinated cooperation of multi-level detection units, dual verification and precise positioning of the leakage point are achieved, effectively solving the problem of inaccurate positioning in traditional technologies.

[0019] Pipeline Detection Unit Structural Design: The pipeline detection unit is arranged along the pipeline's extension direction. It includes a positioning layer, a monitoring layer, and a protective layer, arranged radially from the inside out. These three layers form a nested module. The protective layer, made of corrosion-resistant polyurethane, evenly wraps around the entire monitoring layer, protecting the monitoring and positioning layers from soil erosion. Its function is to ensure that the pipeline's vibration energy is transmitted to the monitoring layer. The monitoring layer captures pipeline vibration in real time through the piezoelectric effect, obtains monitoring data, and compares the monitoring data with a first threshold. When the monitoring data exceeds the first threshold, a detection warning is issued. After the positioning layer issues a detection warning, it activates the monitoring layer and puts it into positioning mode. The monitoring layer amplifies the defect location and, through spatiotemporal matching of the light intensity attenuation gradient with the vibration spectrum of the positioning layer, achieves dual verification of the leakage pattern and accurately locates the leak. Among them, the positioning layer includes an OTDR module, a colorimetric layer located on the outside of the pipeline, and a detection optical fiber. The colorimetric layer covers the entire outer wall of the pipeline, the detection optical fiber is arranged outside the colorimetric layer, and the OTDR module is integrated at the end of the detection optical fiber; the monitoring layer includes a microprocessor located at the end of the pipeline, and a PVDF film and a conductive grid outside the monitoring layer. The PVDF film is tightly attached to the inner surface of the protective layer, and the polarization direction is perpendicular to the pipeline surface. The conductive grid covers the side of the PVDF film close to the positioning layer, and the microprocessor is embedded in the edge of the PVDF film.

[0020] Information display and update: The central platform updates and displays in real time the monitoring status of each reallocated monitoring area and the corresponding management area monitoring platform, as well as the monitoring status of the areas governed by redundant blockchains, enabling managers to intuitively understand the system's operating status and resource allocation, facilitating global management and decision-making.

[0021] It should be noted that: In this application, the central platform serves as the core management body for the entire system, responsible for overall coordination and decision-making. The redundant blockchain serves as a blockchain structure that provides backup and stability within the system. The regional monitoring platform monitors specific areas and feeds relevant task data back to the redundant blockchain.

[0022] The regional monitoring platform stores monitoring data and test results on its redundant blockchain. This not only leverages the complementary nature of redundant blockchains to ensure data security and integrity, but also facilitates subsequent data query, analysis, and processing. The operational status of the redundant blockchains and regional monitoring platform includes their processing efficiency, stability, and failure status.

[0023] The redundant blockchain dynamically adjusts the monitoring tasks and management authority of the regional monitoring platform according to the proportion of the regional monitoring platform's tasks in the redundant blockchain and its operating status. The redundant blockchain takes over the management tasks of the regional monitoring platform within or outside the jurisdiction according to the decision-making instructions of the central platform; The central platform determines whether to directly manage the redundant blockchain or the regional monitoring platform based on the proportion of the redundant blockchain's tasks in the entire central platform, the corresponding jurisdiction and commands for deploying the redundant blockchain, and the proportion of the redundant blockchain's tasks and operating status in the central platform and the proportion of the regional monitoring platform's tasks and operating status in the corresponding redundant blockchain as fed back by the redundant blockchain.

[0024] In this application, redundant blockchains will take over the management tasks of regional monitoring platforms within or outside the jurisdiction according to the decision instructions issued by the central platform. Specifically, when a redundant blockchain has a high proportion of tasks but is in good operating condition, the central platform will select a redundant blockchain with good operating condition, a low proportion of tasks, and adjacent to it based on the proportion of redundant blockchain tasks and operating status. If the task volume of adjacent redundant blockchains is saturated, the redundant blockchains closer to it will be deployed with some of the tasks. This means that after the central platform makes the relevant decision, the redundant blockchain can reallocate and take over the management tasks of the monitoring platforms in different regions to achieve more efficient monitoring and management. Based on the aforementioned indicators, the central platform decides whether to directly manage the redundant blockchain or the regional monitoring platform. For example, if the redundant blockchain has a high workload and is performing well, but the regional monitoring platform has an abnormal workload or is performing poorly, the central platform may directly manage the regional monitoring platform. Conversely, if a redundant blockchain experiences issues, the central platform may directly manage and intervene. This allows for emergency response and optimized resource allocation.

[0025] In this application, the multi-level detection units deployed in the regional monitoring platform are pipeline detection units, edge detection units, and regional detection units. Pipeline detection units are arranged along the pipeline's extension, covering the entire length of the pipeline. Edge detection units are located at both ends of the pipeline, often at the junction of two adjacent pipelines. They monitor the spatial position of both ends of the pipeline, reflecting the operating pressure, flow, and temperature of each pipeline. Regional detection units are composed of edge detection units located at the nodes of the monitoring area. Two edge detection units located at adjacent monitoring area nodes reflect the operating pressure, flow, and temperature of the pipelines within the entire monitoring area, playing a role at key nodes in the area. Edge detection units and regional detection units are primarily used to locate each area and pipeline, and monitor the operating status of these areas and pipelines. For example, this includes determining whether the pipeline is operating normally and whether related facilities within the area are operating stably. This includes, but is not limited to, monitoring the operating pressure, flow, and temperature of the pipeline through sensors. The pipeline detection unit primarily monitors pipeline leaks. Once a leak is detected, it can identify the specific pipeline and the location of the leak point on the leaking pipeline, allowing for timely repair measures.

