Detection device for pressure relief quick connector

Abstract

The invention relates to a detection device for a pressure relief quick connector. The detection device comprises a data acquisition module, a detection execution module and an intelligent control and analysis module. The data acquisition module comprises a pressure sensor, a flow sensor, a displacement sensor and an industrial camera and is used for acquiring internal pressure, pressure relief rate, component displacement and appearance image data of the joint to be detected; the detection execution module comprises a pressure loading system, an environment simulation cabin and a pressure relief control mechanism and is used for simulating working pressure, an extreme environment and triggering pressure relief operation; and the intelligent control and analysis module comprises a detection controller and a data processing unit, and the detection controller is used for generating a detection scheme according to the basic information of the to-be-detected joint and controlling the data acquisition module and the detection execution module to detect the to-be-detected joint according to a detection process. And the data processing unit is used for analyzing the detection data and generating an evaluation report. The method has the effect of improving the joint detection efficiency and the detection quality.

Description

Technical Field

[0001] The present application relates to the field of joint detection technology, and in particular to a detection device for a pressure relief quick joint. Background Art

[0002] A pressure-relief quick connector is a device used to connect fluid or gas pipelines and quickly relieve pressure when needed. It consists of a connector body, a valve seat and valve core, a locking pressure-relief assembly, and a control sleeve. It effectively releases pressure within the pipeline, preventing fluid splashing, component damage, or personal injury caused by sudden pressure release during connector disassembly, significantly improving operational safety.

[0003] As a key component of high-pressure fluid systems, the performance of pressure relief quick connectors directly affects equipment safety. Testing of pressure relief quick connectors requires multiple dimensions, including sealing, pressure relief function, pressure resistance, and material reliability. Traditional testing methods have the following drawbacks:

[0004] 1. Low testing efficiency: manual step-by-step testing (such as pressure testing followed by sealing inspection) is used, and a single test takes more than 30 minutes, which cannot meet the needs of mass production;

[0005] 2. Insufficient accuracy: It is difficult to detect micro leaks (leakage rate > 1×10 - 8 Pa·m 3 / s) and internal structural defects (such as valve seat microcracks);

[0006] 3. Single working condition simulation: It can only test the performance under normal temperature and pressure, and cannot reproduce the failure mode under complex environments such as high temperature and vibration;

[0007] 4. Weak data application: Detection data is not linked to the production process, making it difficult to trace defects and optimize processes.

[0008] Although existing technologies have some automated testing equipment, they lack multi-technical integration and intelligent decision-making capabilities. Therefore, there is an urgent need for a high-precision, high-efficiency testing system with process optimization capabilities. Summary of the Invention

[0009] In order to solve the above problems, the present application provides a detection device for a pressure relief quick connector.

[0010] In a first aspect, the present application provides a detection device for a pressure relief quick connector, which adopts the following technical solution:

[0011] A detection device for a pressure relief quick connector, comprising:

[0012] The data acquisition module includes a pressure sensor, a flow sensor, a displacement sensor, and an industrial camera, which is used to collect internal pressure, pressure relief rate, component displacement, and appearance image data of the joint to be tested;

[0013] A test execution module, including a pressure loading system, an environmental simulation chamber, and a pressure relief control mechanism, used to simulate working pressure, extreme environments, and trigger pressure relief operations; and

[0014] The intelligent control and analysis module includes a detection controller and a data processing unit. The detection controller is used to generate a detection plan based on the basic information of the joint to be detected, and control the data acquisition module and the detection execution module to detect the joint to be detected according to the detection process; the detection plan includes at least one detection item, detection auxiliary equipment, detection process information, equipment configuration parameters and environmental simulation conditions; the data processing unit is used to analyze the detection data and generate an evaluation report.

[0015] Preferably, the data acquisition module further includes:

[0016] Industrial CT scanning unit, used to perform three-dimensional tomography scanning of the internal structure of the joint and construct dynamic displacement images;

[0017] Laser Doppler vibration measuring instrument is used to measure the vibration spectrum of the joint surface in real time and identify abnormal vibration modes.

