System and method for monitoring and evaluating corrosion of buried PE (Poly Ethylene) pipeline

Abstract

The invention relates to the field of pipeline engineering and intelligent detection, and discloses a buried PE pipeline corrosion monitoring and evaluating system which comprises a distributed optical fiber sensor module which is used for being installed on the outer wall of a buried PE pipeline, collecting pipeline temperature and stress change data and sensing the pipeline corrosion condition through optical fiber sensing signals; the invention further discloses a buried PE pipeline corrosion monitoring and evaluating method which comprises the following steps: S1, installing the distributed optical fiber sensors on the outer wall of the buried PE pipeline, and collecting temperature and stress change data of the pipeline in real time; through a distributed optical fiber sensor and an intelligent algorithm, real-time monitoring and accurate corrosion evaluation of the buried PE pipeline are realized, the monitoring efficiency is remarkably improved, a signal acquisition system based on the Internet of Things reduces the construction and maintenance cost, and meanwhile, the flexibility and data security of the system are enhanced; the multi-modal data fusion and deep learning technology optimizes corrosion risk prediction, and reduces misinformation and missing report.

Description

Technical Field

[0001] The present invention relates to the field of pipeline engineering and intelligent detection, and in particular to a buried PE pipeline corrosion monitoring and evaluation system and method. Background Art

[0002] At present, the corrosion monitoring of buried polyethylene (PE) pipelines mainly relies on regular manual inspections and excavation inspections. Although these methods can provide pipeline health status detection to a certain extent, their efficiency is low and it is difficult to achieve real-time monitoring. Manual inspections are usually based on periodic inspections, which cannot grasp the health status of the pipeline in a timely manner. There are problems such as long inspection cycles and large errors, which not only increase safety hazards but also increase maintenance costs. Especially in key areas such as urban gas pipelines, oil pipelines and underground pipelines in chemical parks, traditional monitoring methods cannot ensure real-time and comprehensive assessment of pipeline health status, and thus cannot effectively prevent potential safety accidents.

[0003] Most existing technical solutions use a combination of ground detection instruments and excavation verification. Although these technologies can detect pipeline corrosion to a certain extent, they rely on regular manual intervention and lack real-time and continuity. During the monitoring process, the detection instruments are greatly affected by factors such as terrain and soil conditions, resulting in limitations in data accuracy and stability. For example, when using ultrasonic or electromagnetic detection instruments to detect corrosion on the outer wall of a pipeline, the signal may be attenuated due to different geological structures, affecting the reliability of the monitoring results. In addition, these equipment and technologies usually require excavation or exposure of the pipeline, which is a cumbersome, time-consuming and costly process.

[0004] These traditional monitoring methods fail to address the need for real-time monitoring. Especially in a dynamically changing environment, regular inspections cannot detect potential corrosion risks in a timely manner, and it is difficult to continuously update the health status of the pipeline. Therefore, how to achieve real-time monitoring of buried PE pipelines and improve the accuracy of corrosion risk assessment has become an urgent problem to be solved. The existing technology is difficult to provide continuous monitoring and real-time early warning, resulting in the inability to effectively prevent potential corrosion problems from developing into accidents such as pipeline leakage or breakage, thereby increasing the difficulty and safety risks of pipeline maintenance. In response to the above problems, the present invention proposes a buried PE pipeline corrosion monitoring and evaluation system and method. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a buried PE pipeline corrosion monitoring and assessment system and method, which solves the problem that traditional buried PE pipeline corrosion monitoring methods cannot achieve real-time, continuous monitoring and accurate assessment of corrosion risks.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a buried PE pipeline corrosion monitoring and evaluation system, comprising: Distributed fiber optic sensor modules are installed on the outer wall of buried PE pipelines to collect data on pipeline temperature and stress changes, and sense pipeline corrosion through fiber optic sensing signals; The signal acquisition and edge computing module is used to receive and process the signal data transmitted by the optical fiber sensor, perform signal amplification, filtering, anomaly detection, and data compression before transmitting it to the cloud; Cloud data storage and analysis module, which is used to store sensor data and analyze the data based on machine learning algorithms to assess pipeline corrosion risks and generate risk assessment reports; Multimodal data fusion module, which is used to fuse data from multiple sensors including distributed fiber optic sensors, soil moisture sensors and pressure sensors, and conduct comprehensive analysis using deep learning algorithms for corrosion assessment; The corrosion assessment and early warning system module is used to generate a risk assessment report on pipeline corrosion based on the analysis results, and send out early warning signals to notify pipeline maintenance personnel.

[0007] Preferably, the distributed optical fiber sensor module comprises: Fiber optic sensor unit, used to sense temperature and stress changes on the outer wall of the pipeline based on the principle of Brillouin scattering; The sensing node unit is used to transmit the sensing signal of the optical fiber sensor to the signal acquisition module through wavelength division multiplexing technology; The signal processing unit is used to remove noise from the transmitted signal and convert it into standardized data.

[0008] Preferably, the signal acquisition and edge computing module includes: A signal amplification unit, used for amplifying the signal transmitted by the optical fiber sensor; A filter circuit unit is used to filter out high-frequency noise in the signal and retain the effective signal; an anomaly detection unit, for performing abnormal data detection; The data compression unit is used to reduce the amount of data transmitted to the cloud and effectively retain key data through compression processing.

[0009] Preferably, the cloud data storage and analysis module includes: A data storage unit, used for storing the data transmitted by the signal acquisition module into a distributed database; A data analysis unit, which is used to perform corrosion risk analysis on the stored data based on machine learning algorithms to assess the health status of the pipeline; The historical data processing unit is used to train and optimize the machine learning model using historical data.

[0010] Preferably, the multimodal data fusion module includes: a multi-sensor data input unit for transmitting data from multiple sensors including a distributed optical fiber sensor, a soil moisture sensor, and a pressure sensor to a fusion module; A data fusion algorithm unit is used to perform weighted fusion of sensor data from different sources; The deep learning analysis unit is used to analyze the fused data through a deep learning algorithm to perform corrosion assessment.

