A pipeline blockage monitoring system and method based on multi-source data fusion

Abstract

The application provides a pipeline blockage monitoring system and method based on multi-source data fusion, which comprises: a hardware perception layer, a plurality of monitoring nodes are arranged along the pipeline, each node is integrated with an infrasound sensor and a pressure transmitter; a data processing and analysis layer configured to perform frequency spectrum analysis on the vibration signal, extract the energy time sequence characteristics of the predetermined characteristic frequency band, and determine the blockage trend according to the change rate; when it is determined that there is a blockage trend, the spatial distribution gradient is analyzed by fusing the pressure data, the pressure abnormal area is associated with the vibration signal intensity space to locate the blockage point, and the risk is evaluated based on the energy time sequence characteristics, the blockage point position and the operating parameters; an application interaction layer for outputting the risk evaluation result and the blockage point position. Through the multi-source fusion of infrasound and pressure data, the application realizes early identification of blockage trend, accurate positioning of spatial position and quantitative evaluation of risk, effectively improving the timeliness, accuracy and intelligent level of pipeline blockage monitoring.

Description

Technical Field

[0001] This invention relates to the field of data processing and system monitoring technology, and in particular to a pipeline blockage monitoring system and method based on multi-source data fusion. Background Technology

[0002] Due to the unique characteristics of slurry media (high concentration, high hardness, and non-Newtonian fluid properties), pipeline blockage is one of the main operational risks during slurry pipeline transportation. Blockage can lead to production interruptions, equipment damage, and even safety accidents. Therefore, early and accurate identification and location of blockage risks are of great significance. Currently, monitoring of slurry pipeline blockage mainly relies on the following methods: 1. Manual inspection and single-point threshold alarm This is the most traditional method. It involves regular inspections by maintenance personnel or monitoring using single-point pressure transmitters and flow meters installed on the pipeline. An alarm is triggered when the pressure or flow exceeds a preset threshold. However, manual inspections are inefficient, have slow response times, and are ill-suited for handling sudden blockages. Single-point threshold alarms cannot distinguish between blockages and parameter fluctuations caused by other operating conditions (such as valve movement or pump start / stop), resulting in a high false alarm rate. Furthermore, they cannot provide the location of the blockage, making it difficult to guide precise maintenance.

[0003] 2. Trend analysis based on flow or pressure Some improved methods assess the likelihood of blockage by monitoring the rate of decrease in flow or pressure. However, this method is extremely insensitive to early, localized blockages, and it is only detected when the blockage has developed to a certain extent and significantly affected fluid dynamic parameters, often missing the optimal time for intervention.

[0004] 3. Leakage and blockage detection based on infrasound. In recent years, infrasound sensing technology, due to its long propagation distance and low attenuation, has been explored for detecting pipeline leaks and blockages. For example, Chinese patent application CN118961900A proposes using an infrasound sensor array for passive or active sensing of pipelines to identify abnormal events. However, most existing infrasound applications remain at the level of signal acquisition or simple energy threshold judgment, failing to delve into the spectral characteristics of infrasound signals under different operating conditions, nor clarifying how to use infrasound data to identify early trends of blockages. More importantly, existing technologies have failed to effectively integrate infrasound data with process data such as pressure and flow rate to form a complete automated system from early warning and precise location to risk quantification assessment. For example, relying solely on infrasound makes it difficult to distinguish vibrations caused by blockages from other mechanical vibrations; relying solely on pressure fails to identify early trends before pressure changes occur.

[0005] 4. Other attempts at multi-source data fusion In other pipeline monitoring fields (such as crude oil and natural gas), some research has been conducted on multi-source data fusion. For example, Chinese patent application CN120256979A discloses a multimodal sensor data fusion method for crude oil pipelines to detect pipeline damage and corrosion; CN120312994A proposes a gas leak prediction method based on an LSTM-CNN hybrid model, fusing multi-dimensional data such as concentration, sedimentation, and complaints. However, these solutions address leaks or corrosion problems in single-phase flows (crude oil and gas), and their physical mechanisms, feature extraction methods, and fusion logic are fundamentally different from those for solid-liquid two-phase flow (slurry) blockage problems. Crude oil and natural gas monitoring focuses on imaging or diffusion prediction of leak points, while slurry blockage monitoring needs to focus on early spectral characteristic changes caused by solid particle friction, and the resulting changes in flow resistance (pressure gradient anomalies). Existing technologies have not revealed this physical mechanism, nor have they provided a dedicated fusion analysis scheme for slurry blockage. Summary of the Invention

