A method and system for monitoring the liquid fill rate of a pipeline

By combining a neural network prediction model with temperature and vibration parameters, the filling rate of high-energy pipelines can be monitored in real time, solving the problems of insufficient reliability and timeliness in existing technologies. This achieves efficient and low-cost filling rate monitoring, avoiding engineering modifications.

CN117366482BActive Publication Date: 2026-11-10易感科技有限公司 +1
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202310736087.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2023-04-10
Filing Date
2023-06-20
Publication Date
2026-11-10
Estimated Expiration
2043-06-20

AI Technical Summary

Technical Problem

Existing high-energy pipeline filling rate monitoring technologies are inadequate in terms of reliability, accuracy, and timeliness, which affects industrial safety and production. Furthermore, existing methods require engineering modifications and incur high costs.

Method used

A neural network prediction model is used in conjunction with temperature and vibration parameters. The temperature and vibration data of the pipeline are monitored in real time through a sensing device. The neural network prediction model is constructed to accurately monitor the filling rate. The sensing device adopts wireless transmission and self-powered mode, eliminating the need for wiring and engineering modification.

Benefits of technology

It enables accurate monitoring of the filling rate of high-energy pipelines, improves the timeliness and reliability of monitoring, and reduces installation and maintenance costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117366482B_ABST
    Figure CN117366482B_ABST
Patent Text Reader

Abstract

The application provides a pipeline liquid filling rate monitoring method and system, and the pipeline liquid filling rate monitoring method comprises the following steps: obtaining and processing historical temperature data flow and historical vibration parameter data flow of a pipeline; constructing a neural network prediction model by using the historical temperature data flow and the historical vibration parameter data flow; obtaining the temperature and vibration parameters of the pipeline in real time and inputting the same into the neural network prediction model to obtain the liquid filling rate of the pipeline. The neural network prediction model of the liquid filling rate is constructed by using the historical temperature data flow and the historical vibration parameter data flow of the pipeline, artificial neural network is used for learning and training, and the data flow high-energy pipeline liquid filling rate identification based on neural network prediction is realized. The temperature and vibration data collected in real time are used as the input of the neural network prediction model, the final predicted high-energy pipeline liquid filling rate is obtained, and therefore, the purpose of monitoring the high-energy pipeline liquid filling rate based on the temperature and vibration parameters is realized, and the prediction effect is high in accuracy and timeliness.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of pipeline monitoring technology, and in particular to a method and system for monitoring pipeline filling rate. Background Technology

[0002] In industries such as nuclear energy, power, petrochemicals, and metallurgy, there are numerous high-energy pipelines transporting high-temperature liquid working fluids (such as high-temperature water). During the transport of these working fluids, if leaks occur due to corrosion or other reasons, causing continuous leakage, the working fluid inside the pipeline will no longer be in a stable operating state but will transition to a two-phase state of vapor and liquid with an increasingly lower liquid filling rate. Monitoring the physical state of the working fluid inside high-energy pipelines is of great significance to the thermal system in which these pipelines are located. For example, a deviation of the liquid filling rate from the design allowable range will cause unexpected vibrations in the pipeline. If the pipeline is subjected to alternating stress caused by vibration for a long period of time, fatigue failure may occur, especially at weak points such as welds, which can easily lead to more serious working fluid leakage accidents such as weld breakage and pipeline fracture, resulting in serious property damage, threatening personal safety, and causing harm to the environment. Therefore, taking appropriate measures to monitor the liquid filling rate of high-energy pipelines, detecting and correcting deviations from normal operating conditions as early as possible, eliminating potential safety hazards, and preventing them from inducing more serious industrial production safety accidents are crucial.

[0003] Currently, high-energy pipeline filling rate monitoring technologies are mostly based on differential pressure flow meters or ultrasonic flow meters installed on the pipeline, or on monitoring the working fluid that has already leaked from the pipeline. Since the working fluid in the pipeline is generally a high-temperature, high-pressure fluid, the leaked high-temperature working fluid will gradually cause the ambient temperature near the pipeline to rise. At the same time, the working fluid is eventually collected in a sump within the plant where the pipeline is located. Existing technologies generally detect the leakage indirectly by monitoring ambient temperature and sump liquid level, and then infer the pipeline filling rate. A typical example is the leakage monitoring system of a pressurized water reactor nuclear power plant, which is used to monitor the integrity of the reactor coolant system pressure boundary (RCPB). The method is to determine the leakage of the reactor coolant system by calculating the changes in sump liquid level (and thus calculate the filling rate of the primary coolant pipeline).