[0011] Preferably, the data fusion algorithm unit uses the following formula to perform weighted fusion of sensor data from different sources: D fusion =αD fiber +βD soil +γD flow Among them, D fiber , D soil , D flow are the data from different sensors, and α, β, and γ are weighting coefficients.

[0012] Preferably, the corrosion assessment and early warning system module includes: A corrosion assessment unit, which is used to assess the corrosion risk level of the pipeline based on the analysis results and generate a risk assessment report; An early warning signal generation unit is used to trigger different levels of early warning signals according to the evaluation results, including low risk, medium risk and high risk. The signals are sent to pipeline maintenance personnel via SMS, email or mobile application; The threshold adjustment unit is used to dynamically adjust the warning threshold according to historical data and external environmental factors.

[0013] A method for monitoring and evaluating the corrosion of a buried PE pipeline, used in conjunction with the above-mentioned system for monitoring and evaluating the corrosion of a buried PE pipeline, comprises the following steps: S1. Install the distributed optical fiber sensor on the outer wall of the buried PE pipeline to collect the temperature and stress change data of the pipeline in real time; S2, amplify and filter the sensor signal through the signal acquisition and edge computing module, apply the local anomaly factor algorithm to perform preliminary anomaly detection on the data, and compress the abnormal data and upload it to the cloud; S3, storing the uploaded data in the cloud data storage and analysis module, and using machine learning models to analyze the corrosion risk of the pipeline; S4. The data from different sensors are fused through the multimodal data fusion module, and the corrosion risk is evaluated using the deep learning algorithm to improve the evaluation accuracy; S5. Generate a risk assessment report on pipeline corrosion through the corrosion assessment and early warning system module, and trigger different levels of early warning signals according to preset thresholds to notify pipeline maintenance personnel. Preferably, the signal acquisition and edge computing module in step S2 uses the following LOF anomaly detection algorithm when performing signal processing: Among them, p is the current data point, N k (p) is the k nearest neighbors of p, and lrd(p) is the local density of point p.

[0014] Preferably, the risk assessment report in step S5 includes the current corrosion status of the pipeline, risk prediction results and pipeline maintenance recommendations, and the threshold in step S5 is set based on monitoring data of temperature, stress and pressure changes.

[0015] The present invention provides a buried PE pipeline corrosion monitoring and evaluation system and method. It has the following beneficial effects: 1. The present invention realizes all-weather, real-time monitoring of the corrosion of buried PE pipelines by adopting distributed optical fiber sensors and intelligent monitoring algorithms. Through the application of distributed optical fiber sensors, the health status of the pipeline can be monitored at any time. The optical fiber sensor is based on the Brillouin scattering principle and can accurately capture small changes in temperature and stress, providing key data for the discovery of early corrosion. Compared with traditional regular manual inspections and excavation inspections, the present invention provides a more efficient and rapid monitoring method, significantly improving the monitoring response speed. Real-time data collection and processing enable the corrosion status of the pipeline to be reflected in a timely manner, providing timely and accurate information for pipeline maintenance, helping to quickly take measures, reduce potential risks, and avoid accidents.

[0016] 2. The present invention can significantly reduce the construction and maintenance costs of the pipeline monitoring system by constructing a signal acquisition and transmission system based on the Internet of Things. The Internet of Things technology realizes remote wireless transmission of data between sensors, reduces the deployment and maintenance costs of traditional wired networks, and enhances the flexible deployment capabilities of the system. Each sensor can be freely deployed as needed without being restricted by the constraints of physical lines. In addition, the IoT architecture also effectively improves the data security of the system. Through encrypted transmission and data compression technology, the security and integrity of data during transmission are ensured to prevent data leakage or loss. This design not only improves the scalability of the system, but also provides a safer and more reliable technical guarantee for pipeline monitoring, and is particularly suitable for pipeline safety management in complex geographical environments.

[0017] 3. The present invention introduces multimodal data fusion technology. By weighted fusion of data from multiple sensors, a more comprehensive and accurate assessment of pipeline corrosion is achieved. Through multimodal data fusion, the system can comprehensively consider the feedback information of different sensors on the health status of the pipeline, eliminate the errors that may be caused by a single sensor, and improve the accuracy of corrosion risk prediction. At the same time, the application of deep learning algorithms effectively reduces the false alarm and missed alarm problems caused by data noise or external interference, thereby providing more reliable support for pipeline maintenance decisions. Based on accurate corrosion risk prediction, maintenance personnel can take corresponding measures in time, optimize maintenance strategies, and improve the overall safety and service life of the pipeline. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a framework diagram of the system of the present invention; Figure 2 It is a framework diagram of a distributed optical fiber sensor module of the present invention; Figure 3 It is a framework diagram of the signal acquisition and edge computing module of the present invention; Figure 4 This is a framework diagram of the cloud data storage and analysis module of the present invention; Figure 5 It is a framework diagram of the multimodal data fusion module of the present invention; Figure 6 It is a framework diagram of the corrosion assessment and early warning system module of the present invention; Figure 7 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0020] Please see attached Figure 1 The embodiment of the present invention provides a buried PE pipeline corrosion monitoring and evaluation system, comprising: Distributed fiber optic sensor modules are installed on the outer wall of buried PE pipelines to collect data on pipeline temperature and stress changes, and sense pipeline corrosion through fiber optic sensing signals; The signal acquisition and edge computing module is used to receive and process the signal data transmitted by the optical fiber sensor, perform signal amplification, filtering, anomaly detection, and data compression before transmitting it to the cloud; Cloud data storage and analysis module, which is used to store sensor data and analyze the data based on machine learning algorithms to assess pipeline corrosion risks and generate risk assessment reports; Multimodal data fusion module, which is used to fuse data from multiple sensors including distributed fiber optic sensors, soil moisture sensors and pressure sensors, and conduct comprehensive analysis using deep learning algorithms for corrosion assessment; The corrosion assessment and early warning system module is used to generate a risk assessment report on pipeline corrosion based on the analysis results, and send out early warning signals to notify pipeline maintenance personnel.