[0018] Preferably, the detection controller of the intelligent control and analysis module includes:

[0019] The solution matching model is a machine learning model that is iteratively trained based on a greedy algorithm and historical error data. It is used to generate a personalized detection solution based on the basic information of the joint to be detected; the basic information of the joint to be detected includes the joint model, material, design parameters, application scenario, historical working conditions, and detection standards;

[0020] The AI visual inspection model is based on the YOLOv8 model or ChatGPT-4.0 model and is iteratively trained using historical appearance inspection data of the joint to be inspected. It is used to identify surface cracks, wear, and seal deformation of the joint.

[0021] Preferably, the environmental simulation chamber comprises:

[0022] A multi-field coupled loading device is used to simultaneously apply -40°C to 120°C temperature, 0-50MPa pressure, and 10-2000Hz vibration to create the environmental simulation conditions required for the testing solution;

[0023] The multi-axis mounting table is used to dynamically adjust the joint installation angle and simulate the horizontal, vertical or different tilt angle installation postures of the joint to be tested.

[0024] Preferably, the detection execution module further includes:

[0025] The plugging and unplugging drive unit includes a multi-axis manipulator for performing the plugging and unplugging operation of the connector and a strain gauge force sensor provided at the output end of the multi-axis manipulator for real-time monitoring of plugging and unplugging force changes;

[0026] The multi-media fast switching system includes a solenoid valve group and multiple media output sources, which are used to switch hydraulic oil, compressed air and corrosive media according to the instructions of the detection controller.

[0027] Preferably, the data processing unit includes:

[0028] The digital twin module is used to generate a three-dimensional digital twin model of the joint, dynamically displaying pressure distribution, deformation areas, and leakage risk points according to pre-set color coding rules based on the test data;

[0029] The predictive analysis module includes a defect prediction model, which is used to predict and generate defect risks during the joint inspection process based on the obtained inspection data and the historical working conditions of the joint to be inspected, and dynamically adjust the inspection item combination of the inspection plan, the execution order of items in the inspection process, and the parameter configuration; the defect prediction model is an LSTM neural network model obtained by iterative training through historical inspection data.

[0030] Preferably, the detection controller generates a detection plan based on the basic information of the joint to be detected, and controls the data acquisition module and the detection execution module to detect the joint to be detected, specifically comprising the following steps:

[0031] The industrial camera recognizes the pre-set QR code on the joint to be inspected, obtains basic information about the joint to be inspected, and generates a personalized inspection plan through the plan matching model.

[0032] Use the AI visual inspection model to perform preliminary screening for appearance defects on the joints to be inspected and eliminate unqualified products;

[0033] Dispatch detection auxiliary equipment according to the detection plan and send equipment configuration parameters for parameter configuration;

[0034] Generate control instructions according to the detection plan and detection process, control the detection auxiliary equipment, data acquisition module and detection execution module to detect the joint to be detected according to the detection process.

[0035] Preferably, the data processing unit analyzes the detection data and generates an evaluation report specifically including the following steps:

[0036] The digital twin module is called to map the inspection data to the 3D digital twin model of the joint. The 3D visualization results of the pressure distribution cloud map, internal structural deformation area, leakage risk points and joint characteristic data are dynamically displayed according to the pre-set color coding rules.

[0037] Use the defect prediction model of the predictive analysis module to perform time series analysis on real-time inspection data and historical working condition data to predict the potential defect risk of the joint and dynamically adjust the inspection item combination of the inspection plan, the execution order of items in the inspection process, and the parameter configuration;

[0038] Integrate test data, digital twin visualization results, defect prediction reports, and adjusted test plans to generate a multi-dimensional evaluation report that includes pressure-time curves, CT slice image annotations, and environmental test performance fluctuation ranges.

[0039] Preferably, the digital twin module is also used to superimpose the performance change data of the joint under different environmental simulation conditions onto the three-dimensional digital twin model to generate a dynamic performance evolution video.

[0040] Preferably, the predictive analysis module also includes a defect-cause association reasoning unit, which is used to establish a defect-cause association knowledge base, match the causes of predicted defect risks, and generate possible causes and solution suggestions corresponding to the defect risks in the assessment report.