[0004] Therefore, the existing high-energy pipeline filling rate monitoring technology has the following shortcomings in terms of actual operation and maintenance availability: 1) Relatively low reliability and accuracy: There are many factors affecting the ambient temperature near the pipeline, such as room ventilation and the operation of the air conditioning system. At the same time, there are also many factors affecting the sump level, such as waste liquid generated during external cleaning. These may interfere with the judgment of pipeline leakage, thus affecting the reliability and accuracy of the entire filling rate calculation; 2) Insufficient monitoring timeliness: The working fluid that diffuses into the surrounding environment due to leakage needs to accumulate to a certain mass before it can cause the plant ambient temperature or the sump level to rise. This results in the entire monitoring system requiring a long response time, which poses a threat to industrial safety production. Summary of the Invention

[0005] In order to solve the problems existing in the prior art, the present invention aims to provide a method and system for monitoring pipeline filling rate, which can accurately monitor the pipeline filling rate.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A method for monitoring pipeline filling rate includes the following steps:

[0008] Acquire and process historical temperature data streams and historical vibration parameter data streams of the pipeline;

[0009] A neural network prediction model was constructed using historical temperature data streams and historical vibration parameter data streams.

[0010] The temperature and vibration parameters of the pipeline are acquired in real time and input into the neural network prediction model to obtain the filling rate of the pipeline.

[0011] As a further improvement of the present invention, the step of acquiring and processing the historical temperature data stream and historical vibration parameter data stream of the pipeline includes:

[0012] Acquire historical temperature data streams and historical vibration parameter data streams for the pipeline;

[0013] Preprocess the historical temperature data stream and historical vibration parameter data stream;

[0014] Feature parameters are extracted from historical temperature data streams and historical vibration parameter data streams.

[0015] As a further improvement of the present invention, the step of preprocessing the historical temperature data stream and historical vibration parameter data stream of the pipeline includes:

[0016] In the historical temperature data stream and historical vibration parameter data stream, data segments, isolated data segments, and abnormal data points that are running in an unsteady state are removed;

[0017] Supplement the missing data.

[0018] As a further improvement of the present invention, the step of constructing a neural network prediction model using historical temperature data streams and historical vibration parameter data streams includes:

[0019] Construct a neural network prediction model using feature parameters;

[0020] The feature parameters are divided into a first training set and a first test set;

[0021] The historical filling rate data was divided into a second training set and a second test set.

[0022] The neural network prediction model is trained by using the first training set as input and the second training set as output until it converges.

[0023] The neural network prediction model is tested using the first test set as input and the second training set as output to obtain the neural network prediction model.

[0024] As a further improvement of the present invention, when testing the neural network prediction model, if the accuracy is less than a set value, the model is re-divided into a first training set, a first test set, a second training set, and a second test set, and the neural network prediction model is retrained and tested.

[0025] As a further improvement of the present invention, the characteristic parameters include: compression ratio, average rising slope of vibration data, average falling slope of vibration data, average rising slope of temperature difference data, average falling slope of temperature difference data, number of extreme points of vibration data, average value of vibration data stream, and root mean square error of vibration data stream.

[0026] In addition, the present invention also provides a pipeline filling rate monitoring system, comprising:

[0027] Sensing devices;

[0028] A computer device is connected to the sensing device to acquire historical temperature data streams, historical vibration parameter data streams, real-time temperature, and real-time vibration parameters of the pipeline. The computer device includes a processor and a memory. The memory stores at least one instruction, at least one program, code set, or instruction set. The at least one instruction, at least one program, code set, or instruction set is loaded and executed by the processor to implement the above-described pipeline filling rate monitoring method.

[0029] As a further improvement of the present invention, the sensing device includes:

[0030] Thermoelectric conversion module, installed on the pipe wall, is used to convert the heat of the pipe into electrical energy;

[0031] A power management module, connected to the thermoelectric conversion module, is used to acquire electrical energy and boost its voltage.

[0032] A temperature sensor is installed on the pipe wall to obtain the temperature; the temperature sensor is connected to the power management module to obtain electrical energy.

[0033] A vibration sensor is installed on the pipe wall to acquire the vibration parameters; the vibration sensor is connected to the power management module to obtain electrical energy.

[0034] An ambient temperature sensor is installed in the environment in which the pipeline operates to acquire the ambient temperature; the ambient temperature sensor is connected to the power management module to obtain electrical energy.