[0006] The technical problem to be solved by this invention is: In view of the technical problems existing in the prior art, this invention provides a pipeline blockage monitoring system and method based on multi-source data fusion, which aims to overcome the shortcomings of existing slurry pipeline blockage monitoring technology, such as delayed early warning, high false alarm rate, inability to accurately locate, and lack of risk assessment.

[0007] To solve the above-mentioned technical problems, the technical solution proposed by this invention is as follows: A pipeline blockage monitoring system based on multi-source data fusion includes: The hardware sensing layer includes multiple monitoring nodes deployed along the pipeline, each of which includes at least an infrasound sensor for collecting pipeline vibration signals and a pressure transmitter for collecting pressure inside the pipeline. The data processing and analysis layer includes at least one processor, the processor being configured as follows: Trend identification: The vibration signal collected by the infrasound sensor is subjected to spectrum analysis to extract the energy time-series characteristics of the predetermined characteristic frequency band, and the presence of a blockage trend is determined based on the rate of change of the energy time-series characteristics. Location verification: When the blockage trend is determined, the pressure data collected by the pressure transmitter is retrieved, the spatial distribution gradient of pressure along the pipeline is analyzed, and the abnormal pressure area is spatially correlated with the vibration signal intensity to determine the location information of the blockage point; Risk assessment: Based on the energy timing characteristics, the location information of the blockage point, and the operating parameters of the pipeline, a risk assessment result is generated; An application interaction layer is used to output the risk assessment results and / or the location information of the blockage points.

[0008] As a further improvement of the present invention, the hardware sensing layer also includes a timing module, which is used to add a unified timestamp to the data collected by the infrasound sensor and pressure transmitter of the same monitoring node, so as to realize the synchronization of multi-source data in time.

[0009] As a further improvement of the present invention: in the trend identification, the spectrum analysis includes: performing wavelet packet transform decomposition on the vibration signal collected by the infrasound sensor, extracting the energy proportion of a predetermined characteristic frequency band as a feature value, and establishing a time series curve of the feature value, calculating the rate of change of the time series curve through a sliding window; when the rate of change continuously exceeds a preset threshold, it is determined that there is a blockage trend and a primary warning is triggered.

[0010] As a further improvement of the present invention: in the positioning verification, the step of spatially correlating the abnormal pressure area with the vibration signal intensity to determine the location of the blockage point includes: Calculate the pressure difference between adjacent monitoring nodes. When the pressure gradient of a certain pipe section increases abnormally and the pressure difference exceeds the preset threshold, the section is identified as a pressure abnormality area. Compare the infrasound signal energy values ​​of each monitoring node in the predetermined characteristic frequency band, and take the node with the largest energy value as the monitoring node location with the strongest vibration signal; The coordinates of the blockage point are calculated using a weighted algorithm based on the abnormal pressure area and the location of the monitoring node with the strongest vibration signal intensity.

[0011] As a further improvement to the present invention: Based on the abnormal pressure area and the location of the monitoring node with the strongest vibration signal intensity, the coordinates of the blockage point are calculated using a weighted algorithm; the weighted algorithm is as follows:

[0012] in, The coordinates of the blockage point. The coordinates of the center point of the pressure anomaly region are: The coordinates of the monitoring node with the strongest vibration signal intensity; The weights are dynamically set based on the degree of pressure gradient anomaly. The weights are dynamically set based on the vibration signal intensity, satisfying:

[0013]

[0014] in, The pressure difference in the aforementioned pressure anomaly region. The pressure difference threshold The infrasound energy value of the strongest node. This represents the average infrasound energy across all monitoring nodes.

[0015] As a further improvement of the present invention: in the risk assessment, the operating parameters include at least one of slurry flow rate and concentration; the risk assessment results include at least one of risk level, expected development speed, and disposal recommendations.