[0041] In summary, this application includes at least one of the following beneficial technical effects:

[0042] 1. This application can accurately and efficiently match personalized detection plans for different joints based on the actual conditions and historical working conditions of the joints, uniformly dispatch data collection and detection execution modules, and achieve an upgrade from "single indicator detection" to "personalized full-scenario performance verification" according to the detection plan. This not only ensures the reliability of the test results, but also improves the detection efficiency and economy by automating the detection process and dynamically optimizing the detection plan based on real-time detection data and historical working conditions, effectively improving the efficiency and quality of joint detection.

[0043] 2. Through the setting of the predictive analysis module, based on the LSTM neural network, the time series data characteristics are mined. By analyzing the historical and real-time inspection data of joints of the same model and process, high-frequency defect patterns are identified and potential risk points of the current batch to be inspected are predicted. Based on the predicted defect risks, the project combination, execution sequence and parameter configuration of the inspection plan are dynamically adjusted to achieve the inspection goals of "low cost, high efficiency and full coverage", which helps to improve the efficiency and accuracy of joint inspection.

[0044] 3. Integrate multi-source data and intelligent analysis results, and through the collaboration of functional modules such as digital twin visualization, defect risk prediction, and detection plan adjustment, achieve efficient data utilization and dynamic optimization of detection plans, and ultimately form a three-dimensional and structured detection conclusion; the report content covers key information such as graphical data, CT defect annotation, and performance fluctuation range, which not only meets the technical personnel's need for detailed tracing, but also provides intuitive decision-making basis for management, realizing multi-dimensional output of detection value. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a system block diagram of a detection device for a pressure relief quick connector according to an embodiment of the present application;

[0046] Figure 2 This is a flow chart of a method for matching and generating a personalized detection plan in an embodiment of the present application;

[0047] Figure 3 This is a flow chart of the method for analyzing test data and generating an evaluation report in an embodiment of the present application.

[0048] Explanation of the accompanying drawings: 1. Data acquisition module; 11. Pressure sensor; 12. Flow sensor; 13. Displacement sensor; 14. Industrial camera; 15. Industrial CT scanning unit; 16. Laser Doppler vibrometer; 2. Detection execution module; 21. Pressure loading system; 22. Environmental simulation chamber; 221. Multi-field coupling loading device; 222. Multi-axis mounting platform; 23. Pressure relief control mechanism; 24. Plug-in drive unit; 25. Multi-media rapid switching system; 3. Intelligent control and analysis module; 31. Detection controller; 311. Scheme matching model; 312. AI visual detection model; 32. Data processing unit; 321. Digital twin module; 322. Predictive analysis module; 323. Defect-cause association reasoning unit. DETAILED DESCRIPTION

[0049] The following is combined with Figure 1-3 This application is described in further detail.

[0050] The embodiment of the present application discloses a detection device for a pressure relief quick connector. Figure 1, a detection device for a pressure relief quick connector, comprising a data acquisition module 1, a detection execution module 2 and an intelligent control and analysis module 3. The data acquisition module 1, the detection execution module 2 and the intelligent control and analysis module 3 are connected to each other in communication. When facing different connectors, the present application can accurately and efficiently match personalized detection schemes according to the actual conditions and historical working conditions of the connectors, uniformly dispatch the data acquisition and detection execution module 2, and realize the upgrade from "single indicator detection" to "personalized full-scene performance verification" according to the detection scheme, which not only ensures the reliability of the detection results, but also improves the detection efficiency and economy through the automation of the detection process and the dynamic optimization of the detection scheme according to real-time detection data and historical working conditions, thereby effectively improving the detection efficiency and quality of the connectors.