[0035] The microcontroller unit is connected to the temperature sensor and the vibration sensor to obtain the temperature and vibration parameters; the microcontroller unit is also connected to the power management module to obtain electrical energy.

[0036] A wireless transmission module, connected to the microcontroller unit, is used to send the pipe's temperature and vibration parameters to the computer device under the control of the microcontroller unit; the wireless transmission module is also connected to the power management module to obtain electrical energy.

[0037] As a further improvement of the present invention, the present invention also includes: a shielding box, wherein the power management module, the microcontroller unit, and the wireless transmission module are placed inside the shielding box.

[0038] The beneficial effects of this invention are as follows: A neural network prediction model reflecting the filling rate of high-energy pipelines is constructed using historical temperature and vibration parameter data streams of the pipeline. An artificial neural network is used for learning and training, thereby achieving high-energy pipeline filling rate identification based on the neural network prediction data stream. Finally, real-time collected temperature and vibration data are used as input to the neural network prediction model to obtain the final predicted high-energy pipeline filling rate. This achieves the goal of monitoring the filling rate of high-energy pipelines based on temperature and vibration parameters, with high accuracy and timeliness in prediction. Attached Figure Description

[0039] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0040] Figure 1 This is a flowchart of the pipeline filling rate monitoring method described in Example 1;

[0041] Figure 2 This is another flowchart of the pipeline filling rate monitoring method described in Example 1;

[0042] Figure 3This is a schematic diagram of the pipeline filling rate monitoring structure described in Example 2; Detailed Implementation

[0043] To make the technical problems solved by the present invention, the technical solutions adopted, and the technical effects achieved clearer, the technical solutions of the embodiments of the present invention will be further described in detail below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] Example 1

[0045] This embodiment provides a method for monitoring the liquid filling rate of a pipeline, such as... Figure 1 and Figure 2 As shown, the steps include:

[0046] S1. Acquire and process historical temperature and vibration parameter data streams of the pipeline. A data stream is a continuous, infinite, rapid sequence of time-varying data elements, generating large amounts of data quickly over an approximately infinite time range. Sensors collect temperature and vibration data from the pipeline at a specific frequency (typically 2Hz to 5000Hz), resulting in a stable data stream.

[0047] Specifically, step S1 includes:

[0048] S12. Obtain the historical temperature data stream and historical vibration parameter data stream of the pipeline.

[0049] S13. Preprocess the historical temperature data stream and historical vibration parameter data stream: Preprocessing mainly includes data removal and data supplementation.

[0050] In historical temperature data streams and historical vibration parameter data streams, data segments, isolated data segments, and abnormal data points that are not in steady-state operation are removed. During unsteady-state operation, the fluid in the high-energy pipeline is in an unstable flow state, and the filling rate is constantly changing. Typical unsteady-state operation includes pipeline filling with water, preheating, pump start-up / shutdown in the pipeline loop, and opening of valves on the pipeline (steam vent valves, drain valves, etc.). When performing trend analysis on historical temperature data streams and historical vibration parameter data streams, such unsteady-state operation data segments should be removed to ensure the accuracy of the trend analysis.

[0051] Due to data transmission failures, sensor malfunctions, and short-term continuous external electromagnetic interference, isolated data segments exist in the historical monitoring data stream. These isolated segments are characterized by excessively large time intervals between adjacent data points or too few data points within a given period. Since these isolated segments cannot reflect the data's trend, they should be deleted. During operation, various factors cause momentary fluctuations in sensor signals, resulting in numerous outliers in the monitoring data. To avoid the impact of these outliers on subsequent analysis, they must be removed. Box plots or other methods can be used to identify and eliminate outliers. For example, the upper and lower edges of normal values ​​can be obtained by calculating the first and third quartiles of temperature and vibration data, and data outside these edges can be discarded.

[0052] Meanwhile, considering the possibility of data packet loss during wireless signal transmission, it is also necessary to supplement the missing data, such as interpolating temperature and vibration data to obtain replacement values ​​for missing and abnormal data.

[0053] S14. Extract feature parameters from historical temperature data stream and historical vibration parameter data stream. In this embodiment, a variable sliding window algorithm is used for trend analysis of historical temperature data stream and historical vibration parameter data stream to reasonably segment the data sequence and improve the accuracy of trend analysis. Dynamic data stream trend analysis extracts the segmentation points that describe the curve characteristics. The curve features of the data stream are extracted using the parameters of each segmentation point and the data segments between the segmentation points. Dynamic feature extraction and judgment are performed on the data stream, and each feature data is used as the input parameters of the neural network to model the feature parameters in order to achieve the purpose of identifying and classifying the curves of each feature parameter. Then, pattern recognition is performed based on the data stream features to accurately classify the data, thereby achieving effective identification of the filling rate status of high-energy pipelines.

