Resin tank sealing performance detection method and device and storage medium

Abstract

The invention discloses a resin tank sealing performance detection method and device and a storage medium, and belongs to the technical field of detection, and the method comprises the steps: carrying out the key point analysis of a target resin tank; using condition analysis is carried out on the application scene information, and resin tank sealing performance detection condition parameters are determined; arranging a pressure sensor group on the key point position set of the resin tank, and detecting the sealing state of the resin tank based on the sealing detection condition parameters of the resin tank and the pressure sensor group; state detection is conducted on a resin tank cover of the target resin tank through an industrial camera and a displacement sensor; and performing sealing performance evaluation on the target resin tank based on the resin tank pressure sensing data flow set, the resin tank cover deformation image set and the resin tank cover displacement data flow. According to the invention, the technical problem of missing detection or misjudgment caused by low precision, poor efficiency and dependence on artificial experience of a traditional resin tank sealing detection method in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of detection technology, and in particular to a method, equipment and storage medium for detecting the sealing performance of a resin tank. Background Art

[0002] Resin tanks are widely used in the fields of chemical, food, and medical. As containers for storing and transporting liquids or gases, their sealing performance is crucial. Poor sealing not only leads to material leakage, environmental pollution, and waste of resources, but also affects product quality and reduces production safety. With the development of industrial automation and precision testing technology, efficient and accurate testing of the sealing of resin tanks has become a key requirement for improving product quality and ensuring safety.

[0003] At present, most methods for testing the sealing of resin tanks use a single detection method, such as using a mechanical sensor to perform a pressure leak test on the tank body. However, traditional pressure testing methods are difficult to fully evaluate the various working conditions encountered by resin tanks in actual applications, and lack accurate monitoring of details such as the deformation and displacement of the resin tank cover, which easily overlooks potential leakage risks. In addition, existing testing equipment usually lacks full integration with resin tank production design information and application scenarios, and it is difficult to optimize and adjust according to different conditions. With the advancement of industrial intelligence and sensor technology, there is an urgent need for a sealing detection method that integrates a variety of advanced sensing technologies and intelligent data analysis methods to improve the comprehensiveness, accuracy and automation level of detection and meet the high standards of modern industrial production. Summary of the invention

[0004] The present application provides a resin tank sealing detection method, equipment and storage medium, aiming to solve the problems of low accuracy, poor efficiency and reliance on manual experience in the conventional resin tank sealing detection method in the prior art, thereby leading to technical problems of missed detection or misjudgment.

[0005] In a first aspect, the present application provides a resin tank sealing detection method, the method comprising: fixing a target resin tank on a sealing detection platform, performing key point analysis on the target resin tank, and obtaining a resin tank key point set; obtaining application scenario information of the target resin tank, performing usage condition analysis on the application scenario information, and determining resin tank sealing detection condition parameters; deploying a pressure sensor group on the resin tank key point set, performing resin tank sealing status detection based on the resin tank sealing detection condition parameters and the pressure sensor group, and obtaining a resin tank pressure sensing data stream set; installing an industrial camera and a displacement sensor on the sealing detection platform, performing status detection on a resin tank cover of the target resin tank through the industrial camera and the displacement sensor, and obtaining a resin tank cover deformation image set and a resin tank cover displacement data stream; performing sealing performance evaluation on the target resin tank based on the resin tank pressure sensing data stream set, the resin tank cover deformation image set, and the resin tank cover displacement data stream, and generating a resin tank sealing performance detection result.

[0006] In a second aspect, the present application provides an electronic device, comprising: a processor; a memory for storing instructions executable by the processor; wherein the processor is used to execute the resin tank sealing detection method provided in the present application.

[0007] In a third aspect, the present application provides a computer-readable storage medium storing a computer program, and the computer program is used to execute the resin tank sealing detection method provided in the present application.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: By adopting the technical solution of integrating mechanical sensor groups, industrial cameras and displacement sensors at key points of resin tanks, combined with application scenario information analysis and multi-source data fusion analysis, the traditional resin tank sealing detection method in the prior art has solved the problems of low precision, poor efficiency and reliance on manual experience, which leads to missed detection or misjudgment, and achieved high-precision, high-efficiency and automated sealing detection. This method uses multiple sensors to monitor the pressure changes, tank cover deformation and displacement of the resin tank in real time, and uses intelligent data processing technology to comprehensively cover and accurately analyze all potential sealing weaknesses. The results not only improve the accuracy and reliability of detection, but also significantly reduce manual operations, improve detection efficiency, ensure the sealing performance and safety of the resin tank in actual use, and meet the strict requirements of modern industry for high-standard sealing detection.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A flow chart of a method for detecting the sealing performance of a resin tank is provided for this application.