[0051] Reference Figure 1 The data acquisition module 1 includes a pressure sensor 11 for collecting the internal pressure of the joint to be tested, a flow sensor 12 for collecting the pressure relief rate of the joint to be tested, a displacement sensor 13 for collecting the displacement of the joint assembly to be tested, an industrial camera 14 for collecting image data of the joint's appearance, an industrial CT scanning unit 15 for performing three-dimensional tomography scanning of the joint's internal structure to construct a dynamic displacement image, and a laser Doppler vibrometer 16 for real-time measurement of the joint's surface vibration spectrum and identification of abnormal vibration modes. The industrial CT scanning unit 15 uses micro-focus industrial CT to perform three-dimensional tomography scanning of the joint's internal valve seat, valve core, and other precision structures, constructing a dynamic displacement image to detect hidden defects such as internal cracks and abnormal assembly clearances. It also performs simultaneous dynamic scanning to observe changes in internal component displacement during pressure testing. The laser Doppler vibrometer 16 uses a laser beam to measure the joint's surface vibration spectrum in real time, identifying abnormal vibration modes caused by seal wear or structural looseness. Compared to traditional vibration testing that only measures amplitude and frequency, it can obtain more comprehensive vibration characteristic information. The data acquisition module 1 integrates multiple types of sensors and testing equipment to comprehensively collect multi-dimensional data, including pressure, flow, displacement, appearance, internal structure, and vibration. The application of an industrial CT scanning unit 15 and a laser Doppler vibrometer 16 overcomes the bottleneck of traditional testing in identifying internal defects and complex vibration modes. The former enables nondestructive detection and dynamic monitoring of internal defects, while the latter obtains in-depth vibration information, significantly improving the ability to detect hidden defects and the accuracy of vibration fault diagnosis.

[0052] Reference Figure 1The test execution module 2 includes a pressure loading system 21, an environmental simulation chamber 22, and a pressure relief control mechanism 23, which are used to simulate the working pressure, extreme environment, and trigger the pressure relief operation when the joint is in use. Among them, the pressure loading system 21 and the pressure relief control mechanism 23 are existing equipment and are not described in detail here. The environmental simulation chamber 22 includes a multi-field coupling loading device 221 and a multi-axis mounting platform 222. The multi-field coupling loading device 221 integrates temperature control equipment, pressure loading equipment, and vibration generating equipment. It can simultaneously apply 0-50MPa pressure and 10-2000Hz vibration in a temperature range of -40℃ to 120℃ to simulate extreme working conditions in aerospace and other fields. The multi-axis mounting platform 222 includes an electrically controlled multi-axis pan-tilt stage and fixing parts such as clamps for fixing joints. It is used to dynamically adjust the joint installation angle and simulate the horizontal, vertical, or different tilt angle installation postures of the joint to be tested to show the performance changes of the joint to be tested in different installation postures such as horizontal and vertical, such as the difference in pressure relief efficiency. Through the collaboration of multiple components, the detection execution module 2 can highly reproduce the actual working scenarios and extreme working conditions of the joint, improving the practicality and reliability of the test results. The multi-field coupling loading device 221 realizes the simultaneous loading of multiple physical fields such as temperature, pressure, and vibration, accurately replicating the complex stress environments in harsh scenarios such as aerospace, and providing a simulated environment for various testing projects. The multi-axis mounting platform 222, based on the working principle of the joint, can demonstrate the performance changes of the joint under test in different installation postures such as horizontal and vertical, helping to improve detection accuracy.

[0053] Reference Figure 1The intelligent control and analysis module 3 includes a detection controller 31 and a data processing unit 32. The detection controller 31 is used to generate a detection plan based on the basic information of the joint to be detected, and control the data acquisition module 1 and the detection execution module 2 to detect the joint to be detected according to the detection process. The detection controller 31 includes a solution matching model 311 and an AI visual detection model 312. Among them, the solution matching model 311 is a machine learning model obtained by iterative training based on a greedy algorithm and historical error data. The specific training steps are existing technologies and are not described in detail. The solution matching model 311 is used to generate a personalized detection plan based on the basic information of the joint to be detected. The AI visual detection model 312 is obtained by iterative training based on the historical appearance detection data of the joint to be detected using the YOLOv8 model or the ChatGPT-4.0 model. The specific training steps are existing technologies and are not described in detail. The AI visual detection model 312 is used to identify cracks, wear and seal deformation on the joint surface. In this embodiment, the ChatGPT-4.0 model is preferably used for the initial mixed inspection of multiple types of joints. Relying on its powerful image recognition capabilities, it can achieve high-efficiency mixed inspection of multiple types of joints without sorting. When faced with already sorted joints, the YOLOv8 model is preferred to achieve efficient and accurate detection. Through the setting of the solution matching model 311, a personalized detection solution is generated based on the basic information of the joint to be detected, and the optimal detection path can be automatically generated based on the joint model, material and other information. When faced with different joints, a personalized detection solution can be accurately and efficiently matched according to the actual situation and historical working conditions of the joint, taking into account the detection timeliness and detection difficulty requirements of the inspectors, so as to effectively improve the efficiency and quality of joint detection. Through the setting of the AI visual inspection model 312, unqualified products can be efficiently screened out in advance through visual inspection before joint detection, reducing the time spent on invalid detection.