[0016] As a further improvement of the present invention, the risk assessment further includes: inputting the energy timing characteristics, the location information of the blockage point and the operating parameters into a pre-trained machine learning model, and having the machine learning model output the risk assessment result.

[0017] As a further improvement of the present invention: the application interaction layer includes a visualization interface for highlighting the location of the blockage point on the pipeline geographic information view and displaying the risk assessment results.

[0018] This invention also provides a pipeline blockage monitoring method based on multi-source data fusion, comprising the following steps: Step S1: Synchronously collect vibration signals and pressure data from multiple monitoring nodes along the pipeline; Step S2: Perform spectrum analysis on the vibration signal, extract the energy time-series characteristics of the predetermined characteristic frequency band, and determine whether there is a blockage trend based on the rate of change of the energy time-series characteristics; Step S3: If a blockage trend is determined, retrieve the pressure data, analyze the spatial distribution gradient of pressure along the pipeline, and spatially correlate the abnormal pressure area with the vibration signal intensity to determine the location of the blockage point; Step S4: Based on the energy timing characteristics, the location information of the blockage point, and the pipeline operating parameters, generate a risk assessment result; Step S5: Output the risk assessment results and / or the location information of the blockage point.

[0019] As a further improvement of the present invention: in step S3, the spatial correlation between the abnormal pressure region and the vibration signal intensity includes: The pipe segment interval with abnormally high pressure in the pressure gradient distribution is identified as the pressure anomaly region; Determine the location of the monitoring node with the strongest vibration signal intensity; The coordinates of the blockage point are calculated using a weighted algorithm based on the abnormal pressure area and the location of the monitoring node with the strongest vibration signal intensity.

[0020] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention collects pipeline vibration signals and pressure data through a hardware sensing layer to form a multi-source data foundation. It relies on trend identification, location verification, and risk assessment in the data processing and analysis layer to achieve the fusion analysis of multi-source data. It outputs relevant results through the application interaction layer, thus constructing an integrated pipeline blockage monitoring system. This effectively solves the technical defects of existing pipeline blockage monitoring technologies, such as delayed early warning, high false alarm rate, inability to accurately locate, and lack of systematic risk assessment.

[0021] 2. This invention performs spectral analysis on vibration signals collected by infrasound sensors and extracts the energy temporal characteristics of predetermined characteristic frequency bands. Combined with the rate of change of these characteristics, it determines the blockage trend, accurately capturing early signs of pipeline blockage and significantly improving the timeliness of blockage warnings. This represents a significant improvement over existing monitoring methods that are insensitive to early, localized blockages. After determining a blockage trend, it retrieves pressure data collected by pressure transmitters, analyzes the spatial distribution gradient of pressure along the pipeline, and spatially correlates abnormal pressure areas with vibration signal intensity to determine the location of the blockage point. This effectively distinguishes pressure changes caused by different operating conditions, reduces false alarm rates, and achieves precise blockage location, solving the problem of traditional monitoring methods being unable to locate blockages. Simultaneously, based on energy temporal characteristics, blockage location information, and pipeline operating parameters, it generates risk assessment results, providing a scientific and quantitative basis for pipeline blockage management. This changes the current situation where existing monitoring can only provide simple alarms without subsequent risk assessment, comprehensively improving the automation and intelligence level of pipeline blockage monitoring, effectively avoiding production interruptions and equipment damage caused by pipeline blockages, and ensuring the safe and efficient operation of pipeline transportation. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the pipeline blockage monitoring system architecture based on multi-source data fusion in an embodiment of the present invention.

[0023] Figure 2 This is a schematic diagram of the hardware composition of the monitoring node in an embodiment of the present invention.

[0024] Figure 3 This is a flowchart of the core algorithm module of the data processing and analysis layer in this embodiment of the invention.