[0011] Figure 2 A schematic diagram of the structure of an electronic device provided in this application.

[0012] Description of the reference numerals: processor 21 , memory 22 , input device 23 , output device 24 . DETAILED DESCRIPTION

[0013] The overall idea of ​​the technical solution provided by this application is as follows: The embodiments of the present application provide a method, device and storage medium for testing the sealing performance of a resin tank. First, the resin tank is fixed on a testing platform, and the parts that need to be monitored are identified through key point analysis. Then, the test condition parameters are set according to the application scenario information, and a pressure sensor group is arranged at the key points. An industrial camera and a displacement sensor are installed to monitor the deformation and displacement of the tank cover. Finally, the pressure, deformation and displacement data are integrated to evaluate the sealing performance through data fusion analysis.

[0014] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically described below in conjunction with the drawings in the specification.

[0015] Embodiment 1, as Figure 1 As shown, the embodiment of the present application provides a method for detecting the sealing property of a resin tank, the method comprising: Step S100: a target resin tank is fixedly placed on a sealing detection platform, and key point analysis is performed on the target resin tank to obtain a resin tank key point set.

[0016] Specifically, the target resin tank refers to the resin tank that needs to be tested for sealing, which is usually a container for storing or transporting chemical materials such as resin. In this step, the target resin tank is the object to be tested. The sealing test platform refers to a dedicated platform used to support and fix the resin tank and provide the required conditions. The platform is usually integrated with various sensors and testing equipment, which can be used to accurately obtain various data of the resin tank. Key point analysis refers to the analysis of the structure of the resin tank to identify the key parts (such as seams, welding points, valves, lids, etc.) that need special attention during the sealing test. These parts are usually where sealing performance problems are most likely to occur. The resin tank key point set is a set of multiple key points, which represent the parts on the resin tank that affect the sealing performance. Through these key points, the resin tank can be tested in more detail.

[0017] First, fix the resin tank on the sealing test platform. The platform usually includes a stable support system to ensure the stability of the resin tank during the test. The platform is equipped with a laser locator, industrial camera or other precision instruments to ensure the accurate positioning of the resin tank.

[0018] Next, the key point analysis identifies areas where sealing problems occur by analyzing the design drawings, production processes or computer-aided design (CAD) models of the resin tank. These areas usually include structural weaknesses such as the seams, interfaces, covers, and welds of the resin tank. In some applications, especially in environments with large pressure and temperature changes, the material and shape of the sealing parts are particularly important. For example, suppose the design drawings of the target resin tank show that there is a welding point at the connection between the top cover and the tank body, where the seal is prone to failure due to factors such as thermal expansion. Then, this welding point will be identified as a key point and included in the key point set of the resin tank. In this way, all sealing weaknesses can be accurately located.

[0019] Specifically, the key point analysis uses computer-aided design (CAD) software to model and combines finite element analysis (FEA) methods to perform stress analysis on the resin tank to predict the performance of each key point under different working conditions. Laser scanning and 3D modeling technology are also applied to the actual resin tank to further improve the accuracy of the analysis.

[0020] This step ensures that there are no blind spots in the sealing test of the resin tank. Not only does it cover all the weak points of the sealing, but it can also conduct targeted testing based on the actual design and use environment of the resin tank.

[0021] Step S200: Acquire application scenario information of the target resin tank, perform usage condition analysis on the application scenario information, and determine resin tank sealing detection condition parameters.

[0022] Specifically, the application scenario information includes the environment and condition information of the target resin tank during actual use, such as storage environment (temperature, humidity), transportation method (pressure change, vibration), operating parameters during use (temperature, pressure), and external influencing factors (chemical corrosion, mechanical stress). The sealing test condition parameters refer to the specific test parameters determined based on the results of the use condition analysis, such as the test pressure range, temperature range, applied mechanical stress, test time period, etc. These parameters are used to simulate the working state of the resin tank in actual use to ensure the comprehensiveness and accuracy of the sealing test.

