A kind of operation and maintenance management system suitable for coke oven string leakage prevention

Abstract

The application discloses a kind of operation and maintenance management systems suitable for coke oven string leakage prevention, belong to coke oven string leakage detection identification field, specifically including operation and maintenance management platform, working condition data acquisition module, data processing analysis module, diagnosis positioning module and hierarchical early warning module;The application is to construct multi-point distribution monitoring network for the key parts of coke oven object, deeply fusion temperature monitoring, pressure monitoring and masonry apparent monitoring, as the front guide early warning of auxiliary gas leakage detection, based on the cross-validation analysis of multi-source data fusion, reach early warning and accurate positioning of string leakage, to solve the hysteresis problem of single gas monitoring, realize the operation and maintenance management system of string leakage prevention of string leakage coke oven with "predictive proactive maintenance" as core.

Description

Technical Field

[0001] This invention relates to the field of coke oven leakage detection and identification technology, and more specifically, to an operation and maintenance management system suitable for preventing coke oven leakage. Background Technology

[0002] During the operation of coke ovens, leakage within the oven body is a relatively common problem. Because the coke oven body needs to withstand high temperatures for a long time, the refractory material is prone to peeling and cracking, resulting in gaps in the oven body. In addition, improper operation and aging of equipment sealing materials may also lead to leakage problems. Once leakage occurs in the coke oven body, it will cause changes in the chemical composition during the coking process, thereby affecting the quality of coke.

[0003] Currently, by studying the causes and patterns of furnace leakage, effective measures can be taken for prevention and control, ensuring the stability of the coking process and the reliability of coke quality, and improving the market competitiveness of products. For example, patent number CN120176946A discloses a coke oven leakage identification and positioning system based on data acquisition and analysis. The core of this patent is to identify and locate leakage through internal and external gas composition analysis. However, relying solely on gas analysis can easily lead to a problem of delayed early warning.

[0004] To enhance the early warning capability and comprehensiveness of the system, an operation and maintenance management system suitable for coke oven leakage prevention is proposed. By introducing temperature monitoring, pressure monitoring, and masonry surface anomaly monitoring, a multi-dimensional, early warning leakage prevention operation and maintenance management system is constructed. Summary of the Invention

[0005] The purpose of this invention is to solve practical problems and provide an operation and maintenance management system suitable for preventing coke oven leakage compared with existing technologies.

[0006] The objective of this invention can be achieved through the following technical solution: an operation and maintenance management system suitable for preventing coke oven leakage, including an operation and maintenance management platform, an operating condition data acquisition module, a data processing and analysis module, a diagnostic and location module, and a graded early warning module;

[0007] The operation and maintenance management platform is used to mark temperature measuring points, pressure measuring points, gas leakage measuring points, and surface image monitoring surfaces for key parts of the coke oven, and to build a multi-point distributed monitoring network.

[0008] The operating condition data acquisition module includes an infrared temperature field monitoring module, a pressure field monitoring module, a leak gas monitoring module, and an image acquisition module. These modules acquire operating condition data for temperature measuring points, pressure measuring points, and gas leak measuring points, respectively. Based on the operating condition data, they generate temperature sequences, pressure sequences, and gas leak concentration sequences, respectively. They also acquire masonry surface images for the surface image monitoring surface. Together, these modules form the operating condition dataset and send it to the diagnostic positioning module.

[0009] The diagnostic positioning module receives the operating condition dataset, performs anomaly judgment analysis on temperature and pressure measuring points based on temperature and pressure sequences, identifies and locates abnormal defect features on the surface monitoring surface based on masonry surface images, performs anomaly judgment analysis on gas leak measuring points based on gas leak concentration sequences, performs cross-validation based on the anomaly analysis results to generate corresponding serial leak early warning signals, and sends them to the graded early warning module for graded early warning prompts.

[0010] Furthermore, the process of constructing the operating condition dataset includes:

[0011] The infrared temperature of the temperature measuring point is collected by the infrared temperature field monitoring module to obtain the temperature sequence for each temperature measuring point. The pressure of the pressure measuring point is collected by the pressure field monitoring module to obtain the pressure sequence for each pressure measuring point. The gas (CO) concentration is continuously monitored by the gas leakage monitoring module to obtain the gas leakage concentration sequence for each gas leakage measuring point. The appearance image monitoring surface is periodically imaged by the image acquisition module to obtain the appearance image of the masonry.