[0025] Figure 4 This is a schematic diagram of the human-computer interface layout using the interaction layer in an embodiment of the present invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0027] like Figure 1 As shown, this embodiment provides a pipeline blockage monitoring system based on multi-source data fusion, including: The hardware sensing layer includes multiple monitoring nodes deployed along the pipeline. Each monitoring node includes at least an infrasound sensor for collecting pipeline vibration signals and a pressure transmitter for collecting pressure inside the pipeline. The data processing and analysis layer includes at least one processor, configured as follows: Trend identification: Perform spectrum analysis on the vibration signal collected by the infrasound sensor, extract the energy time-series characteristics of the predetermined characteristic frequency band, and determine whether there is a blockage trend based on the rate of change of the energy time-series characteristics; Location verification: When a blockage trend is determined, the pressure data collected by the pressure transmitter is retrieved, the spatial distribution gradient of the pressure along the pipeline is analyzed, and the abnormal pressure area is spatially correlated with the vibration signal intensity to determine the location information of the blockage point. Risk assessment: Based on energy timing characteristics, location information of blockage points, and pipeline operating parameters, risk assessment results are generated; The application interaction layer is used to output risk assessment results and / or the location information of blockage points.

[0028] In a specific application embodiment, the hardware sensing layer consists of multiple monitoring nodes spaced apart along the slurry conveying pipeline. Each monitoring node integrates at least the following devices: an infrasound sensor for collecting infrasound signals of pipeline vibration, a pressure transmitter for collecting pressure inside the pipeline, and a signal acquisition terminal responsible for signal collection and preliminary processing.

[0029] In this embodiment, the hardware sensing layer also includes a timing module, which is used to add a unified timestamp to the data collected by the infrasound sensor and pressure transmitter of the same monitoring node, so as to realize the synchronization of multi-source data in time.

[0030] Specifically, a GPS timing module is built into the signal acquisition terminal to add a high-precision timestamp to the data collected by the infrasound sensor and pressure transmitter. The server receives data packets from each node, aligns them according to the timestamps, and forms a multi-source data snapshot of the entire pipeline at the same moment.

[0031] The system in this embodiment also includes a data transmission layer, which consists of two parts: the front-end part uses an industrial network switch to interconnect the signal acquisition terminals of each monitoring node; the back-end part uses a communication module to upload the multi-source data acquired by the sensing layer to a remote data processing center.

[0032] In this embodiment, the trend identification includes: performing wavelet packet transform decomposition on the vibration signal collected by the infrasound sensor, extracting the energy proportion of a predetermined characteristic frequency band as a feature value, establishing a time series curve of the feature value, and calculating the rate of change of the time series curve through a sliding window; when the rate of change continuously exceeds a preset threshold, it is determined that there is a blockage trend and a primary warning is triggered.

[0033] In the early stages of slurry pipeline blockage, the blockage point has not yet fully formed, but the deposition or localized accumulation of solid particles on the pipe wall reduces the fluid cross-section, significantly increasing friction between particles and the pipe wall, and between particles themselves. The vibrational energy generated by this frictional excitation is mainly concentrated in the low-frequency infrasound band (1-20Hz). This frequency band has the characteristics of long propagation distance, slow attenuation, and sensitivity to changes in flow state, and can reflect the changes in frictional characteristics in the early stages of blockage.

[0034] Frequency bands above 20Hz are susceptible to interference from operating noises such as pump start-up and shutdown, valve operation, and mechanical vibration, making it difficult to extract stable features related to blockage. Frequency bands below 1Hz mainly exhibit global pressure fluctuations or temperature drift, and are not sensitive to local blockages. Therefore, selecting 1-20Hz as the characteristic frequency band can effectively capture the frictional excitation characteristics in the early stages of blockage while avoiding major noise interference, ensuring the accuracy and robustness of trend identification.

[0035] The determination of this frequency band range is based on spectral analysis experiments using actual operating data from slurry pipelines. In multiple simulated blockage tests, energy variation trends in different frequency bands were extracted. It was found that the energy in the 1-20Hz band showed a significant upward trend 0.5 to 2 hours before blockage occurred, while other frequency bands did not exhibit a stable pattern. Combined with theoretical modeling and spectral response analysis of particle-pipe wall friction excitation, the physical rationality of this frequency band was further verified.