[0054] The above-mentioned testing plan includes at least one test item, auxiliary testing equipment, test process information, equipment configuration parameters, and environmental simulation conditions. Basic information about the joint to be tested includes the joint model, material, design parameters, application scenario, historical operating conditions, and testing standards. Auxiliary testing equipment is equipment specifically required for the test item. For example, the auxiliary testing equipment required for a helium mass spectrometer leak test item includes a helium mass spectrometer leak detector for detecting the concentration of leaked helium gas, and an infrared thermal imager for assisting in locating the leak point and visually displaying the leak area through temperature changes.

[0055] Reference Figure 1The data processing unit 32 includes a digital twin module 321 and a predictive analysis module 322. The digital twin module 321 is used to generate a three-dimensional digital twin model of the joint, dynamically displaying the pressure distribution, deformation area, and leakage risk points according to the pre-set color coding rules based on the test data. The color coding rules are pre-set by the management personnel, such as red ≥ 80% rated pressure, yellow 50%-80% rated pressure, and green 0-50% rated pressure. The data processing unit 32 uses digital twin technology to construct a virtual mapping of the joint, realizing three-dimensional visualization of the test data and real-time labeling of risk areas, greatly reducing the difficulty of manual interpretation and the risk of misjudgment. The predictive analysis module 322 has a built-in defect prediction model, which is used to predict and generate defect risks during the joint inspection process based on the obtained test data and the historical working conditions of the joint to be inspected, and dynamically adjust the inspection project combination of the inspection plan, the execution order of the projects in the inspection process, and the parameter configuration. The defect prediction model is an LSTM neural network model obtained by iterative training through historical inspection data. Through the setting of the predictive analysis module 322, the time series data features are mined based on the LSTM neural network. By analyzing the historical inspection data and real-time inspection data of the joints of the same model and the same process, high-frequency defect patterns (such as a high leakage rate of seals in a certain batch and cracks in the valve core heat treatment) are identified, and the potential risk points of the current batch to be inspected are predicted. According to the predicted defect risks, the project combination, execution sequence and parameter configuration of the inspection plan are dynamically adjusted to achieve the inspection goal of "low cost, high efficiency and full coverage", and realize full inspection of high-risk products and random inspection of low-risk products, which helps to improve the efficiency and accuracy of joint inspection.

[0056] For example, when inspecting Type A joints, dense microcracks of 0.1 mm were found. According to historical data, this had occurred before. Type A titanium alloy joints are prone to 0.1 mm microcracks after laser welding. However, because the frequency of occurrence is low and only occurs once, the inspection plan generated by the scheme matching model 311 does not increase the frequency of industrial CT scanning in related inspection items. The defect prediction model outputs microcrack risk prediction based on real-time data and historical data, and dynamically adjusts the inspection plan to increase the frequency of industrial CT scanning and efficiently detect defective products.

[0057] Furthermore, the digital twin module 321 can present personalized inspection items for different joints in different visualization formats, such as areas of abnormal vibration (e.g., flashing icons), flow fluctuation ranges (e.g., dynamic curve overlays), and areas of performance variation under different media. This implementation of a diversified visualization strategy ensures the efficient communication and accurate interpretation of complex inspection data, meeting the differentiated inspection needs of multiple scenarios and types of joints, and improving the efficiency of human-computer interaction.