[0021] Specifically, the present invention provides a buried PE pipeline corrosion monitoring and evaluation system, which aims to ensure the safe operation of the pipeline through real-time monitoring, data analysis and intelligent early warning. The system adopts multiple technical means such as advanced distributed optical fiber sensing technology, edge computing, cloud data analysis and deep learning algorithms to comprehensively improve the accuracy and response speed of pipeline corrosion monitoring and evaluation.

[0022] First, the system installs fiber optic sensors on the outer wall of the buried PE pipeline through distributed fiber optic sensor modules. These sensors can collect temperature and stress change data of the pipeline in real time, and monitor the corrosion of the pipeline surface based on the Brillouin scattering principle. Distributed fiber optic sensors can collect key signals reflecting the health status of the pipeline at different locations of the pipeline and transmit these signals to subsequent processing modules.

[0023] Next, the signal acquisition and edge computing module pre-processes the collected signals. This module is responsible for signal amplification and filtering, removing noise, and ensuring the effectiveness of the signal. At the same time, the module also uses a local anomaly factor algorithm for anomaly detection to identify potential corrosion areas. Through edge computing, the system can perform preliminary analysis and compression of signal data, effectively reducing the amount of data transmission and ensuring the accuracy and real-time nature of information transmission.

[0024] The processed data is uploaded to the cloud data storage and analysis module. This module not only stores sensor data, but also conducts in-depth analysis of the data through integrated machine learning algorithms to assess the corrosion risk of the pipeline. These algorithms can make accurate judgments on the current corrosion status of the pipeline by learning from historical data, and generate detailed risk assessment reports to help maintenance personnel understand the health status of the pipeline.

[0025] On this basis, the system further combines the multimodal data fusion module to integrate data from different sensors, including distributed fiber optic sensors, soil moisture sensors, and pressure sensors. These data are comprehensively analyzed through weighted fusion and deep learning algorithms to improve the accuracy and reliability of corrosion assessment. This module not only enhances the ability to perceive the impact of the external environment on the pipeline, but also more accurately predicts potential corrosion risks.

[0026] Finally, the corrosion assessment and early warning system module generates a pipeline corrosion risk assessment report based on the analysis results, and generates early warning signals of different levels based on the set thresholds. These signals are notified to pipeline maintenance personnel via SMS, email or APP push. The module also has the ability to dynamically adjust thresholds to cope with changes in corrosion risks under different environmental conditions, thereby providing a more intelligent and flexible early warning mechanism.

[0027] Through the collaborative work of these modules, the buried PE pipeline corrosion monitoring and evaluation system of the present invention can achieve all-weather, real-time pipeline monitoring, accurately evaluate corrosion risks and provide timely warnings, greatly improving the efficiency and accuracy of pipeline safety management.

[0028] Please see attached Figure 2 , the distributed fiber optic sensor module includes: Fiber optic sensor unit, used to sense temperature and stress changes on the outer wall of the pipeline based on the principle of Brillouin scattering; The sensing node unit is used to transmit the sensing signal of the optical fiber sensor to the signal acquisition module through wavelength division multiplexing technology; The signal processing unit is used to remove noise from the transmitted signal and convert it into standardized data.

[0029] Specifically, the distributed optical fiber sensor module of the present invention plays a core role in the buried PE pipeline corrosion monitoring system, and is mainly used to realize the real-time perception of pipeline temperature and stress changes, and then identify the risk of pipeline corrosion. The module combines optical fiber sensing technology, wavelength division multiplexing technology and signal processing technology, and can efficiently and accurately collect data on the health status of the pipeline and provide reliable information for subsequent analysis. The optical fiber sensor module includes three main components: an optical fiber sensor unit, a sensor node unit and a signal processing unit. The units work together to ensure that the data collected from the pipeline surface is efficiently and accurately transferred, processed and analyzed in the system.

[0030] In this embodiment, the optical fiber sensor unit senses the temperature and stress changes of the outer wall of the pipeline through the Brillouin scattering principle. The Brillouin scattering principle uses the interaction between the light waves in the optical fiber and the acoustic waves on the surface of the pipeline, which causes the frequency of the light waves to shift. This frequency shift is proportional to the changes in temperature and stress. Therefore, by monitoring the Brillouin frequency shift, the temperature and stress changes of the pipeline can be inferred.

[0031] The working principle of fiber optic sensors is based on the physical properties of fiber optic materials and external temperature and stress changes. Specifically, when the light waves in the fiber optic sensor pass through the surface of the pipeline, the temperature and stress changes in the pipeline will cause a slight change in the frequency of the light waves in the fiber optic. This frequency change can be detected by the fiber optic reflection signal.

[0032] The formula is described as follows: Δf B =C T ΔT+C σ Δσ Where: Δf B is the Brillouin frequency shift (Hz), which is proportional to the temperature and stress changes; C T ,C σ are the sensitivity coefficients of temperature and stress respectively; ΔT is the temperature change (unit: ℃); Δσ is the stress change (unit: Pa).

[0033] This formula shows that the linear relationship between the Brillouin frequency shift and the temperature and stress changes makes it possible to accurately obtain the temperature and stress state of the outer wall of the pipeline by monitoring the frequency shift, which is crucial for early detection of corrosion risks.

[0034] In this embodiment, the role of the sensor node unit is to transmit the sensing signals of each optical fiber sensor module to the signal acquisition module through wavelength division multiplexing technology. Each sensor node corresponds to an optical fiber sensor, and the signal is transmitted through a specific wavelength to ensure that the multi-sensor signals will not interfere when transmitted on the same optical fiber.

[0035] Wavelength division multiplexing technology is a technology that uses light waves of different wavelengths to transmit multiple information simultaneously. Each fiber optic sensor unit is distinguished according to different working wavelengths and sends the signal to the signal processing center through the optical fiber network. The sensor node unit transmits the data of multiple fiber optic sensors in parallel through wavelength division multiplexing technology, which not only improves the transmission efficiency but also reduces the complexity and cost of the system. Each sensor node can work independently, and the system can flexibly schedule and transmit data according to the needs of different sensor nodes.