[0036] Specifically, in this embodiment, after preprocessing the raw infrasound signal of each node by denoising and standardization, wavelet packet decomposition is performed to extract the energy proportion of the 1-20Hz frequency band as a feature value. This energy band will significantly increase when the slurry flow is obstructed and particle friction intensifies. The low-frequency energy ratio (E_low) of this frequency band to the total energy is calculated, and a time series of the low-frequency energy ratio value of each node is established. The gradient of its change is calculated using a sliding window (e.g., 10 minutes). If the gradient value of a node exceeds the empirical threshold for several consecutive cycles (e.g., 3 cycles), it is determined that there is a blockage trend in the area near the node, triggering a primary warning, i.e., a "caution" level warning. It should be noted that the sliding window duration and analysis cycle in this embodiment are not fixed values. Those skilled in the art can adapt and adjust them according to actual working conditions and monitoring needs.

[0037] In this embodiment, the location verification involves spatially correlating the abnormal pressure area with the vibration signal intensity to determine the location of the blockage point, including: Calculate the pressure difference between adjacent monitoring nodes. When the pressure gradient of a certain pipe section increases abnormally and the pressure difference exceeds the preset threshold, the section is identified as a pressure abnormality area. Compare the infrasound signal energy values ​​of each monitoring node in the predetermined characteristic frequency band, and take the node with the largest energy value as the monitoring node location with the strongest vibration signal; The coordinates of the blockage point are calculated using a weighted algorithm based on the location of the abnormal pressure area and the monitoring node with the strongest vibration signal intensity.

[0038] The pressure anomaly region is a one-dimensional pipe segment interval, with its center point taken as the spatial reference point; the node with the strongest vibration signal is a discrete point. By weighted fusion of the interval's center point and the strongest node, the spatial association between the "interval" and the "point" is achieved. When the pressure gradient anomaly is significant, As the pressure increases, the positioning results shift towards the center of the abnormal pressure area; when the vibration signal energy concentration is high... The increase in amplitude shifts the location result towards the node with the strongest vibration. This method takes into account the inconsistency in the spatial distribution of the two types of physical fields, improving the robustness of blockage point location.

[0039] The weighting algorithm in this embodiment is: (1) in, The coordinates of the blockage point. The coordinates of the center point of the pressure anomaly region. The coordinates of the monitoring node with the strongest vibration signal intensity; The weight is dynamically set based on the degree of pressure gradient anomaly, and is used to characterize the degree to which the pressure difference in the pressure anomaly area exceeds the normal threshold. The larger the value, the more the positioning result shifts towards the center of the pressure anomaly area. The weight is dynamically set based on the intensity of the vibration signal and is used to characterize the concentration of the infrasound energy of the strongest monitoring node relative to the average energy of the entire line. The larger the value, the more the positioning result is offset towards the location of that node. , satisfy: (2) (3) in, The pressure difference in the area of ​​abnormal pressure. The pressure difference threshold The infrasound energy value of the strongest node. This represents the average infrasound energy across all monitoring nodes.

[0040] In a specific application embodiment, when the data processing and analysis layer receives a "Caution" warning, the system automatically retrieves the pressure data for the entire pipeline during the warning period. It analyzes the pressure distribution curve to identify pipe sections with significantly increased pressure gradients (i.e., areas where upstream pressure is high and downstream pressure is low). This abnormal pressure range is spatially correlated with the location of the node with the strongest infrasound energy. Using a preset pipeline distance model, the most probable location of the blockage point is calculated, and the warning is upgraded to a "Warning" level.

[0041] In this embodiment, the pipeline distance model adopts a geographic information system (GIS)-based approach. A linear reference model is used to establish a mapping relationship between the mileage coordinates along the pipeline and geographical coordinates. The model is represented as follows: (4) in, Here is the arc length parameter along the centerline of the pipeline. The arc length parameter at the starting point of the pipeline. The geographical coordinates of a point on the centerline of the pipeline. This refers to the mileage coordinates of a point on the pipeline. The coordinates of the starting point of the pipeline.

[0042] This model is built based on the centerline coordinate data in the pipeline construction design drawings, through... Calibration is performed on key measured points (such as elbows, valves, and joint locations) to correct deviations between design data and actual installation locations. High-precision calibration is used during the calibration process. The geographic coordinates and mileage values ​​of each monitoring node and pipeline feature point are collected, and a piecewise linear interpolation function is constructed to realize the bidirectional mapping between mileage coordinates and geographic coordinates.