[0058] The digital twin module 321 also overlays the performance change data of the joint under different environmental simulation conditions onto the three-dimensional digital twin model to generate a dynamic performance evolution video, visually presenting the trend of joint performance changes with environmental factors in a timeline format. First, the dynamic video visualizes the changes in joint performance parameters under multiple coupled environments, such as temperature, pressure, and vibration. Compared to static data or charts, inspectors and engineers can more intuitively and quickly observe the trend of joint performance changes with environmental factors. For example, how the joint sealing performance gradually decreases under high temperature, high pressure, and vibration combined environments, helping users quickly grasp key information. Second, the dynamic performance evolution video can fully demonstrate the entire process from normal performance to abnormal performance. Combined with the digital twin model's display of internal structure and pressure distribution, when a joint performance failure occurs, technicians can review the video to accurately locate the time and specific location of the failure and analyze the relationship between the failure and environmental factors. For example, under a certain temperature and pressure combination, a specific part of the joint begins to deform and leak, thereby more efficiently determining the cause of the failure. Thirdly, by viewing dynamic videos, R&D personnel can comprehensively assess the impact of different environmental conditions on product performance and understand the product's weaknesses under extreme operating conditions. Based on this information, they can optimize product design, improve manufacturing processes, and enhance product quality and reliability. For example, they can adjust the joint's structural design to enhance its stability in complex environments. Finally, the generated dynamic videos are retained as part of the test data and, together with other test data, form a digital archive of the joint's entire lifecycle, allowing for easy subsequent review and traceability, providing comprehensive data support for product quality traceability and performance evaluation.

[0059] Reference Figure 1 The predictive analysis module 322 is additionally provided with a defect-cause association reasoning unit 323, which is used to establish a defect-cause association knowledge base, match the causes of the predicted defect risks, and generate possible causes and solution suggestions corresponding to the defect risks in the assessment report. Among them, the establishment of the defect-cause association knowledge base is based on historical detection data combined with expert verification conclusions to construct a dynamically updated defect-cause association knowledge base. Through the setting of the defect-cause association reasoning unit 323, the defect-cause association reasoning unit 323 constructs a closed-loop logic of "risk prediction-cause analysis-solution suggestion", establishes a probabilistic association model of defects and causes based on historical data, and realizes the intelligent extension from "discovering problems" to "solving problems"; by providing specific cause annotations and maintenance suggestions in the assessment report, it significantly reduces the cost of manual troubleshooting, which helps enterprises to efficiently optimize iterative processes.

[0060] Reference Figure 2The detection controller 31 generates a detection plan based on the basic information of the joint to be detected, and controls the data acquisition module 1 and the detection execution module 2 to detect the joint to be detected. Specifically, the following steps are included:

[0061] A1. Generate a personalized inspection plan: Use the industrial camera 14 to identify the QR code pre-set on the joint to be inspected, obtain basic information about the joint to be inspected, and generate a personalized inspection plan using the plan matching model 311;

[0062] A2. Perform preliminary screening for appearance defects: Use the AI visual inspection model 312 to perform preliminary screening for appearance defects on the joints to be inspected, and reject unqualified products.

[0063] A3. Parameter configuration: Dispatch auxiliary detection equipment according to the detection plan and send device configuration parameters for parameter configuration;

[0064] A4. Inspect the joints to be inspected: Generate control instructions according to the inspection plan and the inspection process, control the inspection auxiliary equipment, data acquisition module 1 and inspection execution module 2 to inspect the joints to be inspected according to the inspection process. Through the above steps, the inspection process is fully automated and controlled. The QR code recognition technology quickly obtains the basic information of the joints, and combines the solution matching model 311 to generate a customized inspection plan. When facing different joints, it can accurately and efficiently match personalized inspection plans based on the actual situation and historical working conditions of the joints, taking into account the inspection timeliness and inspection difficulty requirements of the inspectors, so as to effectively improve the inspection efficiency and inspection quality of the joints; AI visual initial screening eliminates unqualified products in real time, reduces invalid inspection processes, and improves inspection efficiency; the automation of equipment scheduling and parameter configuration ensures the standardization and consistency of the inspection process and reduces the risk of human operation.