[0054] The characteristic parameters include:

[0055] Compression ratio Num, Num = data volume of data stream / number of split points;

[0056] Average rise slope of vibration data avg_A_up:

[0057]

[0058] Among them: i=1, 2...n; n<Num; A_up i The slope of the vibration data stream is the slope of the data segment that is positive.

[0059] Average slope of vibration data avg_A_down:

[0060]

[0061] Among them: i=1, 2...n; n<Num; A_down i The slope of the vibration data stream is the negative slope of the data segment.

[0062] Average slope of temperature difference data avg_T_up:

[0063]

[0064] Among them: i=1, 2...n; n<Num; T_up i The slope of the data segment is positive when the slope of the difference between the pipe wall temperature and the ambient temperature.

[0065] Average slope of temperature difference data avg_T_down:

[0066]

[0067] Among them: i=1, 2...n; n<Num; T_down i The slope of the data segment is negative, representing the slope of the change in the difference between the pipe wall temperature and the ambient temperature.

[0068] Number of extreme points in vibration data: Extreme_num (If the slope signs of the data segments before and after the split point are opposite, the split point is an extreme point).

[0069] Vibration data stream average value avg_D:

[0070]

[0071] Where N is the data volume of the data stream, and D i Let be the value of the i-th vibration data.

[0072] The mean square error θ of the vibration data stream:

[0073]

[0074] Where N is the data volume of the data stream, and D i Let be the value of the i-th vibration data.

[0075] S2. Construct a neural network prediction model using historical temperature data streams and historical vibration parameter data streams. The following sections will further explain the specific steps:

[0076] S21. Construct a neural network prediction model using feature parameters. The number of nodes in the input layer, hidden layer, and output layer of the neural network prediction model is set according to specific requirements. In this embodiment, the input parameters are the aforementioned eight feature parameters, and the output layer is one node representing the pipe filling rate state. The neural network prediction model can be represented as:

[0077] s=net(Num, avg_A_up, avg_A_down, avg_T_up,

[0078] avg-T-down, Extreme_num, avg_D, θ)

[0079] The hidden layers of the neural network prediction model use the tangent sigmoid function, and the output layer uses the logarithmic sigmoid function. This embodiment does not limit the number of hidden layer units; preferably, 13 hidden layer nodes are used.

[0080] S22. Divide the feature parameters into the first training set and the first test set.

[0081] S23. Divide the historical filling rate data into a second training set and a second test set.

[0082] S24. Using the first training set as input and the second training set as output, train the neural network prediction model until it converges, as shown in Table 1:

[0083] Table 1 shows some examples of training samples.

[0084]

[0085] S25. Using the first test set as input and the second training set as output, test the neural network prediction model to obtain the neural network prediction model. When testing the neural network prediction model, if the accuracy is less than 95%, divide it into the first training set, the first test set, the second training set, and the second test set, and retrain and test the neural network prediction model until the accuracy is greater than or equal to 95%.

[0086] S3. Real-time acquisition of the temperature and vibration parameters of the pipeline, extraction of characteristic parameters from the temperature and vibration parameters, and input into the neural network prediction model to obtain the pipeline filling rate, as shown in Table 2.

[0087] Table 2

[0088] 0<=s<0.55 100% filling rate 0.55<=s<0.7 90% ≤ Filling rate < 100% 0.7<=s<0.8 80% ≤ Filling rate < 90% 0.8<=s<0.95 60% ≤ Filling rate < 80% 0.95<=s<1 Filling rate <60%

[0089] This embodiment leverages the data stream characteristics of high-energy pipeline temperature and vibration parameters. It utilizes data stream technology to analyze the historical temperature and vibration data of the high-energy pipeline, extracting the dynamic trends during its operation. Based on this, a neural network prediction model reflecting the filling rate of the high-energy pipeline is constructed using feature parameters. An artificial neural network is used for learning and training, thereby realizing a high-energy pipeline filling rate identification model based on neural network prediction data streams. Finally, trend analysis is performed on the real-time acquired temperature and vibration data streams to extract feature parameters. These feature data are used as input to the neural network prediction model to obtain the final predicted filling rate of the high-energy pipeline. This achieves the goal of monitoring the filling rate of high-energy pipelines based on temperature and vibration parameters, with high accuracy and timeliness in prediction.