[0012] The temperature sequence, pressure sequence, gas leakage concentration sequence, and masonry appearance image of the corresponding multi-point distributed monitoring network together constitute the operating condition dataset.

[0013] Furthermore, the process of anomaly detection and analysis of temperature and pressure measuring points based on temperature and pressure sequences includes:

[0014] The diagnostic positioning module sets a standard temperature range and a temperature gradient baseline for each temperature measuring point. It calculates the difference between the real-time infrared temperature in the temperature sequence and the upper limit of the set standard temperature range to obtain a temperature gradient sequence for each temperature measuring point. Based on the temperature gradient sequence, it plots a temperature gradient curve and plots the temperature gradient baseline on the same coordinate system as the temperature gradient curve. When the temperature gradient curve intersects with the temperature gradient baseline and the temperature gradient curve continues to rise after the intersection point, the temperature measuring point is determined to be a temperature anomaly calibration point, and an abnormal temperature rise signal is generated.

[0015] Obtain the standard atmospheric pressure value, calculate the pressure fluctuation value by comparing the real-time pressure in the pressure sequence with the standard atmospheric pressure value, obtain the pressure fluctuation sequence for each pressure measuring point based on the pressure fluctuation value, compare the absolute value of the pressure fluctuation value in the pressure fluctuation sequence with the pressure fluctuation threshold, and if there is a preset number of pressure fluctuation values ​​in the pressure fluctuation sequence whose absolute value is greater than the pressure fluctuation threshold, then the pressure measuring point is determined to be a pressure anomaly calibration point, and an abnormal pressure fluctuation signal is generated.

[0016] Furthermore, the process of anomaly detection, identification, and localization of the monitored surface based on the masonry surface image includes:

[0017] A convolutional neural network is used to preprocess the masonry appearance image. A pre-trained and optimized masonry appearance defect recognition model is used to perform feature recognition on the preprocessed masonry appearance image to determine whether there are defects in the masonry appearance image. If there are defects, abnormal defect features are identified, appearance abnormal signals are generated, and edge localization algorithm is used to determine the location of the identified abnormal defect features and mark them as abnormal defect calibration points.

[0018] Furthermore, the process of anomaly detection and analysis of gas leak monitoring points based on gas leak concentration sequences includes:

[0019] The diagnostic location module receives the gas leak concentration sequence, calculates the fluctuation of flue gas concentration before and after the gas leak concentration sequence, obtains the gas leak fluctuation sequence for each gas leak point, compares the gas (CO) concentration in the gas leak concentration sequence with the preset gas (CO) upper limit concentration threshold, and generates a gas abnormal leak signal when the gas (CO) concentration in the gas leak concentration sequence is greater than the preset gas (CO) upper limit concentration threshold or the gas leak fluctuation sequence is a data rising sequence. Then, the gas leak measurement point is determined to be a gas abnormal leak calibration point.

[0020] Furthermore, the diagnostic positioning module counts the number of temperature anomaly calibration points, pressure anomaly calibration points, abnormal defect calibration points, and abnormal gas leak calibration points. Based on the signal reception and the statistics of the number of calibration points, cross-validation is performed to generate corresponding serial leak warning signals.

[0021] Furthermore, the process of generating a cross-validation warning signal based on the anomaly analysis results includes:

[0022] When only one of the following signals is generated: abnormal temperature rise signal, abnormal pressure fluctuation signal, or apparent abnormal signal, and the number of temperature abnormality calibration points, the number of pressure abnormality calibration points, and the number of abnormal defect calibration points are all within the preset safe calibration point number, a potential leakage warning signal is generated.

[0023] When one, two, or three of the following signals are generated: abnormal temperature rise signal, abnormal pressure fluctuation signal, or apparent abnormal signal, and the number of temperature abnormality calibration points, pressure abnormality calibration points, or abnormal defect calibration points exceeds the preset safety calibration point number, a low confidence leakage warning signal is generated.

[0024] When a potential leakage warning signal or a low-confidence leakage warning signal is generated, a gas abnormal leakage signal is generated simultaneously. If the gas abnormal leakage calibration point is within the preset number of safety calibration points, a medium-confidence leakage warning signal is generated.

[0025] When a potential leakage warning signal or a low-confidence leakage warning signal is generated, a gas abnormal leakage signal is generated simultaneously, and when the number of gas abnormal leakage calibration points exceeds the preset number of safety calibration points, a high-confidence leakage warning signal is generated.