[0026] In this application, the monitoring layer has two main functions: first, it amplifies areas where defects may exist to more clearly identify problems; second, it performs a spatiotemporal matching of the light intensity attenuation gradient with the vibration spectrum of the positioning layer to achieve dual verification of the leakage pattern. Based on this information, the location of the pipeline leak is then precisely determined. This allows for real-time monitoring of the pipeline and precise location of leaks, enabling timely detection and resolution of pipeline leaks.

[0027] In this application, the task volume ratio is calculated as follows: The proportion of regional monitoring platform tasks = (current task processing time × task priority coefficient) / regional platform maximum load capacity; The proportion of redundant blockchain tasks = ∑ (the proportion of tasks on the platform in the corresponding region) × blockchain weight coefficient; The central platform generates a load heat map based on the proportion results and optimizes the task allocation path through the gradient descent algorithm.

[0028] Furthermore, the updating of the self-learning module includes: Build a decision result database to record the success rate, false alarm rate and positioning error of leak verification; A reinforcement learning model is used to dynamically adjust the flow reduction threshold, vibration spectrum matching weight and task allocation strategy with the goal of minimizing the false alarm rate and positioning error.

[0029] Furthermore, the hierarchical management of the permissions of the redundant blockchain is specifically as follows: The core chain will prioritize monitoring tasks in areas with a high incidence of historical leaks, cross-border pipelines, or pipelines transporting high-pressure oil products; The backup chain is normally in low-power monitoring mode. When the core chain load exceeds the safety threshold or the regional platform fails, it is activated and takes over the task through the consensus algorithm.

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0031] Example 1: Carry out oilfield pipeline area division and platform deployment, as shown in the attached Figure 1 and 2 As shown, based on the geographic distribution of oilfield pipelines (such as geological stability and historical leak frequency) and environmental factors (such as climate and soil corrosivity), pipelines 1 through 12 in the pipeline cluster are divided into four independent monitoring areas (A, B, C, and D). Each area is deployed with a regional monitoring platform (including edge computing nodes), namely regional monitoring platforms A and D. Adjacent regional platforms are linked through smart contracts to form redundant blockchains (for example, regional monitoring platforms A and B form redundant blockchain I, and regional monitoring platforms C and D form redundant blockchain II), enabling data sharing and mutual backup. The central platform is deployed in the oilfield control center and connected to each regional monitoring platform and redundant blockchain nodes (based on the Hyperledger Fabric architecture) via a 5G private network.

[0032] The regional monitoring platform is equipped with three levels of monitoring units, including pipeline detection units 17, edge detection units 13, and regional detection units 18. Pipeline detection units 17 are arranged along the extension direction of the pipeline, edge detection units 13 are set at both ends of the pipeline, and edge detection units located at the nodes of the monitoring area constitute regional detection units 18. The schematic diagram of their location and installation is shown in the attached figure. Figure 2 As shown; Pipeline detection unit 2: located outside the pipeline (if attached Figure 3 The structure shown includes a positioning layer 14, a protective layer 15, and a protective layer 16 that are sequentially connected from the inside to the outside along the radial direction of the pipeline.

[0033] Positioning layer 14: spirally wound distributed single-mode optical fiber, coated with a color-developing layer (which changes color in the event of oil and gas leakage), and an integrated OTDR module (wavelength 1550nm, sampling rate 10kHz) at the end of the optical fiber.

[0034] Protective layer 15: PVDF film (50 μm thick) adheres closely to the outer wall of the pipe, and the surface is covered with a diamond-shaped conductive grid (2 cm side length). The grid nodes are connected to the microprocessor (STM32 series) through copper foil electrodes.

[0035] Protective layer 16: The outer layer is covered with an anti-corrosion polyurethane coating and has a copper foil grounding wire embedded inside.

[0036] Edge detection unit 13: deployed at both ends of the pipeline, integrating a MEMS accelerometer (to monitor vibration), a PT100 temperature sensor, a pressure sensor, and a flow meter; Area detection unit 18: Located at the junction node of the monitoring area (such as the junction of area AB), it integrates the LoRa wireless communication module and GPS positioning module for global status perception.

[0037] 1. Dynamic detection cycle optimization: Adjacent area monitoring platforms (such as Area A and Area B) achieve data sharing and task collaboration by deploying smart contracts (chain code based on Hyperledger Fabric) to form a redundant blockchain.

[0038] In the "Dynamic Allocation of Redundant Blockchain Tasks" section, change "Area B Platform takes over 30% of its tasks through smart contracts" to a specific description of the smart contract's execution logic. For example, the Area B platform automatically triggers the task migration process through a predefined smart contract (such as a load balancing contract) and takes over 30% of Area A's data analysis tasks.

[0039] Multi-level verification mechanism: "If the vibration energy ratio of a pipeline increases by 20% for three consecutive hours, the regional monitoring platform will determine it as an abnormal pipeline. It will further obtain the flow data of the adjacent monitoring area (such as area B) and the pressure fluctuation data of the edge detection unit. If there is a temporal and spatial match (such as the flow drop is synchronized with the vibration anomaly), the leak is confirmed."

[0040] Dynamic monitoring mechanism: "For high-risk pipelines with confirmed leakage (such as P3), the detection cycle is shortened from 30 minutes to 5 minutes; for low-risk pipelines (data fluctuations of ±5%), it is extended to 1 hour."