[0023] First, application scenario information is obtained through a variety of data collection methods, including collecting relevant data from the resin tank's manual, production records, on-site monitoring equipment (such as temperature sensors, pressure sensors), and user feedback. For example, if the resin tank is used to store resin in a high-temperature environment, the relevant temperature data and environmental description will be recorded.

[0024] Next, conditional analysis is used to conduct in-depth analysis of the collected application scenario information. Data analysis software (such as MATLAB and Python data analysis libraries) is used to process environmental parameters and identify key factors that affect sealing performance. For example, through analysis, it is found that resin tanks are often subjected to high pressure and severe vibration during transportation. These factors will serve as an important reference for sealing detection.

[0025] Based on the analytical results, the parameters of the sealing test conditions are determined. This includes setting the specific pressure range, temperature range, applied vibration frequency and amplitude required for the test. For example, if the analytical results show that the highest pressure experienced by the resin tank during transportation is 5 bar, then the pressure parameters of the sealing test will be set in a range slightly higher than 5 bar to ensure that the tank can withstand the corresponding pressure without leakage in actual use.

[0026] By acquiring and analyzing the application scenario information of the target resin tank, the condition parameters of the sealing test can be accurately determined, making the testing process closer to the actual use environment. It not only improves the accuracy and reliability of sealing testing, but also can identify sealing problems that occur under specific use conditions in advance, thereby effectively preventing risks such as resin leakage, environmental pollution and equipment damage caused by sealing failure.

[0027] Step S300: deploying a pressure sensor group on the resin tank key point set, performing a resin tank sealing state detection based on the resin tank sealing detection condition parameters and the pressure sensor group, and acquiring a resin tank pressure sensing data stream set.

[0028] Specifically, the pressure sensor group is a collection of multiple pressure sensors, which are arranged at key points of the resin tank and are used to monitor and record in real time the pressure changes at key points of the resin tank during operation. The resin tank sealing test condition parameters refer to specific test parameters determined based on application scenario information and usage condition analysis, including test pressure range, temperature conditions, applied mechanical stress, etc. These parameters are used to simulate the working state of the resin tank in actual use. The pressure sensing data stream set is a collection of pressure data continuously collected by the pressure sensor group during the sealing test process, which is used for subsequent analysis and sealing performance evaluation.

[0029] First, according to the determined set of key points of the resin tank, a pressure sensor group is arranged at each key point. These pressure sensor groups usually include high-precision pressure sensors, data acquisition modules and signal processing units. Next, based on the determined resin tank sealing test condition parameters, the working parameters of the pressure sensor group, such as the measurement range, sampling frequency and calibration settings, are configured. Then, the resin tank sealing status detection process is started. During this process, the pressure sensor group will continuously monitor the pressure changes at each key point of the resin tank under the preset detection conditions. For example, after applying a certain pressure, the sensor group will record the pressure data of each key point to evaluate whether the sealing performance meets the requirements.

[0030] Acquiring the resin tank pressure sensing data stream set means aggregating the real-time pressure data collected by all pressure sensor groups during the detection process into a complete data set. These data stream sets can be transmitted to the central processing system via wired or wireless networks for subsequent data analysis and sealing performance evaluation.

[0031] By placing pressure sensor groups at key points of the resin tank and performing sealing status detection based on detection condition parameters, efficient and accurate sealing performance monitoring can be achieved.

[0032] Step S400: installing an industrial camera and a displacement sensor on the sealing detection platform, and performing state detection on the resin tank cover of the target resin tank through the industrial camera and the displacement sensor to obtain a resin tank cover deformation image set and a resin tank cover displacement data stream.

[0033] Specifically, the resin tank cover is the upper closed part of the resin tank, which is used to seal the tank body to prevent resin leakage and external material intrusion. The sealing performance of the tank cover directly affects the sealing and safety of the entire resin tank. The resin tank cover deformation image set refers to the set of deformation images generated by the resin tank cover during the inspection process captured by the industrial camera. These images are used to analyze the structural changes and potential sealing problems of the tank cover. The resin tank cover displacement data stream is a set of resin tank cover displacement data collected in real time by the displacement sensor. These data reflect the specific displacement of the tank cover under applied load or environmental changes, which is used for further sealing performance evaluation.