[0036] Through the coordinated work of sensor node units, the system can simultaneously acquire data at multiple points and transmit the data to the back-end processing module efficiently and quickly. In this way, the entire system can achieve efficient real-time monitoring of long-distance and complex pipeline systems.

[0037] The main task of the signal processing unit is to denoise, filter and standardize the signals transmitted from the optical fiber sensor to provide clean and standardized input data for the subsequent analysis modules.

[0038] Denoising: Since there are often various noise sources in the pipeline monitoring environment (such as external electromagnetic interference, ambient temperature changes, etc.), it is necessary to denoise the collected signals. Common denoising methods include low-pass filtering, Gaussian filtering, etc. The low-pass filter can remove high-frequency noise and retain the effective low-frequency signal to ensure that the transmitted signal data represents the real physical changes.

[0039] Filtering: After the signal is denoised, it still needs to be further filtered to remove interference components in a specific frequency band. During the signal processing, the filter circuit removes unnecessary high-frequency parts through the set cutoff frequency to maintain the integrity of the data.

[0040] Standardization: The signal processing unit also standardizes the signal to make the data from different sensors consistent. This process is achieved through a normalization algorithm, which adjusts the data collected by different sensors to a unified standard range to facilitate subsequent fusion analysis. The standardized data not only eliminates the differences between different fiber optic sensors, but also provides a standard input for signal analysis and data fusion.

[0041] Through the above processing, the signal processing unit effectively improves the quality of the signal, eliminates external interference, and provides accurate input for subsequent data storage, analysis, and evaluation.

[0042] Please see attached Figure 3 , the signal acquisition and edge computing modules include: A signal amplification unit, used for amplifying the signal transmitted by the optical fiber sensor; A filter circuit unit is used to filter out high-frequency noise in the signal and retain the effective signal; an anomaly detection unit, for performing abnormal data detection; The data compression unit is used to reduce the amount of data transmitted to the cloud and effectively retain key data through compression processing.

[0043] Specifically, the signal acquisition and edge computing module plays an important role in the buried PE pipeline corrosion monitoring and evaluation system of the present invention. It is mainly responsible for receiving the signal transmitted from the optical fiber sensor module, and amplifying, filtering, abnormal detection and compression processing of the signal to ensure the quality and efficiency of data transmission. The design of this module ensures that the data collected from the sensor is optimized during the processing process, so that the data can be efficiently transmitted to the cloud for storage and analysis, thereby realizing real-time monitoring and corrosion risk assessment of the pipeline.

[0044] The module consists of four main units: signal amplification unit, filter circuit unit, anomaly detection unit and data compression unit. Each unit plays a key role in the signal processing flow.

[0045] In this embodiment, the signal amplification unit is mainly used to amplify the weak signal transmitted from the optical fiber sensor to ensure that the signal is suitable for subsequent processing and analysis. Since the signal captured by the optical fiber sensor is usually weak, the signal strength may be attenuated after a certain distance of transmission, so it is necessary to increase the signal strength through the amplification unit to avoid signal loss or interference.

[0046] The signal amplification unit uses an operational amplifier circuit structure for signal gain. The gain formula is as follows: Among them: A v is the magnification; R f is the feedback resistor; R in is the input resistance.

[0047] By reasonably selecting the amplification factor, the signal amplification unit can amplify the signal output by the optical fiber sensor to a suitable range, providing strong support for further signal processing.

[0048] In this embodiment, the main function of the filter circuit unit is to filter out high-frequency noise in the signal and retain the valid signal. Since there may be external factors such as electromagnetic interference and environmental noise in the pipeline monitoring environment, the signal is easily contaminated by noise during transmission, which affects the signal quality. The filter circuit unit effectively removes high-frequency noise and retains low-frequency signals related to pipeline temperature and stress changes by using a low-pass filter or a band-pass filter.

[0049] The transfer function of the low-pass filter is given by the following formula: Where: c is the cutoff frequency of the filter; Q is the quality factor of the filter, which determines the bandwidth of the filter.

[0050] By properly selecting the cutoff frequency, the filter circuit can filter out irrelevant frequency components in the signal, retain the signals of key physical quantities such as temperature and stress changes, and ensure that the signal is not disturbed during subsequent processing.

[0051] In this embodiment, the anomaly detection unit is used to perform anomaly detection on the transmitted signal data and identify potential abnormal data points. During the signal acquisition process, it may be affected by various factors, such as sensor failure, electromagnetic interference, or drastic changes in the external environment, causing data anomalies. In order to ensure the accuracy of the data, the anomaly detection unit uses a local anomaly factor (LOF) algorithm to analyze the signal data and detect abnormal data that is significantly different from normal data points.

[0052] The basic principle of the LOF algorithm is to evaluate the abnormality of a data point by calculating the local reachability density of the data point and other data points in its neighborhood. The specific formula is as follows: in: LOF(p) is the local outlier factor of data point p; N k (p) is the k nearest neighbors of data point p; lrd(p) is the local reachable density of point p; lrd(o) is the local reachable density of point o.

[0053] When the LOF(p) value exceeds the set threshold, it means that the data point p is abnormal data, and the system will mark and further process the data point.

[0054] In this embodiment, the data compression unit is used to compress the data after amplification, filtering and anomaly detection to reduce the amount of data transmitted to the cloud. In a large-scale pipeline monitoring system, the amount of data from the optical fiber sensor may be very large. Directly transmitting all the data will consume a lot of bandwidth and storage resources. Therefore, it is necessary to use a compression algorithm to effectively reduce the amount of data transmission.

[0055] Data compression methods use compression techniques such as wavelet transform or Huffman coding. Wavelet transform can decompose the signal into different frequency components, thereby compressing the data according to the importance of different frequency bands, retaining key information while removing redundant parts. Huffman coding achieves data compression by assigning different lengths of codes to data of different frequencies.

[0056] The goal of the data compression unit is to reduce the amount of data transmitted to the cloud while ensuring that important information is not lost, thus ensuring the accuracy of subsequent analysis.