[0041] 2. Leakage monitoring and early warning process: Leak Alert Confirmation: If a pipeline's operating data exceeds the normal threshold and indicates an anomaly, the pipeline is initially identified as an anomaly. Further verification is conducted based on whether the overall operating data trend within the monitoring area is abnormal. For example, if the vibration energy percentage of a pipeline increases by 20% for three consecutive hours, the pipeline is initially identified as an anomaly. Furthermore, if the overall flow rate data of the pipeline, as monitored by edge monitoring units located at both ends of the anomaly pipeline, decreases, and the pressure at both ends fluctuates abnormally, the pipeline is further verified (determined) as high-risk and a leak alert is issued. Based on the verification results, the regional monitoring platform dynamically adjusts the anomaly detection cycle to strengthen monitoring of high-risk pipelines and identify the leak location. For pipelines identified as high-risk, the detection cycle is shortened; for pipelines identified as low-risk, the detection cycle is extended. For example, if a pipeline (such as P3) is identified as high-risk, the detection cycle is shortened from 30 minutes to 5 minutes. If the data percentage fluctuation range of a pipeline is within ±5%, indicating a low-risk status, the detection cycle is extended to 1 hour.

[0042] At the same time, the regional monitoring platform stores the monitoring data and test results in its redundant blockchain, using the characteristics of the redundant blockchain to ensure the security and integrity of the data, facilitating subsequent query, analysis and processing.

[0043] At the same time, the regional monitoring platform collects real-time monitoring data (such as pressure, flow, temperature, vibration, etc.) and operating parameters from each pipeline within its jurisdiction, compiling them into a first summary. It dynamically calculates the changing trend of each pipeline's monitoring data and operating parameters in the first summary. The central platform aggregates the workload and operating status of redundant blockchains globally to form a second summary. It dynamically calculates the workload and operating status of each redundant blockchain within the entire central platform, as well as the changing trend of its operating status within the second summary. Based on this, it dynamically adjusts the jurisdiction and commands of each redundant blockchain. Based on the workload and operating status of both the redundant blockchain and the regional monitoring platform, it decides whether to directly manage the redundant blockchain or the regional monitoring platform. The redundant blockchain dynamically adjusts the monitoring tasks and management authority of the regional monitoring platform based on the regional monitoring platform's workload and operating status within the redundant blockchain. Furthermore, based on the central platform's decision-making instructions, it takes over management tasks from regional monitoring platforms within or outside its jurisdiction.

[0044] 3. Redundant blockchain tasks and permissions management: The redundant blockchain dynamically adjusts the monitoring tasks and management permissions of regional monitoring platforms based on their respective workload and operational status within the redundant blockchain. If a regional monitoring platform experiences excessive workload or poor performance, the redundant blockchain can allocate some of its tasks to other regional monitoring platforms with better performance and less workload. Based on the central platform's decisions, the redundant blockchains will take over management tasks from regional monitoring platforms within or outside of their jurisdiction. For example, if a redundant blockchain has a high workload but performs well, the central platform will select a neighboring redundant blockchain with a lower workload and better performance based on the global workload and operational status of the redundant blockchains. If all neighboring redundant blockchains are fully loaded, the central platform will allocate some of its tasks to the nearest redundant blockchain, assuming proximity.

[0045] For example, dynamic allocation of redundant chain tasks: Redundant blockchains adjust task permissions based on the load of each platform (CPU utilization > 80%): Load balancing: When the platform in Area A is overloaded (e.g., CPU utilization exceeds 80%), the platform in Area B takes over 30% of its tasks (e.g., taking over data analysis in Pipeline 3) through smart contracts.

[0046] Global coordination: The central platform monitors the proportion of redundant blockchain tasks. If a redundant blockchain (such as chain AB) accounts for more than 60%, some tasks under its jurisdiction will be migrated to the low-load chain (such as redundant blockchain C or redundant blockchain D). Generally, the distance to the chain with the smallest proportion is the first consideration, and the distance to the chain with the smallest proportion is the second consideration.

[0047] 4. Leakage location and emergency linkage: Leak trigger: The OTDR module detects a sudden change in the optical fiber light intensity attenuation gradient (attenuation rate > 0.5 dB / km), triggering a level 1 warning.

[0048] The PVDF film simultaneously captures the abnormal vibration spectrum (energy rise of 30% in the 10Hz to 200Hz frequency band), and the microprocessor activates the positioning mode.

[0049] Precise positioning: Fiber optic positioning: The OTDR determines the distance from the leak point to the starting point. It then compares the vibration signal delay at each end of the pipeline and calculates the ratio of the distance between the leak point and the two ends. This outputs the spatial coordinates of the leak point and simultaneously sends feedback to the central platform, which dispatches the nearest repair team.

[0050] The dynamic adjustment of the jurisdiction of redundant blockchains is as follows: A new pipeline in Monitoring Area D is added, and the load on Regional Monitoring Platform C, previously in Monitoring Area C, reaches 45%. The central platform then assigns Monitoring Area D to the redundant blockchain governing Monitoring Area B, migrating 20% ​​of Monitoring Area B's tasks to Monitoring Area A. The original redundant blockchain, Block 1, then forms a new redundant blockchain governing Monitoring Area A, part of Monitoring Area B, and Monitoring Area D. The GIS interface uses color to indicate the dynamically adjusted areas.

[0051] Dynamic adjustments to the jurisdiction of monitoring regional platforms, specifically: A minor leak occurs in Pipeline 6 in Area B (a sudden 25% increase in vibration energy), but Regional Monitoring Platform B responds late due to a hardware failure. Redundant Blockchain Block 1 detects a timeout in the Regional Monitoring Platform's zone task, triggering Regional Monitoring Platform A in Area A to take over monitoring authority for Pipeline 6 and increase the detection period for Pipeline 6.

[0052] 5. Data model update and monitoring accuracy improvement: It should be noted in this embodiment that: In this application, the implementation steps of the self-learning module update are: 1. Data collection and decision database construction: Regional monitoring platform: real-time collection of pipeline vibration spectrum, light intensity attenuation gradient, pressure, temperature and flow data; Redundant blockchain: records task allocation (task volume ratio, processing time, regional load status); Central platform: stores leak verification results (success / false positive / missed negative), positioning error values, and historical threshold adjustment records.