[0034] First, the industrial cameras and displacement sensors are arranged reasonably on the sealing test platform to ensure that they can cover the key areas of the resin tank cover. The industrial camera is usually installed directly above or on the side of the tank cover, and the high-resolution lens is used to capture the subtle deformation of the tank cover during the test. The displacement sensor is installed at the key connection parts of the tank cover, such as the threaded interface or around the sealing ring, to record the displacement changes of the tank cover in real time when pressure or other loads are applied.

[0035] Next, the status detection program is started to test the resin tank according to the preset sealing test condition parameters (such as pressure, temperature, vibration, etc.). During the test, the industrial camera continuously captures the deformation image of the tank cover and transmits the image data to the central processing system in real time to form a set of deformation images of the resin tank cover. At the same time, the displacement sensor monitors the displacement of the tank cover in real time, and records the measured displacement data through the data acquisition system to form a displacement data stream of the resin tank cover. These images and data will be used for subsequent analysis and evaluation, using image processing technology to identify the slight deformation of the tank cover, and using displacement data to analyze the response of the tank cover under different loads.

[0036] By installing industrial cameras and displacement sensors on the sealing detection platform, combined with real-time monitoring and analysis of deformation images and displacement data, efficient and accurate detection of the sealing performance of resin can lids can be achieved. This multi-dimensional detection method not only improves the comprehensiveness and accuracy of sealing evaluation, but also significantly improves detection efficiency and reliability through automated and intelligent detection processes, meeting the strict requirements of modern industry for high-standard sealing performance testing.

[0037] Step S500: Evaluate the sealing performance of the target resin can based on the resin can pressure sensing data stream set, the resin can cover deformation image set and the resin can cover displacement data stream, and generate a resin can sealing performance test result.

[0038] Specifically, the displacement data stream refers to the set of displacement data of the can cover collected in real time by the displacement sensor during the detection process. These data record the displacement changes of the can cover under the action of external forces and are used to evaluate the stability and sealing performance of the can cover. The sealing performance test result refers to the final test report generated by the sealing performance evaluation, which includes the conclusion of whether the sealing is qualified, as well as specific evaluation indicators and analysis data.

[0039] First, the pressure sensing data stream set, deformation image set and displacement data stream are synchronously collected and transmitted to the central data processing system. These data come from the pressure sensor group, industrial camera and displacement sensor deployed in the previous step. In order to evaluate the sealing performance, it is necessary to use comprehensive data analysis software (such as MATLAB, Python data analysis library, dedicated industrial inspection software) to process and analyze these data.

[0040] Specifically, the system first analyzes the pressure sensing data stream set to identify the pressure distribution and change trend under different pressure conditions. Then, the deformation image set is processed using image processing technology (such as edge detection and deformation analysis algorithm) to identify small deformations, cracks or other structural changes of the tank cover. At the same time, the displacement data stream is evaluated by the displacement analysis algorithm to evaluate the displacement amplitude and direction of the tank cover under different loads.

[0041] Next, the system combines these analysis results and uses machine learning algorithms or rule engines to comprehensively evaluate the sealing performance of the resin tank. For example, algorithms such as support vector machines (SVM), decision trees, or deep neural networks (DNN) can be used to combine pressure, deformation, and displacement data to predict whether the sealing performance meets the standards. Finally, the system generates sealing performance test results, including whether the sealing is qualified, the specific values ​​of each indicator, and the location of potential sealing defects.

[0042] By comprehensively analyzing pressure, deformation and displacement data, a comprehensive evaluation of the sealing performance of the resin tank is achieved, generating accurate and detailed test results. This multi-dimensional, intelligent testing method not only improves the accuracy and efficiency of sealing evaluation, but also enhances the safety and reliability of the use of the resin tank.

[0043] Furthermore, the method of obtaining the key point set of the resin tank includes: obtaining production design information of the target resin tank, performing three-dimensional modeling based on the production design information, and generating a resin tank space model; determining a mesh type and a mesh density according to sealing analysis requirements and model complexity; performing finite element division on the resin tank space model according to the mesh type and mesh density to obtain a finite element mesh model of the resin tank; and performing sealing simulation testing and key point analysis based on the resin tank finite element mesh model to obtain a resin tank key point set.