[0090] Example 2

[0091] Existing liquid filling rate monitoring systems require the installation of flow meters on pipelines. These meters need to be installed directly on high-energy pipelines, and their installation and deployment involve a large number of engineering modifications. For example, the installation of differential pressure flow meters requires cutting off high-energy pipelines. Moreover, the power supply, signal acquisition wiring, and maintenance of the meters all consume a lot of manpower and financial resources.

[0092] To address the aforementioned problems, this embodiment provides a pipeline filling rate monitoring system, such as... Figure 2 As shown, it includes: a sensing device and a computer device. The computer device is connected to the sensing device to acquire historical temperature data stream, historical vibration parameter data stream, real-time temperature, and real-time vibration parameters of the pipeline. The computer device includes a processor and a memory. The memory stores at least one instruction, at least one program, code set, or instruction set. At least one instruction, at least one program, code set, or instruction set is loaded and executed by the processor to implement the pipeline filling rate monitoring method in Embodiment 1.

[0093] The sensing device includes: a thermoelectric conversion module, a power management module, a temperature sensor, a vibration sensor, an ambient temperature sensor, a microcontroller unit, and a wireless transmission module. The thermoelectric conversion module is mounted on the pipe wall and converts heat from the pipe into electrical energy. The power management module is connected to the thermoelectric conversion module and acquires and boosts the electrical energy. The temperature sensor is mounted on the pipe wall and acquires the temperature; it is connected to the power management module to acquire electrical energy. The vibration sensor is mounted on the pipe wall and acquires vibration parameters; it is connected to the power management module to acquire electrical energy. The ambient temperature sensor is located in the environment in which the pipe operates and acquires the ambient temperature; it is connected to the power management module to acquire electrical energy. The microcontroller unit is connected to the temperature sensor and the vibration sensor to acquire temperature and vibration parameters; it is connected to the power management module to acquire electrical energy. The wireless transmission module is connected to the microcontroller unit and, under the control of the microcontroller unit, transmits the pipe's temperature and vibration parameters to a computer device; it is connected to the power management module to acquire electrical energy.

[0094] In the specific implementation process, the thermoelectric conversion module converts the temperature difference between the high-energy pipeline and the environment into electrical energy. The generated electrical energy is transmitted to the power management module for unified management. The power management module realizes functions such as boosting, stabilizing, and storing electrical energy, and provides power to the temperature sensor, vibration sensor, microcontroller unit, and wireless transmission module.

[0095] In addition, this embodiment also includes a shielding box, in which the power management module, microcontroller unit, and wireless transmission module are housed. The shielding box can reduce or eliminate potential environmental electromagnetic interference.

[0096] Because the sensing device in this embodiment uses wireless data transmission and is self-powered, it allows for flexible installation and deployment, offering advantages such as no wiring required and no engineering modifications needed. Combined with the pipeline filling rate monitoring in Embodiment 1, it enables the monitoring of the pipeline filling rate.

[0097] It should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This way of describing the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

[0098] The technical principles of the present invention have been described above with reference to specific embodiments. These descriptions are merely for explaining the principles of the invention and should not be construed as limiting the scope of protection of the invention in any way. Based on this explanation, those skilled in the art can readily conceive of other specific embodiments of the invention without inventive effort, and these embodiments will all fall within the scope of protection of the present invention.