[0053] Database Design: Fields include: timestamp, pipeline ID, flow rate drop value, vibration spectrum characteristics, task priority coefficient, task processing time, false alarm flag, positioning error, etc. Utilize the distributed ledger characteristics of blockchain to store data and ensure immutability and traceability.

[0054] 2. Feature Engineering and Model Input: Perform key feature extraction: Traffic drop characteristics: the percentage of deviation between the current traffic and the historical average, and the traffic change rate; Vibration spectrum characteristics: energy proportion in the 10Hz~200Hz frequency band, distribution of spectrum mutation points; Task allocation characteristics: regional platform task volume ratio (formula: task volume ratio = processing time × priority coefficient / maximum load capacity), redundant blockchain load heat map; Environmental factors: geological risk level of the pipeline and historical leakage frequency.

[0055] Tag definition: Supervised learning labels: true leak events (1) and false positive events (0); Reinforcement Learning Reward Signal: Successfully detected a leak: +1; False positive: -0.5; Missing: -1; For every 0.1 meter decrease in positioning error: +0.2.

[0056] 3. Machine Learning Model Design: Model Architecture: Dual-model synergy mechanism: Supervised learning model (threshold optimization): uses a random forest or XGBoost classification model, inputs traffic reduction characteristics and environmental factors, and outputs dynamic threshold recommendations (for example, adjusting the traffic reduction threshold from 5% to 4.2%). Reinforcement learning model (task allocation optimization): Based on DQN (Deep Q Network), it inputs task allocation characteristics and real-time load data and outputs the optimal task migration path (for example, migrating 30% of tasks in area A to area B).

[0057] Training process: In the initial stage, historical data is used to pre-train the model; During the online learning phase, the model dynamically updates parameters based on real-time data (e.g., iterates once every 24 hours).

[0058] 4. Dynamic optimization and deployment: Dynamic adjustment of threshold: The supervised learning model analyzes traffic data and false alarm rates in real time. If the false alarm rate increases, the threshold is automatically tightened (for example, the traffic drop threshold is lowered from 5% to 4%). Combined with positioning error feedback, adjust the vibration spectrum matching weight (such as increasing the weight of the 10-50Hz frequency band).

[0059] Task allocation optimization: The reinforcement learning model generates a task migration strategy based on the load heat map and calculates the optimal path using a gradient descent algorithm (formula: minimize objective function = ∑(task processing time × priority coefficient) + load imbalance penalty term). The central platform writes the optimization strategy into the smart contract, triggering the automatic migration of redundant blockchains.

[0060] 5. Closed-loop feedback and continuous learning: Feedback mechanism: After each leak event is handled, the actual results (whether there is a leak, positioning accuracy) are recorded and the database is updated; The model recalculates thresholds and task allocation strategies based on new data, forming a closed loop of "monitoring-decision-feedback-optimization".

[0061] Model evaluation and iteration: Key indicators: false alarm rate, missed alarm rate, positioning error, and task response delay; If the model performance deteriorates (for example, the false alarm rate increases by 10%), manual intervention or rollback to a historical stable version will be triggered.

[0062] Ultimately, each time a leak is detected, the initial prediction data, final confirmation results, and the corresponding decision-making process are recorded. This data is fed back into the data model to update and optimize the model.

[0063] This system utilizes a three-tiered hybrid architecture comprised of a central platform, redundant blockchain, and regional monitoring platform, achieving an organic combination of global coordination and regional autonomy through the construction of a three-tiered collaborative system. As the core of the system, the central platform is responsible for global resource allocation, task decision-making, and visual management and control. It generates load heat maps in real time and optimizes task allocation paths using a gradient descent algorithm, ensuring optimal resource allocation at a macro level. The redundant blockchain, formed by interconnecting adjacent regional monitoring platforms through smart contracts, is divided into a core chain and a backup chain. The core chain prioritizes high-risk pipeline monitoring tasks, while the backup chain is activated when the core chain's load exceeds a threshold. This design establishes a dynamic redundant backup mechanism, effectively preventing system failures caused by single points of failure. The regional monitoring platform, deployed in independent monitoring areas, includes multiple detection units that collect pipeline operation data in real time and trigger a multi-level verification mechanism, enabling precise monitoring and rapid response within the region.

[0064] In some embodiments, the present application has a multi-level detection unit collaborative positioning mechanism, where pipeline detection units are arranged along the extension direction of the pipeline, edge detection units are set at both ends of the pipeline, and the edge detection units located at the nodes of the monitoring area constitute the regional detection units, forming a three-level monitoring network. The pipeline detection unit includes nested modules of positioning layer, monitoring layer and protection layer. The positioning layer realizes light intensity attenuation gradient detection through OTDR module and detection optical fiber. The monitoring layer uses the piezoelectric effect of PVDF film to capture vibration signals, and the protective layer ensures signal transmission and equipment protection. The edge detection unit integrates MEMS accelerometer, temperature sensor, pressure sensor and flow meter to monitor the vibration, pressure, flow and other parameters at both ends of the pipeline in real time. The regional detection unit is located at the junction node of the monitoring area and integrates LoRa wireless communication module and GPS positioning module to realize global status perception.