[0044] Specifically, production design information refers to all relevant data and information generated during the design and manufacturing stages of the resin tank, including design drawings, specification parameters, material selection, manufacturing process flow, etc. This information describes in detail the structure, size, shape and components of the resin tank. Three-dimensional modeling refers to the use of computer-aided design (CAD) software to create a three-dimensional digital model of the resin tank based on the production design information. Three-dimensional modeling can accurately reflect the geometric structure and spatial layout of the resin tank, which is convenient for subsequent analysis and simulation. The resin tank spatial model is a digital three-dimensional model of the resin tank generated by three-dimensional modeling, which contains all its structural details and geometric features for further finite element analysis and sealing simulation testing. Mesh type refers to the geometric shape and arrangement of a three-dimensional space model into several small units (meshes) in finite element analysis. Common mesh types include tetrahedral meshes, hexahedral meshes, hybrid meshes, etc. Different types of meshes are suitable for different analysis requirements and model complexity. Mesh density refers to the number of mesh units per unit volume during the finite element division process. The higher the mesh density, the finer the mesh units are divided, and the accuracy of the analysis results is usually higher, but the calculation amount and time cost also increase accordingly.

[0045] First, collect the production design information of the target resin tank, including design drawings, specification parameters and manufacturing process. This information will be input into computer-aided design (CAD) software for 3D modeling to generate a spatial model of the resin tank. For example, using software such as AutoCAD or SolidWorks, a 3D digital model that accurately reflects the structure and size of the resin tank can be created.

[0046] Next, determine the appropriate mesh type and mesh density based on the sealing analysis requirements and model complexity. If the resin tank structure is more complex, a hybrid mesh or hexahedral mesh needs to be used, and the mesh density needs to be increased to improve the analysis accuracy; for parts with simpler structures, a tetrahedral mesh can be used and the mesh density can be appropriately reduced to reduce the consumption of computing resources.

[0047] Subsequently, according to the determined mesh type and mesh density, the finite element analysis (FEA) software (such as ANSYS, Abaqus) is used to perform finite element division on the spatial model of the resin tank to generate a finite element mesh model of the resin tank. This process subdivides the overall structure of the resin tank into a large number of small mesh units, each of which will serve as an independent calculation unit for subsequent physical quantity simulation.

[0048] Finally, based on the generated finite element mesh model, a sealing simulation test and key point analysis were performed. By applying pressure, temperature and other working conditions under actual use conditions, the sealing performance of the resin tank under working conditions was simulated, and the stress distribution and deformation of each key part were calculated.

[0049] Through this step, the potential sealing weaknesses of the resin tank under actual use conditions can be systematically and comprehensively identified, forming an accurate set of key points. This method not only improves the pertinence and accuracy of sealing detection, reduces the hidden dangers missed in traditional detection, but also predicts the performance of sealing performance in advance through finite element analysis, enhancing the scientificity and reliability of detection.

[0050] Furthermore, the obtaining of the key point set of the resin tank includes: defining sealing boundary conditions, and setting pressure test load information according to the working pressure range of the target resin tank; applying the sealing boundary conditions and the pressure test load information to the resin tank finite element mesh model to perform sealing simulation analysis to obtain the stress distribution parameters of the resin tank; based on the resin tank stress distribution parameters, performing stress area identification and deformation assessment to obtain a resin tank stress concentration area set and a resin tank deformation parameter set; based on the resin tank stress concentration area set and the resin tank deformation parameter set, performing key point analysis on the resin tank finite element mesh model to obtain the resin tank key point set.

[0051] Specifically, sealing boundary conditions refer to the boundary limits and environmental conditions set in finite element analysis to simulate the sealing performance of the resin tank under actual working conditions. This includes defining how the sealing parts of the resin tank are isolated from the external environment, such as fixing certain parts, applying specific pressures or temperatures, etc. Pressure test load information refers to the specific pressure data applied to the finite element model of the resin tank in the sealing simulation analysis. This includes the size of the pressure, the location where it is applied, the way it is applied (such as internal pressure, external pressure), and how the pressure changes over time.

[0052] First, define the sealing boundary conditions, including setting the isolation method of the sealing part of the resin tank from the external environment. For example, if the lid of the resin tank is fixed by a threaded connection, the threaded connection can be defined as a fixed boundary in the finite element model to prevent it from moving during the simulation. At the same time, according to the working pressure range of the resin tank, such as 1 bar to 10 bar, set the corresponding pressure load information to ensure that the simulation analysis can cover all pressure conditions in actual use.