[0026] Compared with the prior art, the advantages of this invention are:

[0027] 1. Construct a multi-point distributed monitoring network for key parts of the coke oven, deeply integrating temperature monitoring, pressure monitoring, and masonry surface monitoring as an auxiliary early warning for gas leak detection. Based on cross-validation analysis of multi-source data fusion, achieve early warning and accurate location of leaks, thereby solving the problem of the lag of single gas monitoring and realizing an operation and maintenance management system for preventing coke oven leaks with "predictive proactive maintenance" as its core.

[0028] 2. Based on the above, this solution receives all abnormal signals and data from four units: gas, temperature, pressure, and vision. Based on the temperature monitoring, pressure monitoring, and masonry appearance monitoring network, it achieves comprehensive multi-point and multi-feature monitoring. Before gas leakage, potential leakage risks can be detected through temperature, pressure, or appearance anomalies, realizing a fundamental shift from "post-event detection" to "pre-event warning". Attached Figure Description

[0029] Figure 1 This is a system principle block diagram of the present invention;

[0030] Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0031] 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 some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0032] Example 1: This invention discloses an operation and maintenance management system suitable for preventing coke oven leakage. Please refer to [link / reference]. Figures 1-2 It includes an operation and maintenance management platform, a working condition data acquisition module, a data processing and analysis module, a diagnosis and positioning module, and a graded early warning module;

[0033] The operation and maintenance management platform is used to mark temperature measuring points, pressure measuring points, gas leakage measuring points, and surface image monitoring surfaces for key parts of the coke oven (furnace door, bridge pipe, masonry, etc.), forming a multi-point distributed monitoring network.

[0034] The operating condition data acquisition module includes an infrared temperature field monitoring module, a pressure field monitoring module, a leak gas monitoring module, and an image acquisition module. These modules acquire operating condition data for temperature measuring points, pressure measuring points, and gas leak measuring points, respectively. Based on the operating condition data, they generate temperature sequences, pressure sequences, and gas leak concentration sequences, respectively. They also acquire masonry surface images for the surface image monitoring surface. Together, these modules form the operating condition dataset and send it to the diagnostic positioning module.

[0035] The process of constructing the operating condition dataset includes:

[0036] Temperature measuring points are marked on key parts such as furnace door, furnace frame, riser pipe, and furnace body masonry surface. Thermocouples or infrared thermal imagers are arranged on multiple temperature measuring points to form a temperature monitoring network, realizing all-weather, full-coverage, non-contact monitoring of the temperature field on the furnace body surface, capturing local hot spots caused by leakage, and collecting infrared temperatures of temperature measuring points through infrared temperature field monitoring modules to obtain temperature sequences for each temperature measuring point.

[0037] Pressure measuring points are marked on key parts such as gas collecting pipe, carbonization chamber, combustion chamber, heat storage chamber, and flue gas distribution duct. Pressure sensors are arranged on multiple pressure measuring points to form a pressure monitoring network to monitor the stability of the coke oven pressure system in real time. Since pressure fluctuation is one of the root causes of inducing and aggravating leakage, the pressure at the pressure measuring points is collected by the pressure field monitoring module to obtain the pressure sequence for each pressure measuring point.

[0038] Gas leakage detection points are marked on key parts such as furnace top, furnace end, and furnace door. Multiple online gas sensors are deployed at the gas leakage detection points to form a gas leakage monitoring network. The gas (CO) concentration is continuously monitored through the leakage gas monitoring module to obtain the gas leakage concentration sequence of each gas leakage detection point.

[0039] The surface of the masonry inside the furnace is marked as the surface image monitoring surface. High-definition cameras or industrial endoscopes are placed at the surface image monitoring surface to form an image monitoring network. The surface image monitoring surface is periodically captured by the image acquisition module to obtain the surface image of the masonry.

[0040] Temperature monitoring network, pressure monitoring network, gas leak monitoring network, and image monitoring network together constitute a multi-point distributed monitoring network. The temperature sequence, pressure sequence, gas leak concentration sequence, and masonry appearance image of the corresponding multi-point distributed monitoring network together constitute the operating condition dataset. Based on the existing gas monitoring as one of the core components, temperature, pressure, and appearance monitoring are added to form a multi-dimensional monitoring network. Among them, temperature monitoring serves as an early warning indicator because the early stage of leakage is often accompanied by local temperature rise. Pressure monitoring serves as an auxiliary judgment indicator, as abnormal pressure can indicate furnace sealing problems. Masonry appearance monitoring serves as direct evidence of structural anomalies and can detect masonry damage that may lead to leakage in advance.