[0057] Please see attached Figure 4 ,The cloud data storage and analysis modules include: A data storage unit, used for storing the data transmitted by the signal acquisition module into a distributed database; A data analysis unit, which is used to perform corrosion risk analysis on the stored data based on machine learning algorithms to assess the health status of the pipeline; The historical data processing unit is used to train and optimize the machine learning model using historical data.

[0058] Specifically, the cloud data storage and analysis module is a key component of the buried PE pipeline corrosion monitoring and assessment system. It is mainly responsible for receiving and storing data transmitted from the signal acquisition module, and using advanced machine learning algorithms to perform corrosion risk analysis on these data, and ultimately assess the health of the pipeline. Through this module, the system can extract valuable information from large-scale data, thereby providing pipeline maintenance personnel with real-time, accurate corrosion assessment reports and early warnings. The module consists of three main units: a data storage unit, a data analysis unit, and a historical data processing unit. Each unit plays an important role in the storage, processing, and analysis of data.

[0059] In this embodiment, the data storage unit is used to store the data transmitted by the signal acquisition module in a distributed database. Since the amount of data generated during pipeline monitoring is huge and involves data streams from multiple sensors, a distributed database is used for storage, which can efficiently manage and process large-scale data. The distributed database has high scalability, fault tolerance, and concurrent processing capabilities, which can ensure efficient storage and fast access to data.

[0060] Common distributed databases, such as HBase and Cassandra, can be horizontally expanded as needed to process large amounts of real-time data from various sensors. In this module, the data storage unit is responsible for receiving and storing the raw data, pre-processed data, and analyzed data transmitted by the signal acquisition module. In order to ensure the security and integrity of the data, the data storage unit uses backup and redundancy mechanisms to prevent data loss and ensure high availability of the system.

[0061] The distributed database distributes data to different storage nodes through data sharding technology, ensuring that a large amount of data streams from various sensors can be processed and stored efficiently and stably. At the same time, the system also uses data compression technology to reduce storage space usage and further improve data storage efficiency.

[0062] In this embodiment, the data analysis unit is the core of the cloud data storage and analysis module, which is responsible for using machine learning algorithms to perform corrosion risk analysis on the stored sensor data. The unit conducts in-depth analysis of the monitoring data such as temperature, stress, and pressure of the pipeline, evaluates the health status of the pipeline, and generates a corrosion risk assessment report.

[0063] The data analysis unit usually includes the following aspects of work: Data preprocessing: First, the data analysis unit cleans and standardizes the stored raw data to eliminate noise and outliers to ensure the accuracy and reliability of the data.

[0064] Feature extraction: Based on the physical model and monitoring indicators of pipeline corrosion, key features that affect corrosion are extracted, such as temperature fluctuation range, stress change trend, etc. Through feature engineering, the data analysis unit can extract useful features from a large amount of data and improve the predictive ability of the machine learning model.

[0065] Machine learning model: The data analysis unit uses supervised learning or unsupervised learning algorithms to analyze data. Common algorithms include random forests, support vector machines, neural networks, etc. These algorithms can learn the corrosion pattern of pipelines from a large amount of historical data and then predict the corrosion risk of the current pipeline.

[0066] Specifically, the training and prediction process of the machine learning model includes: in: is the predicted corrosion risk; X is the input data (temperature, stress and other characteristic data); f is the prediction function of the machine learning model.

[0067] The data analysis unit predicts real-time data through the trained model, evaluates the corrosion risk level of the pipeline, and generates detailed reports for maintenance personnel as decision support.

[0068] In this embodiment, the historical data processing unit is used to train and optimize the machine learning model using historical data. Since the assessment of corrosion risk depends on a large amount of historical data, this unit plays a very important role in the system. By continuously updating and optimizing the machine learning model, the historical data processing unit can improve the accuracy and reliability of the model, so that the system can adapt to the corrosion monitoring needs under different pipeline environments.

[0069] Specifically, the historical data processing unit includes the following functions: Data integration and cleaning: This unit is responsible for collecting and organizing historical data, integrating data from different sources, and dealing with issues such as missing values ​​and outliers to ensure data quality.

[0070] Model training: By using the cleaned historical data as a training set, a model for predicting corrosion risk is trained using machine learning algorithms (such as regression analysis, classification algorithms, etc.). These training data include information such as pipeline temperature, stress changes, and historical corrosion conditions.

[0071] Model optimization: The historical data processing unit continuously optimizes and adjusts the machine learning model based on new monitoring data to ensure that the model can maintain high prediction accuracy even in new environments and conditions.

[0072] Through continuous model optimization, the historical data processing unit continuously improves the system's ability to predict corrosion risks, ensuring that the system can always provide accurate and real-time corrosion assessments.

[0073] Please see attached Figure 5 , the multimodal data fusion module includes: a multi-sensor data input unit for transmitting data from multiple sensors including a distributed optical fiber sensor, a soil moisture sensor, and a pressure sensor to a fusion module; A data fusion algorithm unit is used to perform weighted fusion of sensor data from different sources; The deep learning analysis unit is used to analyze the fused data through a deep learning algorithm to perform corrosion assessment.

[0074] The data fusion algorithm unit uses the following formula to perform weighted fusion of sensor data from different sources: D fusion =αD fiber +βD soil +γD flow Among them, D fiber , D soil , D flow are the data from different sensors, and α, β, and γ are weighting coefficients.

[0075] Specifically, in this embodiment, the main function of the multi-sensor data input unit is to collect data from multiple sensors (including distributed optical fiber sensors, soil moisture sensors and pressure sensors) and transmit these data to the data fusion algorithm unit. This unit ensures that the data from different sensors can be transmitted to the fusion module in a timely and accurate manner for further processing.

[0076] Specifically, the sensors include: Distributed fiber optic sensors to measure temperature and stress changes on the outer wall of the pipeline; Soil moisture sensor, used to monitor humidity changes in the pipeline burial environment and reflect the impact of soil on pipeline corrosion; Pressure sensors are used to measure the pressure changes of gas or liquid flowing in pipelines, indirectly reflecting the stress conditions in the pipelines.