[0053] Next, the sealing boundary conditions and pressure test load information are applied to the finite element mesh model. Using finite element analysis (FEA) software (such as ANSYS, Abaqus), predefined boundary conditions and pressure loads are applied to the model to perform a sealing simulation analysis. This analysis will calculate the stress distribution of the resin tank under different pressure conditions and generate stress distribution parameters. For example, the stress value of a welded joint under a pressure of 10 bar is 150MPa.

[0054] Subsequently, stress area identification and deformation assessment are performed based on the obtained stress distribution parameters. By setting the stress threshold, areas where stress exceeds the critical value, such as seams or welds, are identified. At the same time, the deformation of the resin tank under pressure is evaluated, such as the slight deformation of the lid under high pressure, which leads to seal failure.

[0055] Finally, based on the stress concentration area set and deformation parameter set, the finite element mesh model is subjected to key point analysis. All key areas with significant stress concentration and deformation are identified and grouped into a resin tank key point set.

[0056] By defining the sealing boundary conditions and pressure test load information and applying them to the finite element mesh model for sealing simulation analysis, the stress distribution and deformation of the resin tank under different pressure conditions can be accurately obtained. This improves the scientificity and effectiveness of sealing testing and meets the strict requirements of high-standard industrial applications for the sealing performance of resin tanks.

[0057] Furthermore, the generation of the resin tank sealing performance test result includes: setting an abnormal value determination condition, and performing abnormal value identification on the resin tank pressure sensing data stream set and the resin tank cover displacement data stream according to the abnormal value determination condition to obtain an abnormal resin tank detection data set; determining an abnormal data preprocessing process according to the abnormal value determination condition; preprocessing the abnormal resin tank detection data set based on the abnormal data preprocessing process to obtain a standard resin tank pressure sensing data stream set and a standard resin tank cover displacement data stream; performing sealing performance evaluation on the target resin tank based on the standard resin tank pressure sensing data stream set and the standard resin tank cover displacement data stream, as well as the resin tank cover deformation image set, to generate a resin tank sealing performance test result.

[0058] Specifically, the outlier judgment conditions refer to the standards and rules used to identify and distinguish normal data from abnormal data. These conditions can be based on statistical analysis, threshold setting or machine learning models, and are designed to accurately identify abnormal data points that appear during the sealing detection process. The abnormal resin tank detection data set is a set of data points identified as abnormal in the pressure sensing data stream set and the displacement data stream. The standard resin tank pressure sensing data stream set refers to a set of pressure sensing data that meets normal operating conditions after abnormal data preprocessing. These data are used for subsequent sealing performance evaluation as a reference standard. The standard resin tank cover displacement data stream refers to a set of tank cover displacement data that meets normal operating conditions after abnormal data preprocessing. These data are used to evaluate the displacement of the tank cover to ensure that it is within the normal range.

[0059] First, the outlier judgment conditions are set to establish the criteria for identifying normal and abnormal data. These conditions can be based on statistical methods (such as setting upper and lower limits for pressure and displacement), machine learning algorithms (such as anomaly detection models), or industry standards. For example, the normal range of pressure sensing data is set to 0-10 bar, and the normal range of displacement data is set to 0-1 mm. Data points outside these ranges will be considered abnormal.

[0060] Next, the pressure sensing data stream set and the displacement data stream are identified as outliers according to the outlier judgment criteria to obtain the abnormal resin tank detection data set. This process is usually automated using data analysis software (such as MATLAB, Python's Pandas and Scikit-learn libraries). By writing scripts or using ready-made algorithms, large amounts of data can be quickly scanned to identify abnormal points that exceed the preset range. For example, in the pressure data, the pressure at a key point suddenly rises to 15 bar, which obviously exceeds the set upper limit of 10 bar, and is identified as an abnormality.

[0061] Subsequently, the abnormal data preprocessing process is determined based on the outlier determination criteria. This step involves selecting appropriate preprocessing methods, such as data cleaning (deleting noise data), data correction (correcting erroneous measurements), or data interpolation (filling missing values). The selected preprocessing method depends on the nature of the outliers and the purpose of detection. Then, the abnormal resin tank detection dataset is preprocessed based on the abnormal data preprocessing process to obtain a standard resin tank pressure sensing data stream set and a standard resin tank cover displacement data stream. This process ensures that the subsequent sealing performance evaluation is based on clean and accurate data. For example, after preprocessing, the abnormal points in the pressure dataset are removed or corrected to ensure that the dataset contains only valid data reflecting the normal operating status of the resin tank.