[0065] Multi-level detection units work together to achieve dual verification and precise positioning of leak points. When the pipeline detection unit detects that the vibration energy exceeds the first threshold, a first-level warning is triggered. The edge detection unit simultaneously analyzes the vibration signal delay at both ends of the pipeline and calculates the distance between the leak point and the near end. The regional detection unit verifies the time-space matching in conjunction with the GIS system. For example, the OTDR module determines the distance L1 between the leak point and the starting point of the optical fiber, compares the vibration signal delay Δt at both ends, and calculates L2 = (Δt × sound speed) / 2. The coordinates (L1, L2) are output to the GIS system, and the positioning accuracy can reach ±0.5 meters. At the same time, through the global perception of the regional detection unit, related anomalies in adjacent areas can be discovered in a timely manner. For example, a leak can be confirmed when the difference between the flow valley time and the vibration peak time is less than 10 seconds, further improving the accuracy and reliability of monitoring.

[0066] In some embodiments, the present application has a collaborative anti-interference design of the positioning layer and the monitoring layer. The positioning layer includes an OTDR module, a color development layer and a detection optical fiber. The color development layer covers the outer wall of the pipe, the detection optical fiber is arranged on the outside, and the OTDR module is integrated at the end, which can detect leaks through the light intensity attenuation gradient. The monitoring layer includes a microprocessor, a PVDF film and a conductive grid. The PVDF film is tightly attached to the inner surface of the protective layer, and the polarization direction is perpendicular to the pipe surface. The conductive grid covers the side close to the positioning layer. The microprocessor is embedded at the edge and captures vibration signals through the piezoelectric effect. The protective layer is an anti-corrosion polyurethane layer, and copper foil electrodes are embedded on the inner surface, which are evenly distributed along the circumference and connected to the conductive grid of the positioning layer to eliminate electromagnetic interference.

[0067] The collaborative anti-interference design of the positioning and monitoring layers significantly improves the system's signal quality and monitoring reliability. When the PVDF film captures the vibration signal, a microprocessor extracts covariance features, including the CIR tap autocovariance matrix of the normal vibration spectrum (training set) and the cross-covariance vector between the abnormal state and the training set (target set). A linear prediction program is used to calculate the vibration spectrum residual. When the residual exceeds a second threshold, OTDR positioning is triggered. The copper foil electrodes in the protective layer improve the signal-to-noise ratio to ≥30dB, a 20dB improvement compared to traditional designs without anti-interference, effectively suppressing the impact of soil electromagnetic interference and industrial noise on the monitoring signal. This collaborative mechanism enables dual verification of leakage patterns. The monitoring layer amplifies the vibration signature of the defect location, while the spatial and temporal matching of the light intensity attenuation gradient with the vibration spectrum ensures accurate leak detection. For example, in oilfields with strong electromagnetic interference, traditional monitoring systems often suffer from signal distortion, leading to false alarms or missed detections. However, this design, through electromagnetic shielding and signal enhancement provided by the copper foil electrodes, ensures stable operation in complex electromagnetic environments, providing a reliable signal foundation for leak detection.

[0068] In some embodiments, the present application has a dynamic authority allocation and task migration mechanism, which can dynamically migrate tasks to low-load nodes based on the proportion of regional platform tasks and the proportion of blockchain tasks. When the proportion of regional platform tasks is greater than 80% or the CPU utilization rate is greater than the safety threshold, the smart contract will migrate the excess tasks to the adjacent low-load platform. The central platform dynamically adjusts the redundant blockchain jurisdiction boundaries based on the global load heat map, and newly added monitoring areas are assigned to the low-load blockchain according to distance priority. The migration task volume satisfies the migration volume = original task volume × (1-remaining load rate of the target blockchain). The task volume ratio calculation combines the task processing time, priority coefficient, regional platform maximum load capacity and blockchain weight coefficient, where the priority coefficient and weight coefficient are dynamically adjusted based on the pipeline risk level, historical success rate, etc.

[0069] The dynamic permission allocation mechanism enables intelligent scheduling and load balancing of system resources, fundamentally changing the resource rigidity of traditional technologies. Traditional monitoring systems often use a fixed task allocation model, which often results in overloaded platforms in some areas while others remain idle. This mechanism, however, monitors the load status of each node in real time. When the workload of a platform in a certain area exceeds 80%, it triggers a task migration process within 10 seconds. For example, 30% of data analysis tasks in Area A could be migrated to Area B, improving overall system load balance by 60%. Using a global load heat map, the central platform dynamically adjusts the jurisdictional boundaries of redundant blockchains. For example, when a new pipeline in Area D is added, it can be assigned to the blockchain in Area B, which has a 45% load factor. Simultaneously, 20% of tasks in Area B can be migrated to Area A to ensure balanced load across all blockchains. This dynamic adjustment mechanism not only improves resource utilization but also enhances system scalability. When new monitoring areas or pipelines are added, the system automatically optimizes task allocation without manual intervention, significantly reducing operational and maintenance costs. In addition, the dynamic calculation of task priority coefficients enables monitoring tasks for high-risk pipelines (such as areas with a high historical incidence of leakage and high-pressure oil pipelines) to be prioritized, ensuring the timeliness and reliability of monitoring in key areas.

[0070] In some embodiments, the present application features a multi-level verification mechanism and dynamic monitoring cycle adjustment. When a regional platform detects a pipeline vibration energy increase of >20% for three consecutive hours, it is marked as an abnormal pipeline and the adjacent regional flow data and edge unit pressure fluctuation values ​​are obtained. If the spatial and temporal matching conditions of regional flow drop > dynamic threshold and the time difference between flow valley and vibration peak is <10 seconds are met, a leak is determined and positioning mode is activated. Otherwise, the detection cycle is extended to 1 hour. Based on the verification results, the detection cycle is shortened for high-risk pipelines and extended for low-risk pipelines, achieving dynamic optimization of monitoring resources.