[0043] Based on the calibrated pipeline distance model described above, the steps for calculating the blockage point are as follows: 1. Identify areas of abnormal pressure. Take the mileage of its center point ; 2. Determine the mileage of the node with the strongest infrasound energy. ; 3. Calculate the congestion point mileage based on the above weighted algorithm. ; 4. Using the pipeline distance model, the mileage is... Mapped to geographic coordinates And it is highlighted in the visualization interface.

[0044] This method integrates the calculation of the pressure anomaly range (macroscopic positioning range) with the node with the strongest vibration signal (microscopic sound source direction), achieving precise positioning from "knowing only which segment" to "knowing the specific location". It takes into account the spatial information of the pressure field and the sound field, improving positioning accuracy and engineering usability.

[0045] In this embodiment of risk assessment, the operating parameters include at least one of slurry flow rate and concentration; the risk assessment results include at least one of risk level, expected development speed, and disposal recommendations.

[0046] This embodiment of risk assessment further includes: inputting energy temporal characteristics, blockage location information, and operating parameters into a pre-trained machine learning model, which then outputs risk assessment results. The machine learning model is a blockage risk assessment model trained on historical blockage event data. Inputs include current infrasound energy intensity, pressure variation amplitude, slurry flow velocity, and concentration. The model outputs a risk level (low, medium, high), the predicted rate of blockage escalation, and treatment recommendations, ultimately generating a structured early warning report.

[0047] In this embodiment, the application interaction layer includes a visualization interface for highlighting the location of blockage points on the pipeline geographic information view and displaying the risk assessment results.

[0048] The present invention will be further described below with reference to specific application embodiments. For example... Figure 2 As shown, taking a 30-kilometer-long iron concentrate pipeline as an example, a monitoring node is set up every 3-5 kilometers along the pipeline. Each monitoring node is equipped with one infrasound sensor (piezoelectric type, pressure resistant 14MPa, output 4-20mA) and two pressure transmitters (located before and after the sensor, respectively, with an accuracy of ±0.1%). The sensor signals are connected to a signal acquisition terminal located in an explosion-proof box. This terminal uses a high-performance RTU (Remote Terminal Unit), and the internal structure and functional modules of the RTU include: The power management module supplies power to all modules; The AD conversion module converts the 4-20mA analog signal input from the sensor into a digital signal; The main control processor receives and processes the converted digital signal. The GPS timing module has a timing accuracy of 20ns. It synchronously collects and packages sensor data at a sampling frequency of 100Hz, and adds a precise UTC timestamp to each data sample. The communication interface (RJ45) transmits the packaged data outwards through the industrial network.

[0049] Data transmission layer: Each node RTU is connected to the ring network through an industrial Ethernet switch, and the data is transmitted in real time to the data processing and analysis layer located in the central control room via optical fiber.

[0050] Data Processing and Analysis Layer: The core device is a high-performance server, on which the congestion monitoring and analysis software has a workflow as follows: Figure 3 As shown, the specific steps include: Data reception and synchronization: The server receives data packets from each node, aligns them according to the timestamp, and forms a multi-source data snapshot of the entire pipeline at the same moment.

[0051] Infrasound feature extraction: The raw infrasound signal of each node is preprocessed (denoising and normalization) and wavelet packet decomposition is performed. Special attention is paid to the energy in the 1-20Hz low-frequency band, which significantly increases when slurry flow is obstructed and particle friction intensifies. The proportion of this frequency band to the total energy is calculated and denoted as E_low.

[0052] Trend identification and initial early warning: A time series of E_low values ​​is established for each node, and the gradient of E_low changes is calculated using a sliding time window (e.g., 10 minutes). If the gradient value of a node exceeds the empirical threshold for multiple consecutive periods (e.g., 3 periods), it is determined that a congestion trend has occurred in the vicinity of that node, triggering an "attention" level early warning.

[0053] Multi-source fusion positioning verification: Upon receiving a "Caution" warning, the system automatically retrieves pressure data for the entire pipeline during the warning period, analyzes the pressure distribution curve, and identifies pipe sections with significantly increased pressure gradients (i.e., areas where upstream pressure rises and downstream pressure falls). The system spatially correlates the abnormal pressure intervals with the locations of nodes with the strongest infrasound energy, and, combined with a pre-set pipeline distance model, calculates the most probable location of the blockage point (e.g., "15.3 km from the pumping station"), and upgrades the warning to the "Warning" level.