[0065] Reference Figure 3 The data processing unit 32 analyzes the detection data and generates an evaluation report, which specifically includes the following steps:

[0066] B1. Three-dimensional visualization of test data: Call the digital twin module 321 to map the test data to the three-dimensional digital twin model of the joint, and dynamically display the pressure distribution cloud map, internal structure deformation area, leakage risk points and three-dimensional visualization results of the joint characteristic data according to the pre-set color coding rules; the joint characteristic data is the data generated by the personalized test items of different joints, and the digital twin module 321 can still use different visualization forms to present it, such as presenting abnormal vibration areas (such as flashing icons), flow fluctuation ranges (such as dynamic curve superposition), or performance change areas under the action of different media, etc.; implement diversified visualization strategies to ensure the efficient communication and accurate interpretation of complex test data, meet the differentiated detection needs of multiple scenarios and multiple types of joints, and improve the efficiency of human-computer interaction;

[0067] B2. Predicting potential joint defect risks and dynamically adjusting the inspection plan: Utilizing the defect prediction model of the predictive analysis module 322 to perform time series analysis on real-time inspection data and historical operating condition data, the potential joint defect risks are predicted, and the inspection plan's inspection item combination, the execution order of items in the inspection process, and parameter configuration are dynamically adjusted;

[0068] B3. Generate a multi-dimensional evaluation report: Integrate test data, digital twin visualization results, defect prediction reports, and adjusted test plans to generate a multi-dimensional evaluation report that includes pressure-time curves, CT slice image annotations, and environmental test performance fluctuation ranges. Through the above evaluation report generation process, multi-source data and intelligent analysis results are integrated. Through the collaboration of functional modules such as digital twin visualization, defect risk prediction, and test plan adjustment, efficient data utilization and dynamic optimization of test plans are achieved, ultimately forming a three-dimensional and structured test conclusion. The report content covers key information such as graphical data, CT defect annotations, and performance fluctuation ranges, which not only meets the technical staff's need for detailed tracing, but also provides management with intuitive decision-making basis, realizing multi-dimensional output of test value.

[0069] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the scope of protection of the invention. Obviously, the embodiments described are only some embodiments of the present invention, rather than all embodiments. Based on these embodiments, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in this field can still combine, add, delete or make other adjustments to the features in the various embodiments of the present invention according to the circumstances without conflict, without making creative work, so as to obtain different other technical solutions that do not deviate from the concept of the present invention in essence, and these technical solutions also fall within the scope of protection of the present invention.

Claims

1. A detection device for a pressure relief quick connector, characterized in that: include: A data acquisition module (1) includes a pressure sensor (11), a flow sensor (12), a displacement sensor (13) and an industrial camera (14), and is used to collect internal pressure, pressure relief rate, component displacement and appearance image data of the joint to be tested; A detection execution module (2) includes a pressure loading system (21), an environmental simulation chamber (22) and a pressure relief control mechanism (23), and is used to simulate working pressure, extreme environments and trigger pressure relief operations; and An intelligent control and analysis module (3) comprises a detection controller (31) and a data processing unit (32), wherein the detection controller (31) is used to generate a detection scheme based on basic information of a joint to be detected, and to control the data acquisition module (1) and the detection execution module (2) to detect the joint to be detected according to a detection process; the detection scheme includes at least one detection item, detection auxiliary equipment, detection process information, equipment configuration parameters and environmental simulation conditions; and the data processing unit (32) is used to analyze detection data and generate an evaluation report.

2. A detection device for a pressure relief quick connector according to claim 1, characterized in that: The data acquisition module (1) further comprises: An industrial CT scanning unit (15) is used to perform three-dimensional tomography scanning on the internal structure of the joint and construct a dynamic displacement image; The laser Doppler vibration measuring instrument (16) is used to measure the vibration spectrum of the joint surface in real time and identify abnormal vibration modes.

3. The detection device for a pressure relief quick connector according to claim 1, characterized in that: The detection controller (31) of the intelligent control and analysis module (3) comprises: A scheme matching model (311) is obtained by iteratively training a machine learning model based on a greedy algorithm and historical error data, and is used to generate a personalized detection scheme based on basic information of the joint to be detected; the basic information of the joint to be detected includes the joint model, material, design parameters, application scenario, historical working conditions and detection standards of the joint to be detected; The AI visual inspection model (312) is obtained by iteratively training the historical appearance inspection data of the joint to be inspected based on the YOLOv8 model or the ChatGPT-4.0 model, and is used to identify surface cracks, wear and deformation of the seal on the joint.