[0071] A multi-level verification mechanism significantly improves the accuracy of leak detection and reduces false alarms through multi-dimensional data correlation analysis. This mechanism reduces the false alarm rate to below 1% by leveraging multiple verification methods, including vibration energy growth, flow rate drop, pressure fluctuation, and temporal correlation. For example, if the vibration energy of a pipeline increases by 25% for three consecutive hours, the regional platform initially flags it as an anomaly. It then collects flow data from adjacent areas. If the flow rate drop exceeds a dynamic threshold (e.g., 5%) and the time difference between the flow valley and the vibration peak is 8 seconds, a leak is confirmed. If these conditions are not met, the detection cycle is extended from 30 minutes to 1 hour to avoid false alarms caused by brief fluctuations. Dynamic monitoring cycle adjustment enables precise allocation of monitoring resources. For high-risk pipelines (such as Pipeline 3), the detection cycle is shortened to 5 minutes to promptly detect the development of minor leaks. For low-risk pipelines, the detection cycle is extended to 1 hour, reducing unnecessary testing frequency and boosting overall monitoring efficiency by 50%. This also reduces sensor energy consumption and data processing pressure. This "precise monitoring + intelligent optimization" model not only ensures the reliability of monitoring, but also improves the operating efficiency of the system, providing technical support for the refined operation and maintenance of oilfield pipelines.

[0072] In some embodiments, the present application has a self-learning module and a machine learning optimization mechanism. The self-learning module builds a decision result database to record false alarm rates, positioning errors, etc., and uses a reinforcement learning model to dynamically optimize the flow reduction threshold and vibration spectrum matching weight. The supervised learning model uses a random forest algorithm to analyze historical flow reduction characteristics and environmental factors, and outputs dynamic threshold adjustment suggestions. For every 1% increase in false alarm rate, the threshold is tightened by 0.8%. The reinforcement learning model generates a task migration path based on the DQN network. The reward function R = number of successful detections × 1 + positioning error reduction value × 0.2 - number of false alarms × 0.5. The model parameters are iteratively updated every 24 hours, and manual intervention is triggered when the positioning error is greater than 1 meter.

[0073] The self-learning module uses machine learning algorithms to enable the system to autonomously optimize and continuously evolve, overcoming the inability of traditional technologies to adapt to changing environments. Traditional monitoring systems often have fixed thresholds and strategies, making them incapable of adapting to dynamic changes in pipeline operating conditions and environmental factors, leading to a gradual decline in monitoring accuracy over time. This module, however, continuously accumulates historical data (such as leak verification results, location errors, and task allocation records) and uses a random forest algorithm to analyze the correlation between flow rate drops and environmental factors. It then automatically adjusts dynamic thresholds. For example, if the false alarm rate increases by 1%, the flow rate drop threshold is tightened from 5% to 4.2%, bringing the false alarm rate back below 1%. The reinforcement learning model uses a reward function to guide the optimization of the task transfer strategy. For example, a successful leak detection receives a +1 reward, while a false alarm receives a -0.5 deduction. This encourages the model to continuously learn the optimal task allocation path, reducing task response latency by 30% and reducing location error by an average of 0.3 meters. Model updates are iterated every 24 hours to ensure the system can adapt to new leak patterns and operating environments. For example, if seasonal temperature fluctuations cause changes in pipeline vibration characteristics, the model can automatically adjust the vibration spectrum matching weights to maintain stable monitoring accuracy. This self-learning capability enables the system to upgrade from "passive monitoring" to "active optimization". As the operating time increases, the monitoring performance continues to improve, providing strong guarantees for the long-term safe operation of oilfield pipelines.

[0074] In some embodiments, the present application has a redundant blockchain authority hierarchy and atomic transaction log mechanism. The redundant blockchain adopts authority hierarchical management. The core chain takes over the monitoring tasks of historical leakage high-incidence areas, cross-border pipelines or high-pressure oil pipelines first. The backup chain is in low-power listening mode under normal circumstances. When the core chain load exceeds the safety threshold or the regional platform fails, it is activated and takes over the task through the consensus algorithm. Each blockchain node maintains a global transaction log, stores task descriptors and atomic TLE flags, scans the log identification state discontinuity when power is restored, determines the valid task and restores it to the migration queue. The atomic TLE flag enhancement mechanism includes weak clear and weak set flags, applying a +0.5V bias or injecting 5mA current to reduce the bit error rate to <10^-6.

[0075] The redundant blockchain's permission hierarchy and atomic transaction log mechanism significantly enhances system stability, reliability, and data security. Traditional distributed systems lack effective permission hierarchy and transaction assurance, often leading to task loss or data inconsistencies. The hierarchical design of core and backup chains ensures that high-risk pipeline monitoring tasks are always handled by the core chain with the best performance, ensuring the reliability of critical tasks. The atomic transaction log ensures the atomicity and consistency of tasks. Even in abnormal situations such as power outages, the system can reintegrate abnormal tasks into the migration queue through log scanning and recovery mechanisms, avoiding monitoring vulnerabilities caused by task loss. The atomic TLE flag enhancement mechanism further improves data transmission accuracy through piezoelectric film self-calibration and polarization reset, reducing the bit error rate by two orders of magnitude compared to traditional blockchains and ensuring the reliability and traceability of monitoring data.

[0076] In some embodiments, this application incorporates a pressure boost verification process and optimized positioning accuracy. When the flow difference ΔQ exceeds the flow drop threshold, the boost valve increases the pipeline pressure by 0.5 MPa and re-checks the flow difference ΔQ'. If the expansion rate = (ΔQ'-ΔQ) / ΔQ > 15%, a leak is confirmed, and positioning accuracy is improved to ±0.5 meters, with verification time taking less than 2 minutes. This process further confirms the authenticity of the leak and optimizes positioning accuracy by analyzing the impact of pressure changes on the leak.