[0077] The data from each sensor reflects different aspects of pipeline health, so the data input unit ensures that the information from multiple sensors is integrated and transmitted to downstream modules in a timely manner.

[0078] In this embodiment, the core task of the data fusion algorithm unit is to perform weighted fusion on the data from multiple sensors. Since the data from different sensors have different physical properties and time resolutions, they need to be integrated into a unified evaluation input through a weighted algorithm. Through weighted fusion, the system can give different weights to different sensors according to their reliability and accuracy, ensuring that the fused data reflects the health status of the pipeline as accurately as possible.

[0079] The process of data fusion can be expressed by the following formula: D fusion =αD fiber +βD soil +γD flow Where: D fiber Data from distributed fiber optic sensors, typically including temperature and stress changes; soil Data from soil moisture sensor, indicating soil moisture changes; D flow The data from the pressure sensor indicates the pressure change of the gas or liquid flowing in the pipeline; α, β, and γ are weighting coefficients, which represent the importance of each sensor in the data fusion process.

[0080] The weighting coefficient can be adjusted according to the accuracy and reliability of each sensor in different working environments. For example, if the data of a sensor is relatively stable, it can be given a larger weight.

[0081] Through weighted fusion, the data fusion algorithm unit can generate fused data that comprehensively reflects the health status of the pipeline, providing effective input for subsequent corrosion risk assessment.

[0082] In this embodiment, the main task of the deep learning analysis unit is to analyze the fused data and perform corrosion risk assessment. The deep learning algorithm can effectively handle complex nonlinear relationships and adapt to the analysis needs of multidimensional data. Especially after the fusion of multiple sensor data, it can deeply mine the potential features in the data through models such as deep neural networks or convolutional neural networks, further improving the accuracy of corrosion assessment.

[0083] In this unit, the fused data is used as the input of the neural network model. The network gradually extracts the corrosion pattern of the pipeline through calculations of multiple hidden layers. The goal of this process is to use historical data and current monitoring data to train the model to predict the corrosion risk of the pipeline and give the corresponding evaluation results. The specific steps are as follows: Data input: The fused data is used as input to the neural network model.

[0084] Feature extraction: Extract high-dimensional features from data through a multi-layer network structure.

[0085] Corrosion assessment: The corrosion risk level of the pipeline is predicted through the trained model, and the corrosion assessment results are output.

[0086] The deep learning model can automatically adjust its weights and parameters to continuously optimize the evaluation results during the training process and improve the ability to recognize and predict corrosion patterns.

[0087] Please see attached Figure 6 , the corrosion assessment and early warning system modules include: A corrosion assessment unit, which is used to assess the corrosion risk level of the pipeline based on the analysis results and generate a risk assessment report; An early warning signal generation unit is used to trigger different levels of early warning signals according to the evaluation results, including low risk, medium risk and high risk. The signals are sent to pipeline maintenance personnel via SMS, email or mobile application; The threshold adjustment unit is used to dynamically adjust the warning threshold according to historical data and external environmental factors.

[0088] Specifically, in this embodiment, the corrosion assessment unit evaluates the corrosion risk level of the pipeline based on the fusion data output by the multimodal data fusion module and the machine learning analysis results. The unit compares historical data with real-time monitoring data, combines multi-dimensional information such as pipeline temperature, stress, soil moisture, and pressure, and evaluates the corrosion status of the pipeline through an algorithm.

[0089] The corrosion assessment process can be divided into the following steps: Data input: The corrosion assessment unit receives data from the multimodal data fusion module, which includes multi-source data from optical fiber sensors, soil moisture sensors, pressure sensors, etc.

[0090] Risk level classification: Through comprehensive analysis of these data, the assessment unit divides the corrosion risk of the pipeline into different levels (such as low risk, medium risk, and high risk) according to the set corrosion risk standards.

[0091] Risk assessment report generation: The assessment unit generates a detailed corrosion risk assessment report based on the analysis results. The report includes: the current health status of the pipeline, the severity of corrosion, possible danger points and recommended maintenance measures.

[0092] During this process, the evaluation unit can use the following model to quantitatively evaluate the corrosion risk: Risk=f(T,σ,P,H,S) Among them: Risk is the corrosion risk assessment result; T is the pipeline temperature data; σ is the pipeline stress data; P is the pressure data inside the pipeline; H is the soil moisture data; S is the historical corrosion data.

[0093] The function f in this formula is generated based on machine learning model training and can output a comprehensive corrosion risk level based on multi-dimensional data input.

[0094] In this embodiment, the warning signal generation unit generates warning signals of different levels according to the output results of the corrosion assessment unit. According to the assessment results, the system divides the pipeline corrosion risk into three levels: low risk, medium risk and high risk, and triggers corresponding warning signals according to different levels of risk.

[0095] The generation process of the early warning signal is as follows: Low-risk warning: When the corrosion risk assessment result of the pipeline is low risk, the system will not generate an emergency warning, but will only generate a reminder message to indicate that the pipeline is in normal condition and continue regular monitoring.

[0096] Medium risk warning: When the corrosion risk assessment result is medium risk, the system will generate a yellow warning signal and recommend that pipeline maintenance personnel conduct regular inspections. At this time, the system will provide the health status of the pipeline, possible corrosion locations, and recommended inspection cycles.

[0097] High-risk warning: When the assessment result is high risk, the system will trigger a red warning signal to remind that there is a serious corrosion problem in the pipeline and immediate maintenance measures are required. The high-risk warning will be sent to pipeline maintenance personnel via SMS, email, mobile applications, etc., with detailed corrosion location, severity and emergency treatment suggestions.

[0098] The key to early warning signal generation is to accurately determine the risk level and promptly notify maintenance personnel. Therefore, the early warning signal generation unit generates an appropriate notification signal based on a combination of the set risk threshold and the assessment results.