[0041] The diagnostic positioning module is used to receive the operating condition dataset, perform anomaly judgment analysis on temperature measuring points and pressure measuring points based on temperature and pressure sequences, perform anomaly defect feature judgment and positioning on the surface monitoring surface based on masonry surface images, and perform anomaly judgment analysis on gas leak measuring points based on gas leak concentration sequences.

[0042] The process of anomaly detection and analysis of temperature and pressure measuring points based on temperature and pressure sequences includes:

[0043] The diagnostic positioning module sets a standard temperature range and a temperature gradient baseline for each temperature measurement point. It calculates the difference between the real-time infrared temperature in the temperature sequence and the upper limit of the set standard temperature range to obtain the temperature gradient sequence for each temperature measurement point. Based on the temperature gradient sequence, it plots the temperature gradient curve.

[0044] The temperature gradient baseline is plotted on the same coordinate system as the temperature gradient curve. When the temperature gradient curve intersects with the temperature gradient baseline, and the temperature gradient curve continues to rise after the intersection point, the temperature measuring point is determined to be a temperature anomaly calibration point, and an abnormal temperature rise signal is generated.

[0045] Obtain the standard atmospheric pressure value, calculate the pressure fluctuation value by comparing the real-time pressure in the pressure sequence with the standard atmospheric pressure value, obtain the pressure fluctuation sequence for each pressure measuring point based on the pressure fluctuation value, compare the absolute value of the pressure fluctuation value in the pressure fluctuation sequence with the pressure fluctuation threshold, and if there is a preset number of pressure fluctuation values ​​in the pressure fluctuation sequence whose absolute value is greater than the pressure fluctuation threshold, then the pressure measuring point is determined to be a pressure anomaly calibration point and a pressure anomaly fluctuation signal is generated.

[0046] The process of anomaly detection, identification, and localization of monitored surfaces based on masonry surface images includes:

[0047] A convolutional neural network is used to preprocess the masonry appearance image (preprocessing includes noise reduction, standardization and normalization operations to enhance the contrast and clarity of the masonry appearance image, remove irrelevant information in the masonry appearance image and highlight the masonry appearance features). A pre-trained and optimized masonry appearance defect recognition model is used to perform feature recognition on the preprocessed masonry appearance image to determine whether there are defects in the masonry appearance image. If there are, abnormal defect features are identified, appearance anomaly signals are generated, and edge localization algorithm is used to determine the location of the identified abnormal defect features and mark them as abnormal defect calibration points.

[0048] The process of anomaly detection and analysis of gas leak monitoring points based on gas leak concentration sequences includes:

[0049] The diagnostic positioning module receives the gas leak concentration sequence, calculates the fluctuation of flue gas concentration before and after the gas leak concentration sequence, obtains the gas leak fluctuation sequence for each gas leak point, compares the gas (CO) concentration in the gas leak concentration sequence with the preset gas (CO) upper limit concentration threshold, and generates a gas abnormal leak signal when the gas (CO) concentration in the gas leak concentration sequence is greater than the preset gas (CO) upper limit concentration threshold or the gas leak fluctuation sequence is a data rising sequence. Then, the gas leak measurement point is determined to be a gas abnormal leak calibration point.

[0050] The diagnostic positioning module counts the number of temperature anomaly calibration points, pressure anomaly calibration points, abnormal defect calibration points, and abnormal gas leak calibration points. Based on the signal reception and the statistics of the number of calibration points, it performs cross-validation to generate corresponding serial leak early warning signals.

[0051] Specifically, when only one of the following signals is generated: abnormal temperature rise signal, abnormal pressure fluctuation signal, or apparent abnormal signal, and the number of temperature abnormality calibration points, pressure abnormality calibration points, and abnormal defect calibration points are all within the preset safe calibration point number, a potential leakage warning signal is generated.

[0052] When one, two, or three of the following signals are generated: abnormal temperature rise signal, abnormal pressure fluctuation signal, or apparent abnormal signal, and the number of temperature abnormality calibration points, pressure abnormality calibration points, or abnormal defect calibration points exceeds the preset safety calibration point number, a low confidence leakage warning signal is generated.