Claims

1. A method for monitoring pipeline filling rate, characterized in that, Including the following steps: Acquire and process historical temperature data streams and historical vibration parameter data streams of pipelines, including: acquiring historical temperature data streams and historical vibration parameter data streams of pipelines; preprocessing historical temperature data streams and historical vibration parameter data streams; and extracting feature parameters from historical temperature data streams and historical vibration parameter data streams. Specifically, this includes: using a variable sliding window algorithm to segment the data sequence accordingly for trend analysis of historical temperature data streams and historical vibration parameter data streams, thereby improving the accuracy of trend analysis; Dynamic data flow trend analysis extracts segmentation points that describe the curve features of the data flow. It uses the parameters of each segmentation point and the data segments between the segmentation points to extract the curve features of the data flow. It performs dynamic feature extraction and judgment on the data flow and uses each feature data as the input parameters of the neural network to model the feature parameters in order to identify and classify the curves of each feature parameter. Then, it performs pattern recognition based on the data flow features to accurately classify the data flow and thus achieve effective identification of the filling rate status of high-energy pipelines. The characteristic parameters include: Compression ratio Num, Num = data volume of the data stream / number of split points; Average rise slope of vibration data avg_A_up: Where: i = 1, 2, ..., n; n < Num; The slope of the vibration data stream is the slope of the data segment that is positive; Average slope of vibration data avg_A_down: Where: i = 1, 2, ..., n; n < Num; The slope of the vibration data stream is the negative slope of the data segment; Average slope of temperature difference data (avg_T_up): Where: i = 1, 2, ..., n; n < Num; The slope of the data segment is positive when the slope of the difference between the pipe wall temperature and the ambient temperature. Average slope of temperature difference data (avg_T_down): Where: i = 1, 2, ..., n; n <Num; The slope of the data segment where the slope of the change in pipe wall temperature versus ambient temperature is negative; Vibration data extreme points: Extreme_num, if the slope signs of the data segments before and after the division point are opposite, the division point is an extreme point; Vibration data stream average value avg_D: Where N is the data volume of the data stream. This represents the value of the i-th vibration data point; Mean square error of vibration data stream : Where N is the data volume of the data stream. This represents the value of the i-th vibration data point; A neural network prediction model was constructed using historical temperature data streams and historical vibration parameter data streams. The steps for constructing a neural network prediction model using historical temperature data streams and historical vibration parameter data streams include: A neural network prediction model is constructed using feature parameters; the number of nodes in the input layer, hidden layer, and output layer of the neural network prediction model is set according to specific requirements. The neural network prediction model is represented as follows: s=net(Num,avg_A_up,avg_A_down,avg_T_up, avg-T-down,Extreme_num,avg_D, θ) The hidden layers of the neural network prediction model use the tangent sig function, and the output layer uses the logarithmic sig function. The feature parameters are divided into a first training set and a first test set; The historical filling rate data was divided into a second training set and a second test set. The neural network prediction model is trained by using the first training set as input and the second training set as output until it converges. The neural network prediction model is tested using the first test set as input and the second training set as output to obtain the neural network prediction model. The temperature and vibration parameters of the pipeline are acquired in real time and input into the neural network prediction model to obtain the filling rate of the pipeline.

2. The pipeline filling rate monitoring method according to claim 1, characterized in that, The steps for preprocessing the historical temperature data stream and historical vibration parameter data stream of the pipeline include: In the historical temperature data stream and historical vibration parameter data stream, data segments, isolated data segments, and abnormal data points that are running in an unsteady state are removed; Supplement the missing data.

3. The pipeline filling rate monitoring method according to claim 1, characterized in that, When testing the neural network prediction model, if the accuracy is less than the set value, the model is re-divided into the first training set, the first test set, the second training set, and the second test set for retraining and testing.

4. A pipeline filling rate monitoring system, characterized in that, include: Sensing devices; A computer device is connected to the sensing device to acquire historical temperature data stream, historical vibration parameter data stream, real-time temperature, and real-time vibration parameters of the pipeline. The computer device includes a processor and a memory. The memory stores at least one instruction, code set, or instruction set. The at least one instruction, code set, or instruction set is loaded and executed by the processor to implement the pipeline filling rate monitoring method as described in any one of claims 1 to 3. The sensing device includes: Thermoelectric conversion module, installed on the pipe wall, is used to convert the heat of the pipe into electrical energy; A power management module, connected to the thermoelectric conversion module, is used to acquire electrical energy and boost its voltage. A temperature sensor is installed on the pipe wall to obtain the temperature; the temperature sensor is connected to the power management module to obtain electrical energy. A vibration sensor is installed on the pipe wall to acquire the vibration parameters; the vibration sensor is connected to the power management module to obtain electrical energy. An ambient temperature sensor is installed in the environment in which the pipeline operates to acquire the ambient temperature; the ambient temperature sensor is connected to the power management module to obtain electrical energy. The microcontroller unit is connected to the temperature sensor and the vibration sensor to obtain the temperature and vibration parameters; the microcontroller unit is also connected to the power management module to obtain electrical energy. A wireless transmission module, connected to the microcontroller unit, is used to send the pipe's temperature and vibration parameters to the computer device under the control of the microcontroller unit; the wireless transmission module is also connected to the power management module to obtain electrical energy.

5. The pipeline filling rate monitoring system according to claim 4, characterized in that, Also includes: The power management module, the microcontroller unit, and the wireless transmission module are housed within the shielded box.

Citation Information

Patent Citations

  • Oil and gas pipeline leakage detection method based on double parameters of vibration and temperature

    CN112729688A