[0099] In this embodiment, the function of the threshold adjustment unit is to dynamically adjust the warning threshold according to historical data and external environmental factors. Since environmental factors (such as temperature, humidity, soil pH, etc.) can affect the occurrence and progress of the corrosion process, a fixed threshold may not accurately reflect the corrosion risk of pipelines under different environments. Therefore, the threshold adjustment unit dynamically adjusts the warning threshold so that the system can better adapt to corrosion risk assessment under different conditions.

[0100] The threshold adjustment process consists of the following steps: Historical data analysis: Based on historical monitoring data, the threshold adjustment unit analyzes the corrosion development rules under different conditions and dynamically adjusts the warning threshold based on these rules. For example, in an environment with high soil humidity, the threshold may be lowered accordingly, because excessive humidity will accelerate corrosion.

[0101] External environmental impact: Real-time data on external environmental factors (such as weather changes, soil moisture, etc.) is obtained, and the warning threshold is adjusted according to these environmental changes. For example, seasonal climate change may affect the rate of pipeline corrosion, so the threshold needs to be adaptively adjusted.

[0102] Feedback mechanism: As the system runs for a longer time, the threshold adjustment unit can also continuously optimize the warning threshold from the system operation feedback through the machine learning model, further improving the assessment accuracy and warning response capability.

[0103] Specifically, the threshold adjustment unit adjusts the corrosion risk assessment threshold according to the following formula: Threshold adjusted =Threshold base +ΔThreshold env Where: Threshold adjusted is the warning threshold after dynamic adjustment; Threshold base is the benchmark warning threshold; ΔThreshold env It is the threshold change caused by external environmental factors (such as soil moisture, temperature change, etc.).

[0104] Through this dynamic adjustment mechanism, the threshold adjustment unit can ensure the accuracy and flexibility of the system according to different environmental conditions.

[0105] Please see attached Figure 7 A method for monitoring and evaluating the corrosion of a buried PE pipeline, used in conjunction with the above-mentioned system for monitoring and evaluating the corrosion of a buried PE pipeline, comprises the following steps: S1. Install the distributed optical fiber sensor on the outer wall of the buried PE pipeline to collect the temperature and stress change data of the pipeline in real time; S2, amplify and filter the sensor signal through the signal acquisition and edge computing module, apply the local anomaly factor algorithm to perform preliminary anomaly detection on the data, and compress the abnormal data and upload it to the cloud; S3, storing the uploaded data in the cloud data storage and analysis module, and using machine learning models to analyze the corrosion risk of the pipeline; S4. The data from different sensors are fused through the multimodal data fusion module, and the corrosion risk is evaluated using the deep learning algorithm to improve the evaluation accuracy; S5. Generate a risk assessment report on pipeline corrosion through the corrosion assessment and early warning system module, and trigger different levels of early warning signals according to preset thresholds to notify pipeline maintenance personnel.

[0106] In step S2, the signal acquisition and edge computing module uses the following LOF anomaly detection algorithm when performing signal processing: Among them, p is the current data point, N k (p) is the k nearest neighbors of p, and lrd(p) is the local density of point p.

[0107] The risk assessment report in step S5 includes the current corrosion status of the pipeline, risk prediction results and pipeline maintenance recommendations. The thresholds in step S5 are set based on the monitoring data of temperature, stress and pressure changes.

[0108] Specifically, in this embodiment, in step S1, a distributed optical fiber sensor is installed on the outer wall of the buried PE pipeline to collect the temperature and stress change data of the pipeline in real time. Based on the principle of Brillouin scattering, the distributed optical fiber sensor can accurately capture the tiny temperature and stress changes on the surface of the pipeline and provide comprehensive pipeline health data. These data provide a key basis for subsequent corrosion assessment and early warning. Since the distributed optical fiber sensor can be laid along the pipeline, it can continuously monitor the status of the pipeline over a wide range to ensure that there are no blind spots.

[0109] In step S2, the sensor signal is preliminarily processed by the signal acquisition and edge computing module. First, the sensor signal is amplified and filtered to eliminate noise and enhance the signal strength to make it suitable for subsequent analysis. Then, the local outlier factor (LOF) algorithm is used to detect anomalies in the signal data in order to quickly identify potential anomalies. The basic principle of the LOF algorithm is to identify anomalies by comparing the local density of a data point with the density difference of its neighboring data points. The specific algorithm formula is as follows: Where: p is the current data point, N k (p) is the k nearest neighbors of p, and lrd(p) is the local density of point p.

[0110] When the LOF(p) value exceeds the set threshold, it means that the data point p may be abnormal data, and the system will mark and further analyze these abnormal data. Finally, the processed data will be compressed and uploaded to the cloud data storage module to reduce the amount of data transmission and ensure the security of information.

[0111] In step S3, the cloud data storage and analysis module stores the uploaded signal data and uses machine learning models to analyze the data and assess the corrosion risk of the pipeline. The machine learning model uses historical data and real-time data for training and prediction, can identify the corrosion pattern of the pipeline, and output the risk assessment results of the pipeline health status. By analyzing multiple factors such as temperature and stress changes, the model can comprehensively determine whether the pipeline has potential corrosion risks and generate a corresponding corrosion risk report.

[0112] In step S4, the data from different sensors are fused through the multimodal data fusion module, and the corrosion risk is further analyzed using a deep learning algorithm to improve the assessment accuracy. Multimodal data fusion technology can eliminate the errors that may be caused by single sensor data through weighted fusion of multiple sensor data sources, and provide a more comprehensive and accurate corrosion risk assessment. By learning the fused data through a deep learning algorithm, the system can adapt to complex working environments and achieve accurate corrosion risk prediction.

[0113] The weighted formula for data fusion is as follows: D fusion =αD fiber +βD soil +γD flow Among them, D fiber , D soil , D flow are the data from different sensors, and α, β, and γ are weighting coefficients.

[0114] Through this fusion method, the system can provide more accurate corrosion risk prediction based on considering different environmental factors.