[0053] When a potential leakage warning signal or a low-confidence leakage warning signal is generated, a gas abnormal leakage signal is generated simultaneously. If the number of gas abnormal leakage calibration points is within the preset number of safety calibration points, a medium-confidence leakage warning signal is generated.

[0054] When a potential leakage warning signal or a low-confidence leakage warning signal is generated, a gas abnormal leakage signal is generated simultaneously, and when the number of gas abnormal leakage calibration points exceeds the preset number of safety calibration points, a high-confidence leakage warning signal is generated.

[0055] Potential crosstalk leakage warning signals, low-confidence crosstalk leakage warning signals, medium-confidence crosstalk leakage warning signals, and high-confidence crosstalk leakage warning signals are sent to the graded warning module.

[0056] After receiving leakage warning signals at various levels, the tiered early warning module provides tiered warning prompts. When a potential leakage warning signal is received, the abnormal calibration point is identified and included in the preventive maintenance plan. When a low-confidence leakage warning signal is received, indicating that the gas concentration has not exceeded the standard, the system judges it as an "early internal leak or minor leakage," identifies the abnormal calibration point, strengthens monitoring of the abnormal calibration point, or arranges for recent inspection. When a medium-confidence leakage warning signal or a high-confidence leakage warning signal is received, the abnormal calibration point is identified, especially the gas leak calibration point, which is clearly indicated as the location of the flammable gas leak. Through multi-level early warning, direct execution operation and maintenance instructions are given.

[0057] In summary, this approach involves constructing a multi-point distributed monitoring network for key components of the coke oven, deeply integrating temperature monitoring, pressure monitoring, and masonry surface monitoring. This network serves as a preliminary early warning system to assist in gas leak detection. Based on cross-validation analysis using multi-source data fusion, it achieves early warning and precise location of leaks, addressing the lag issue of single-gas monitoring. This constructs a comprehensive governance system with "distributed network monitoring and early warning" as the guide, "predictive proactive maintenance" as the core, and "multi-root cause anomaly judgment management" as the guarantee, effectively improving the prevention and management of coke oven leaks.

[0058] The above description is merely a preferred embodiment of the present invention; however, the scope of protection of the present invention is not limited thereto; any equivalent substitutions or modifications made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. An operation and maintenance management system suitable for preventing coke oven leakage, characterized in that: It includes an operation and maintenance management platform, a working condition data acquisition module, a diagnostic and location module, and a graded early warning module; The operation and maintenance management platform is used to mark temperature measuring points, pressure measuring points, gas leakage measuring points, and surface image monitoring surfaces for key parts of the coke oven, and to build a multi-point distributed monitoring network. The operating condition data acquisition module includes an infrared temperature field monitoring module, a pressure field monitoring module, a leak gas monitoring module, and an image acquisition module. These modules acquire operating condition data for temperature measuring points, pressure measuring points, and gas leak measuring points, respectively. Based on the operating condition data, they generate temperature sequences, pressure sequences, and gas leak concentration sequences, respectively. They also acquire masonry surface images for the surface image monitoring surface. Together, these modules form the operating condition dataset and send it to the diagnostic positioning module. The diagnostic positioning module receives the operating condition dataset, performs anomaly judgment analysis on temperature and pressure measuring points based on temperature and pressure sequences, identifies and locates abnormal defect features on the surface monitoring surface based on masonry surface images, performs anomaly judgment analysis on gas leak measuring points based on gas leak concentration sequences, performs cross-validation based on the anomaly analysis results to generate corresponding serial leak early warning signals, and sends them to the graded early warning module for graded early warning prompts.

2. The operation and maintenance management system for preventing coke oven leakage according to claim 1, characterized in that: The process of constructing the operational condition dataset includes: The infrared temperature of the temperature measuring point is collected by the infrared temperature field monitoring module to obtain the temperature sequence for each temperature measuring point. The pressure of the pressure measuring point is collected by the pressure field monitoring module to obtain the pressure sequence for each pressure measuring point. The gas (CO) concentration is continuously monitored by the gas leakage monitoring module to obtain the gas leakage concentration sequence for each gas leakage measuring point. The appearance image monitoring surface is periodically imaged by the image acquisition module to obtain the appearance image of the masonry. The temperature sequence, pressure sequence, gas leakage concentration sequence, and masonry appearance image of the corresponding multi-point distributed monitoring network together constitute the operating condition dataset.