4. A detection device for a pressure relief quick connector according to claim 1, characterized in that: The environmental simulation cabin (22) comprises: A multi-field coupling loading device (221) is used to simultaneously apply a temperature of -40°C to 120°C, a pressure of 0-50 MPa, and a vibration of 10-2000 Hz to construct the environmental simulation conditions required for the detection scheme; The multi-axis mounting platform (222) is used for dynamically adjusting the joint installation angle and simulating the horizontal, vertical or different tilt angle installation postures of the joint to be tested.

5. The detection device for a pressure relief quick connector according to claim 1, characterized in that: The detection execution module (2) further includes: A plug-in drive unit (24) includes a multi-axis manipulator for performing plug-in and pull-out operations of the connector and a strain gauge force sensor provided at an output end of the multi-axis manipulator for real-time monitoring of plug-in and pull-out force changes; A multi-media fast switching system (25) comprises a solenoid valve group and multiple media output sources, and is used for switching hydraulic oil, compressed air and corrosive media according to instructions from a detection controller (31).

6. The detection device for a pressure relief quick connector according to claim 1, characterized in that: The data processing unit (32) comprises: A digital twin module (321) is used to generate a three-dimensional digital twin model of the joint, and dynamically display the pressure distribution, deformation area and leakage risk point according to the pre-set color coding rules based on the detection data; The predictive analysis module (322) includes a defect prediction model, which is used to predict and generate defect risks and dynamically adjust the detection item combination of the detection plan, the execution order of items in the detection process and parameter configuration based on the obtained detection data and the historical working conditions of the joint to be detected during the joint detection process; the defect prediction model is an LSTM neural network model obtained by iterative training through historical detection data.

7. The detection device for a pressure relief quick connector according to claim 3, characterized in that: The detection controller (31) generates a detection scheme based on the basic information of the joint to be detected, and controls the data acquisition module (1) and the detection execution module (2) to detect the joint to be detected, specifically comprising the following steps: Identifying a pre-set QR code on the joint to be inspected by an industrial camera (14), obtaining basic information of the joint to be inspected, and generating a personalized inspection plan by a plan matching model (311); Perform preliminary appearance defect screening on the joints to be inspected through the AI visual inspection model (312) and eliminate unqualified products; Dispatch detection auxiliary equipment according to the detection plan and send equipment configuration parameters for parameter configuration; According to the detection scheme and the detection process, a control instruction is generated to control the detection auxiliary equipment, the data acquisition module (1) and the detection execution module (2) to detect the joint to be detected according to the detection process.

8. The detection device for a pressure relief quick connector according to claim 6, characterized in that: The data processing unit (32) analyzes the detection data and generates an evaluation report, specifically comprising the following steps: Calling the digital twin module (321), mapping the detection data to the three-dimensional digital twin model of the joint, and dynamically displaying the pressure distribution cloud map, internal structure deformation area, leakage risk point and three-dimensional visualization results of the joint characteristic data according to the preset color coding rules; Using the defect prediction model of the predictive analysis module (322) to perform time series analysis on real-time detection data and historical working condition data, predict potential defect risks of the joint, and dynamically adjust the detection item combination of the detection plan, the execution order of the items in the detection process, and the parameter configuration; Integrate test data, digital twin visualization results, defect prediction reports, and adjusted test plans to generate a multi-dimensional evaluation report that includes pressure-time curves, CT slice image annotations, and environmental test performance fluctuation ranges.

9. The detection device for a pressure relief quick connector according to claim 6, characterized in that: The digital twin module (321) is also used to superimpose the performance change data of the joint under different environmental simulation conditions onto the three-dimensional digital twin model to generate a dynamic performance evolution video.

10. The detection device for a pressure relief quick connector according to claim 6, characterized in that: The predictive analysis module (322) also includes a defect-cause association reasoning unit (323) for establishing a defect-cause association knowledge base, matching the causes of the predicted defect risks, and generating possible causes and solution suggestions corresponding to the defect risks in the assessment report.

Images