[0115] In step S5, the corrosion assessment and early warning system module generates a risk assessment report for pipeline corrosion based on the analysis results, and triggers early warning signals of different levels according to preset thresholds to notify pipeline maintenance personnel. The risk assessment report includes the current corrosion status of the pipeline, risk prediction results, and pipeline maintenance recommendations, providing important decision-making basis for maintenance personnel. The generation of early warning signals is based on the thresholds set by monitoring data such as pipeline temperature, stress, and pressure. The system triggers low-risk, medium-risk, and high-risk early warning signals based on the assessment results, and notifies relevant maintenance personnel through SMS, email, or mobile applications. In this way, maintenance personnel can take corresponding measures in time at the early stage of discovering corrosion risks to avoid the occurrence of safety accidents.

[0116] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A buried PE pipeline corrosion monitoring and evaluation system, characterized in that: include: Distributed fiber optic sensor modules are installed on the outer wall of buried PE pipelines to collect data on pipeline temperature and stress changes, and sense pipeline corrosion through fiber optic sensing signals; The signal acquisition and edge computing module is used to receive and process the signal data transmitted by the optical fiber sensor, perform signal amplification, filtering, anomaly detection, and data compression before transmitting it to the cloud; Cloud data storage and analysis module, which is used to store sensor data and analyze the data based on machine learning algorithms to assess pipeline corrosion risks and generate risk assessment reports; Multimodal data fusion module, which is used to fuse data from multiple sensors including distributed fiber optic sensors, soil moisture sensors and pressure sensors, and conduct comprehensive analysis using deep learning algorithms for corrosion assessment; The corrosion assessment and early warning system module is used to generate a risk assessment report on pipeline corrosion based on the analysis results, and issue an early warning signal to notify pipeline maintenance personnel.

2. A buried PE pipeline corrosion monitoring and evaluation system according to claim 1, characterized in that: The distributed optical fiber sensor module comprises: A fiber optic sensor unit, which is used to sense the temperature and stress changes on the outer wall of the pipeline based on the Brillouin scattering principle; The sensing node unit is used to transmit the sensing signal of the optical fiber sensor to the signal acquisition module through wavelength division multiplexing technology; The signal processing unit is used to remove noise from the transmitted signal and convert it into standardized data.

3. A buried PE pipeline corrosion monitoring and evaluation system according to claim 1, characterized in that: The signal acquisition and edge computing module includes: A signal amplification unit, used for amplifying the signal transmitted by the optical fiber sensor; A filter circuit unit is used to filter out high-frequency noise in the signal and retain the effective signal; an anomaly detection unit, for performing abnormal data detection; The data compression unit is used to reduce the amount of data transmitted to the cloud and effectively retain key data through compression processing.

4. A buried PE pipeline corrosion monitoring and evaluation system according to claim 1, characterized in that: The cloud data storage and analysis module includes: A data storage unit, used for storing the data transmitted by the signal acquisition module into a distributed database; A data analysis unit, which is used to perform corrosion risk analysis on the stored data based on machine learning algorithms to assess the health status of the pipeline; The historical data processing unit is used to train and optimize the machine learning model using historical data.

5. A buried PE pipeline corrosion monitoring and evaluation system according to claim 1, characterized in that: The multimodal data fusion module includes: a multi-sensor data input unit for transmitting data from multiple sensors including a distributed optical fiber sensor, a soil moisture sensor, and a pressure sensor to a fusion module; A data fusion algorithm unit is used to perform weighted fusion of sensor data from different sources; The deep learning analysis unit is used to analyze the fused data through a deep learning algorithm to perform corrosion assessment.

6. A buried PE pipeline corrosion monitoring and evaluation system according to claim 5, characterized in that: The data fusion algorithm unit uses the following formula to perform weighted fusion of sensor data from different sources: D fusion =αD fiber +βD soil +γD flow Among them, D fiber , D soil , D flow are the data from different sensors, and α, β, and γ are weighting coefficients.

7. A buried PE pipeline corrosion monitoring and evaluation system according to claim 1, characterized in that: The corrosion assessment and early warning system module includes: A corrosion assessment unit, which is used to assess the corrosion risk level of the pipeline based on the analysis results and generate a risk assessment report; An early warning signal generation unit is used to trigger different levels of early warning signals according to the evaluation results, including low risk, medium risk and high risk. The signals are sent to pipeline maintenance personnel via SMS, email or mobile application; The threshold adjustment unit is used to dynamically adjust the warning threshold according to historical data and external environmental factors.

8. A method for monitoring and evaluating the corrosion of buried PE pipelines, characterized in that: Used in conjunction with a buried PE pipeline corrosion monitoring and evaluation system as described in any one of claims 1 to 7, comprising the following steps: S1. Install the distributed optical fiber sensor on the outer wall of the buried PE pipeline to collect the temperature and stress change data of the pipeline in real time; S2, amplify and filter the sensor signal through the signal acquisition and edge computing module, apply the local anomaly factor algorithm to perform preliminary anomaly detection on the data, and compress the abnormal data and upload it to the cloud; S3, storing the uploaded data in the cloud data storage and analysis module, and using machine learning models to analyze the corrosion risk of the pipeline; S4. The data from different sensors are fused through the multimodal data fusion module, and the corrosion risk is evaluated using the deep learning algorithm to improve the evaluation accuracy; S5. Generate a risk assessment report of pipeline corrosion through the corrosion assessment and early warning system module, and trigger different levels of early warning signals according to preset thresholds to notify pipeline maintenance personnel.

9. A method for monitoring and evaluating the corrosion of buried PE pipelines according to claim 8, characterized in that: In the step S2, the signal acquisition and edge computing module uses the following LOF anomaly detection algorithm when performing signal processing: Among them, p is the current data point, N k (p) is the k nearest neighbors of p, and lrd(p) is the local density of point p.

10. A method for monitoring and evaluating the corrosion of buried PE pipelines according to claim 8, characterized in that: The risk assessment report in step S5 includes the current corrosion status of the pipeline, risk prediction results and pipeline maintenance suggestions. The threshold in step S5 is set based on the monitoring data of temperature, stress and pressure changes.

Images