[0077] The pressure-boost verification process provides a final confirmation mechanism for leak detection, resolving the difficulty traditional techniques present in diagnosing suspected leaks. In actual monitoring, false positive alarms can occur due to factors such as flow fluctuations and sensor errors. Traditional methods lack effective verification methods and often require time-consuming and labor-intensive manual inspections. This process, however, actively boosts the pressure by 0.5 MPa and observes changes in the flow differential. If the flow differential increases by more than 15%, it indicates a real leak. This is because increased pressure only leads to a significant increase in leakage when the pipeline is damaged. Furthermore, during the pressure-boost verification process, simultaneous analysis of the OTDR's light intensity attenuation gradient and vibration spectrum further pinpoints the leak, improving positioning accuracy from ±1 meter to ±0.5 meter, meeting the high-precision positioning requirements of oilfield pipeline maintenance. Verification takes less than two minutes, ensuring a timely emergency response. Once a leak is confirmed, the system can immediately dispatch a maintenance team, shortening troubleshooting time and minimizing losses caused by the leak. This process, combined with a multi-level verification mechanism and collaborative design of the positioning layer and monitoring layer, forms a complete technical chain from early warning, verification to positioning, and comprehensively improves the technical level of oilfield pipeline leakage monitoring.

Claims

1. An oilfield pipeline cluster leakage monitoring system, characterized in that: The system comprises: Central platform: performs global resource allocation, task decision-making, and visual management and control, generates load heat maps in real time, and optimizes task allocation paths using a gradient descent algorithm; Redundant blockchain: formed by interconnecting adjacent regional monitoring platforms through smart contracts, divided into a core chain and a backup chain. The core chain takes priority over monitoring tasks in areas with a high incidence of historical leaks, cross-border pipelines, or high-pressure oil pipelines. The backup chain is normally in low-power monitoring mode and is activated through a consensus algorithm when the core chain load exceeds the safety threshold. Regional monitoring platform: Deployed in independent monitoring areas, it includes multi-level detection units, collects pipeline operation data in real time and triggers a multi-level verification mechanism; Dynamic authority allocation module: dynamically migrate tasks to low-load nodes based on the proportion of regional platform tasks and blockchain tasks; Self-learning module: Build a decision result database to record false alarm rates and positioning errors, and use a reinforcement learning model to dynamically optimize the flow reduction threshold and vibration spectrum matching weight.

2. The oilfield pipeline cluster leakage monitoring system according to claim 1, characterized in that: The multi-stage detection unit includes a pipeline detection unit, an edge detection unit and an area detection unit; The pipeline detection units are arranged along the extension direction of the pipeline; the edge detection units are arranged at both ends of the pipeline; the edge detection units located at the nodes of the monitoring area constitute the area detection units; the edge detection units and the area detection units are used to locate each area and each pipeline and monitor the operating status of each area and each pipeline; the pipeline detection unit is used to monitor and determine the leaking pipeline and the leakage point of the leaking pipeline; The pipeline detection unit includes a positioning layer, a monitoring layer, and a protective layer, which are arranged in sequence from the inside to the outside along the radial direction of the pipeline; the positioning layer, the monitoring layer, and the protective layer form a nested module; the protective layer ensures that the vibration energy of the pipeline is transmitted to the monitoring layer; the monitoring layer captures the vibration of the pipeline in real time through the piezoelectric effect, obtains monitoring data, and decides whether to issue a detection warning by comparing the monitoring data with a first threshold; after the positioning layer issues a detection warning, the monitoring layer is activated and enters a positioning mode. The monitoring layer amplifies the defect location on the one hand, and on the other hand, achieves dual verification of the leakage mode and accurately locates the leakage location by temporal and spatial matching of the light intensity attenuation gradient with the vibration spectrum of the positioning layer on the other hand; The positioning layer includes an OTDR module, a color development layer located outside the pipeline, and a detection optical fiber; the color development layer covers the entire outer wall of the pipeline; the detection optical fiber is arranged outside the color development layer; the OTDR module is integrated at the end of the detection optical fiber; The monitoring layer includes a microprocessor located at the end of the pipeline, as well as a PVDF film and a conductive grid outside the monitoring layer. The PVDF film is tightly attached to the inner surface of the protective layer, with its polarization direction perpendicular to the pipeline surface. The conductive grid covers the side of the PVDF film closest to the positioning layer, and the microprocessor is embedded in the edge of the PVDF film. The monitoring layer amplifies the defect area and, through spatiotemporal matching of the light intensity attenuation gradient with the vibration spectrum of the positioning layer, achieves dual verification of the leakage mode and accurately locates the leak. The protective layer is an anti-corrosion polyurethane layer, which is evenly wrapped around the entire monitoring layer to protect the monitoring layer and the positioning layer from soil erosion. The inner surface of the protective layer is embedded with copper foil electrodes, which are evenly distributed along the circumference of the pipeline and are used to connect the conductive grid of the positioning layer.

3. The oilfield pipeline cluster leakage monitoring system according to claim 2, characterized in that: The coordination mode between the positioning layer and the monitoring layer is as follows: After the PVDF film captures the vibration signal, the microprocessor extracts the covariance features, including the training set and the target set; the training set is the CIR tap autocovariance matrix of the normal state vibration spectrum; the target set is the cross-covariance vector of the abnormal state vibration spectrum and the training set; The vibration spectrum residual is calculated through a linear prediction program, and OTDR positioning is triggered when the residual is greater than the second threshold; The copper foil electrode of the protective layer eliminates electromagnetic interference and improves the signal-to-noise ratio to ≥30dB.