3. The operation and maintenance management system for preventing coke oven leakage according to claim 2, characterized in that: The process of anomaly detection and analysis of temperature and pressure measuring points based on temperature and pressure sequences includes: The diagnostic positioning module sets a standard temperature range and a temperature gradient baseline for each temperature measuring point. It calculates the difference between the real-time infrared temperature in the temperature sequence and the upper limit of the set standard temperature range to obtain a temperature gradient sequence for each temperature measuring point. Based on the temperature gradient sequence, it plots a temperature gradient curve and plots the temperature gradient baseline on the same coordinate system as the temperature gradient curve. When the temperature gradient curve intersects with the temperature gradient baseline and the temperature gradient curve continues to rise after the intersection point, the temperature measuring point is determined to be a temperature anomaly calibration point, and an abnormal temperature rise signal is generated. Obtain the standard atmospheric pressure value, calculate the pressure fluctuation value by comparing the real-time pressure in the pressure sequence with the standard atmospheric pressure value, obtain the pressure fluctuation sequence for each pressure measuring point based on the pressure fluctuation value, compare the absolute value of the pressure fluctuation value in the pressure fluctuation sequence with the pressure fluctuation threshold, and if there is a preset number of pressure fluctuation values ​​in the pressure fluctuation sequence whose absolute value is greater than the pressure fluctuation threshold, then the pressure measuring point is determined to be a pressure anomaly calibration point, and an abnormal pressure fluctuation signal is generated.

4. The operation and maintenance management system for preventing coke oven leakage according to claim 3, characterized in that: The process of anomaly detection, identification, and localization of monitored surfaces based on masonry surface images includes: A convolutional neural network is used to preprocess the masonry appearance image. A pre-trained and optimized masonry appearance defect recognition model is used to perform feature recognition on the preprocessed masonry appearance image to determine whether there are defects in the masonry appearance image. If there are defects, abnormal defect features are identified, appearance abnormal signals are generated, and edge localization algorithm is used to determine the location of the identified abnormal defect features and mark them as abnormal defect calibration points.

5. The operation and maintenance management system for preventing coke oven leakage according to claim 4, characterized in that: The process of anomaly detection and analysis of gas leak monitoring points based on gas leak concentration sequences includes: The diagnostic location module receives the gas leak concentration sequence, calculates the fluctuation of flue gas concentration before and after the gas leak concentration sequence, obtains the gas leak fluctuation sequence for each gas leak point, compares the gas (CO) concentration in the gas leak concentration sequence with the preset gas (CO) upper limit concentration threshold, and generates a gas abnormal leak signal when the gas (CO) concentration in the gas leak concentration sequence is greater than the preset gas (CO) upper limit concentration threshold or the gas leak fluctuation sequence is a data rising sequence. Then, the gas leak measurement point is determined to be a gas abnormal leak calibration point.

6. The operation and maintenance management system for preventing coke oven leakage according to claim 5, characterized in that: The diagnostic positioning module counts the number of temperature anomaly calibration points, pressure anomaly calibration points, abnormal defect calibration points, and abnormal gas leak calibration points. Based on the signal reception and the statistics of the number of calibration points, it performs cross-validation to generate corresponding serial leak warning signals.

7. The operation and maintenance management system for preventing coke oven leakage according to claim 6, characterized in that: The process of generating a cross-validation warning signal based on anomaly analysis results includes: When only one of the following signals is generated: abnormal temperature rise signal, abnormal pressure fluctuation signal, or apparent abnormal signal, and the number of temperature abnormality calibration points, the number of pressure abnormality calibration points, and the number of abnormal defect calibration points are all within the preset safe calibration point number, a potential leakage warning signal is generated. When one, two, or three of the following signals are generated: abnormal temperature rise signal, abnormal pressure fluctuation signal, or apparent abnormal signal, and the number of temperature abnormality calibration points, pressure abnormality calibration points, or abnormal defect calibration points exceeds the preset safety calibration point number, a low confidence leakage warning signal is generated. When a potential leakage warning signal or a low-confidence leakage warning signal is generated, a gas abnormal leakage signal is generated simultaneously. If the number of gas abnormal leakage calibration points is within the preset number of safety calibration points, a medium-confidence leakage warning signal is generated. When a potential leakage warning signal or a low-confidence leakage warning signal is generated, a gas abnormal leakage signal is generated simultaneously. When the number of gas abnormal leakage calibration points exceeds the preset number of safety calibration points, a high-confidence leakage warning signal is generated.

Images