[0054] Risk Assessment and Report Generation: The system invokes a blockage risk assessment model, which is trained based on historical blockage event data. Input parameters include the current infrasound energy intensity, pressure change amplitude, slurry flow rate, and concentration. The model outputs a risk level (low, medium, high), the expected rate of blockage escalation, and treatment recommendations (e.g., "It is recommended to arrange flushing within 2 hours"), ultimately generating a structured early warning report.

[0055] Application interaction layer: Early warning reports and real-time data are pushed to the operator's workstation. The human-machine interface is as follows: Figure 4 As shown, the main screen displays a geographic information map of the pipeline, highlighting the location of blockages with flashing icons. A sidebar pops up an early warning card, centrally displaying location information, risk level, real-time trend curves, and handling suggestions. Simultaneously, the system triggers audible and visual alarms based on the risk level and can automatically push critical information to management personnel's mobile terminals.

[0056] This embodiment of the system adopts a four-layer architecture: hardware perception layer, data transmission layer, data processing and analysis layer, and application interaction layer. The functions of each layer are decoupled, facilitating flexible deployment and expansion based on pipeline length, geographical environment, and communication conditions. Key parameters such as monitoring node spacing, sliding window duration, and early warning thresholds can all be adjusted according to actual working conditions, adapting to different mineral types, concentrations, and pipe diameters, demonstrating excellent engineering adaptability. Through the collaborative work of the aforementioned hardware and software, the system achieves fully automated intelligent monitoring of slurry pipeline blockages, from "micro-sign detection" to "precise location" and then to "risk quantification assessment," greatly enhancing the pipeline's ability to ensure safe operation.

[0057] This embodiment also provides a pipeline blockage monitoring method based on multi-source data fusion, including the following steps: Step S1: Synchronously collect vibration signals and pressure data from multiple monitoring nodes along the pipeline; Step S2: Perform spectrum analysis on the vibration signal, extract the energy time-series characteristics of the predetermined characteristic frequency band, and determine whether there is a blockage trend based on the rate of change of the energy time-series characteristics; Step S3: If a blockage trend is determined, retrieve the pressure data, analyze the spatial distribution gradient of pressure along the pipeline, and spatially correlate the abnormal pressure area with the vibration signal intensity to determine the location of the blockage point; Step S4: Generate risk assessment results based on energy timing characteristics, location information of blockage points, and pipeline operating parameters; Step S5: Output the risk assessment results and / or the location information of the blockage point.

[0058] In step S3 of this embodiment, spatial correlation is performed between the abnormal pressure region and the vibration signal intensity, including: Identify pipe sections with abnormally high pressure in the pressure gradient distribution as pressure anomaly regions. Determine the location of the monitoring node with the strongest vibration signal intensity; The coordinates of the blockage point are calculated using a weighted algorithm based on the location of the abnormal pressure area and the monitoring node with the strongest vibration signal intensity.

[0059] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention should fall within the protection scope of the present invention.

Claims

1. A pipeline blockage monitoring system based on multi-source data fusion, characterized in that, include: The hardware sensing layer includes multiple monitoring nodes deployed along the pipeline. Each monitoring node includes at least an infrasound sensor for collecting pipeline vibration signals and a pressure transmitter for collecting pressure inside the pipeline. The data processing and analysis layer includes at least one processor, the processor being configured as follows: Trend identification: Perform spectrum analysis on the vibration signal collected by the infrasound sensor, extract the energy time-series characteristics of the predetermined characteristic frequency band, and determine whether there is a blockage trend based on the rate of change of the energy time-series characteristics; Location verification: When a blockage trend is determined, the pressure data collected by the pressure transmitter is retrieved, the spatial distribution gradient of the pressure along the pipeline is analyzed, and the abnormal pressure area is spatially correlated with the vibration signal intensity to determine the location information of the blockage point. Risk assessment: Based on energy timing characteristics, location information of blockage points, and pipeline operating parameters, risk assessment results are generated; The application interaction layer is used to output risk assessment results and / or the location information of blockage points.