4. The oilfield pipeline cluster leakage monitoring system according to claim 2, characterized in that: The monitoring layer performs: The pipeline vibration is captured in real time through the piezoelectric effect, and a first-level warning is triggered when the vibration energy exceeds the first threshold. After entering the positioning mode, the OTDR light intensity attenuation gradient and vibration spectrum are analyzed synchronously, and dual leakage verification is achieved through time and space matching; The calculation method of the leakage point coordinates is: Step 1: Use OTDR to determine the distance L1 between the leakage point and the starting point of the optical fiber. Step 2: Compare the time delay Δt of the vibration signals at both ends of the pipeline and calculate the distance between the leakage point and the near end, L2 = (Δt × speed of sound) / 2; Step 3: Output coordinates (L1, L2) to the GIS system.

5. The oilfield pipeline cluster leakage monitoring system according to claim 1, characterized in that: The dynamic rights allocation module includes: When the task volume of a regional platform exceeds 80% or the CPU utilization exceeds the safety threshold, the smart contract will migrate the excess tasks to an adjacent low-load platform. Based on the global load heat map, the central platform dynamically adjusts the jurisdiction boundaries of redundant blockchains as follows: when adding new monitoring areas, they are assigned to low-load blockchains based on distance priority; when migrating tasks, the migration volume is satisfied: the migration volume = original task volume × (1-remaining load rate of the target blockchain); The dynamic authority allocation module further includes: allocating monitoring authority, including: When the task load of a regional monitoring platform exceeds a preset threshold or the operating status is abnormal, the redundant blockchain will migrate some tasks to an adjacent low-load platform through a smart contract; The central platform divides redundant blockchains into core chains and backup chains based on the global task volume distribution. The core chain takes over high-priority tasks first, while the backup chain responds to sudden loads. The central platform also adjusts the boundaries of the redundant blockchain's jurisdiction in real time based on its task volume ratio, operating status, and regional feedback data. The calculation method of the task volume ratio is: regional monitoring platform task volume ratio = (current task processing time × task priority coefficient) / regional platform maximum load capacity; The proportion of redundant blockchain tasks = ∑ (the proportion of tasks in the platform in the corresponding area) × blockchain weight coefficient; wherein, the task priority coefficient is dynamically calculated based on the pipeline leakage risk level, task type weight and environmental sensitivity, and the blockchain weight coefficient is dynamically adjusted based on the historical task success rate, current load rate and node stability score.

6. The oilfield pipeline cluster leakage monitoring system according to claim 1, characterized in that: The multi-level verification mechanism is: When the regional platform detects that the pipeline vibration energy increases by more than 20% for three consecutive hours, it is marked as an abnormal pipeline; Obtain the flow data of adjacent areas and the pressure fluctuation value of edge units, if the spatiotemporal matching conditions are met: Condition 1: Regional traffic decrease > dynamic threshold; Condition 2: The time difference between the flow valley value and the vibration peak value is less than 10 seconds; If there is a leak, the detection period is extended to 1 hour.

7. The oilfield pipeline cluster leakage monitoring system according to claim 1, characterized in that: The self-learning module performs: Supervised learning model: Uses a random forest algorithm to analyze historical traffic drop characteristics and environmental factors, and outputs dynamic threshold adjustment recommendations. For every 1% increase in false alarm rate, the threshold is tightened by 0.8%; Reinforcement learning model: Generate task migration paths based on the DQN network, with a reward function R = number of successful detections × 1 + positioning error reduction × 0.2 - number of false alarms × 0.5; The model parameters are iteratively updated every 24 hours, and manual intervention is triggered when the positioning error is greater than 1 meter.

8. The oilfield pipeline cluster leakage monitoring system according to claim 1, characterized in that: The central platform updates and displays in real time each reallocated monitoring area and the corresponding management area monitoring platform, as well as the area monitoring governed by the redundant blockchain.

9. The oilfield pipeline cluster leakage monitoring system according to claim 1, characterized in that: The updating of the self-learning module includes: Build a decision result database to record the success rate, false alarm rate and positioning error of leak verification; A reinforcement learning model is used to dynamically adjust the flow reduction threshold, vibration spectrum matching weight and task allocation strategy with the goal of minimizing the false alarm rate and positioning error.

10. The oilfield pipeline cluster leakage monitoring system according to claim 2, characterized in that: The hierarchical management of permissions for the redundant blockchain is specifically as follows: The core chain will prioritize monitoring tasks in areas with a high incidence of historical leaks, cross-border pipelines, or pipelines transporting high-pressure oil products; The backup chain is normally in low-power monitoring mode. When the core chain load exceeds the safety threshold or the regional platform fails, it is activated and takes over the task through the consensus algorithm; The redundant blockchain uses an atomic transaction log: Each blockchain node maintains a global transaction log, where entries store task descriptors and atomic TLE flags; Power failure recovery process: Step 1: Scan the global transaction log entries to identify TLE flag state discontinuities; Step 2: Determine the LTE adjacent to the discontinuity as a valid task; Step 3: weak programming flag execution is enhanced to a stable state; Step 4: Restore abnormal tasks from LTE to the migration queue; The atomic TLE flag enhancement mechanism includes: Weak clear mark: Apply +0.5V bias to the copper foil electrode for 10ms to reset the polarization direction of PVDF; Weak setting flag: injecting 5mA current through the conductive grid triggers the piezoelectric film to self-calibrate; After enhancement, the bit error rate is reduced to <10 6 ; The pressure boost verification process is as follows: When the flow difference ΔQ is greater than the flow drop threshold, the booster valve increases the pipeline pressure by 0.5 MPa; Re-test the flow difference ΔQ'. If the expansion rate = (ΔQ'-ΔQ) / ΔQ>15%, leakage is confirmed. Positioning accuracy is improved to ±0.5 meters, and verification takes less than 2 minutes.

Images