2. The pipeline blockage monitoring system based on multi-source data fusion according to claim 1, characterized in that, The hardware sensing layer also includes a timing module, which is used to add a unified timestamp to the data collected by the infrasound sensor and pressure transmitter at the same monitoring node, so as to realize the synchronization of multi-source data in time.

3. The pipeline blockage monitoring system based on multi-source data fusion according to claim 1, characterized in that, In the trend identification, the spectrum analysis includes: performing wavelet packet transform decomposition on the vibration signal collected by the infrasound sensor, extracting the energy proportion of a predetermined characteristic frequency band as a feature value, and establishing a time-series curve of the feature value; calculating the rate of change of the time-series curve through a sliding window; when the rate of change continuously exceeds a preset threshold, it is determined that there is a blockage trend and a primary warning is triggered.

4. The pipeline blockage monitoring system based on multi-source data fusion according to claim 1, characterized in that, In the location verification, the step of spatially correlating the abnormal pressure area with the vibration signal intensity to determine the location of the blockage point includes: Calculate the pressure difference between adjacent monitoring nodes. When the pressure gradient of a certain pipe section increases abnormally and the pressure difference exceeds the preset threshold, the section is identified as a pressure abnormality area. Compare the infrasound signal energy values ​​of each monitoring node in the predetermined characteristic frequency band, and take the node with the largest energy value as the monitoring node location with the strongest vibration signal; Based on the abnormal pressure area and the location of the monitoring node with the strongest vibration signal intensity, the coordinates of the blockage point are calculated using a weighted algorithm.

5. The pipeline blockage monitoring system based on multi-source data fusion according to claim 4, characterized in that, The weighting algorithm is as follows: in, The coordinates of the blockage point. The coordinates of the center point of the pressure anomaly region. The coordinates of the monitoring node with the strongest vibration signal intensity; The weights are dynamically set based on the degree of pressure gradient anomaly. The weights are dynamically set based on the vibration signal intensity, satisfying: in The pressure difference in the area of ​​abnormal pressure. The pressure difference threshold The infrasound energy value of the strongest node. This represents the average infrasound energy across all monitoring nodes.

6. The pipeline blockage monitoring system based on multi-source data fusion according to claim 1, characterized in that, In the risk assessment, the operating parameters include at least one of slurry flow rate and concentration; the risk assessment results include at least one of risk level, expected development speed, and disposal recommendations.

7. The pipeline blockage monitoring system based on multi-source data fusion according to claim 6, characterized in that, The risk assessment further includes: inputting energy time-series characteristics, blockage location information, and operating parameters into a pre-trained machine learning model, and having the machine learning model output the risk assessment result.

8. The pipeline blockage monitoring system based on multi-source data fusion according to claim 1, characterized in that, The application interaction layer includes a visualization interface for highlighting the location of blockage points on the pipeline geographic information view and displaying risk assessment results.

9. A pipeline blockage monitoring method based on multi-source data fusion, characterized in that, Includes the following steps: Step S1: Synchronously collect vibration signals and pressure data from multiple monitoring nodes along the pipeline; Step S2: Perform spectrum analysis on the vibration signal, extract the energy time-series characteristics of the predetermined characteristic frequency band, and determine whether there is a blockage trend based on the rate of change of the energy time-series characteristics; Step S3: If a blockage trend is determined, retrieve the pressure data, analyze the spatial distribution gradient of pressure along the pipeline, and spatially correlate the abnormal pressure area with the vibration signal intensity to determine the location of the blockage point; Step S4: Based on the energy timing characteristics, the location information of the blockage point, and the pipeline operating parameters, generate a risk assessment result; Step S5: Output the risk assessment results and / or the location information of the blockage point.

10. The pipeline blockage monitoring method based on multi-source data fusion according to claim 9, characterized in that, In step S3, the spatial correlation between the abnormal pressure region and the vibration signal intensity includes: The pipe segment interval with abnormally high pressure in the pressure gradient distribution is identified as the pressure anomaly region; Determine the location of the monitoring node with the strongest vibration signal intensity; The coordinates of the blockage point are calculated using a weighted algorithm based on the location of the abnormal pressure area and the monitoring node with the strongest vibration signal intensity.

Images