[0062] Finally, based on the standard data stream set and the resin tank cover deformation image set, the sealing performance of the target resin tank is evaluated to generate the sealing performance test results. This includes comprehensive analysis of pressure distribution, displacement change and deformation images, and uses data fusion technology and machine learning models to comprehensively judge whether the sealing performance of the resin tank meets the standards.

[0063] Through the system's outlier identification and data preprocessing, combined with standardized data streams and deformation images, a high-precision evaluation of the sealing performance of the resin tank is achieved. This approach not only improves the accuracy and reliability of the test results, but also optimizes the test process and enhances decision support and preventive maintenance capabilities.

[0064] Furthermore, the generation of the resin tank sealing performance test result includes: parsing the standard resin tank pressure sensing data stream set and the standard resin tank cover displacement data stream to obtain the resin tank pressure distribution parameters and the resin tank cover displacement change parameters; performing noise identification and filtering preprocessing on the resin tank cover deformation image set to obtain the available resin tank cover deformation image set; performing tank cover deformation analysis based on the available resin tank cover deformation image set to obtain the resin tank cover deformation parameters; and evaluating the sealing performance of the target resin tank based on the resin tank pressure distribution parameters and the resin tank cover displacement change parameters, as well as the resin tank cover deformation parameters, to obtain the resin tank sealing performance test result.

[0065] Specifically, the pressure distribution parameters of the resin tank refer to the parameters about the pressure distribution of the resin tank under different pressure conditions obtained by parsing the standard pressure sensing data stream set, such as the maximum pressure value and pressure gradient of each key point. The displacement change parameters of the resin tank cover refer to the parameters about the displacement change of the resin tank cover under different pressure conditions obtained by parsing the standard displacement data stream, such as displacement amplitude, displacement rate, etc. The deformation image set of the resin tank cover refers to the set of deformation images of the resin tank cover collected by the industrial camera after noise identification and filtering preprocessing, which are used for further deformation analysis. The available deformation image set of the resin tank cover refers to the set of deformation images of the resin tank cover with qualified quality and no noise interference after noise identification and filtering preprocessing, which is suitable for subsequent deformation analysis. The deformation parameters of the resin tank cover refer to the parameters obtained by deformation analysis of the available deformation image set, such as deformation amplitude, deformation position, deformation rate, etc., which are used to evaluate the structural integrity and sealing performance of the tank cover.

[0066] First, the standard resin tank pressure sensing data stream set and the standard resin tank cover displacement data stream are analyzed to obtain the resin tank pressure distribution parameters and the resin tank cover displacement change parameters. This process is usually performed using data analysis software (such as MATLAB, Python's Pandas and NumPy libraries). By statistically analyzing the pressure data, the maximum pressure value, pressure gradient and other parameters of each key point can be extracted; similarly, the displacement data is analyzed through time series to obtain the displacement amplitude and change trend of the tank cover under different pressure conditions.

[0067] Next, the deformation image set of the resin can lid is subjected to noise identification and filtering preprocessing to obtain a usable deformation image set of the resin can lid. This process is usually performed using image processing software (such as OpenCV, MATLAB's image processing toolbox). By applying filtering algorithms (such as Gaussian filtering, median filtering) and noise identification techniques, noise and interference in the image are removed, and the image quality is improved, making it suitable for subsequent deformation analysis.

[0068] Then, the deformation analysis of the can cover is performed based on the available deformation image set of the resin can cover to obtain the deformation parameters of the resin can cover. This analysis is usually performed using image recognition and computer vision techniques (such as edge detection, feature extraction, and deformation measurement algorithms). By comparing the deformation images at different time points, the deformation amplitude, deformation position, and change rate of the can cover during the inspection process can be accurately measured to obtain the deformation parameters.

[0069] Finally, based on the pressure distribution parameters of the resin tank, the displacement change parameters of the resin tank cover, and the deformation parameters of the resin tank cover, the sealing performance of the target resin tank is evaluated to generate the test results of the sealing performance of the resin tank. This evaluation process is usually carried out in combination with data fusion technology and machine learning algorithms (such as support vector machines, decision trees, and neural networks). Comprehensively analyze the pressure distribution, displacement change, and deformation parameters to determine whether the sealing performance of the resin tank meets the standards, and generate a detailed test report, including whether the sealing is qualified, the specific values ​​of various indicators, and the location and cause analysis of existing sealing defects.

[0070] By analyzing and processing the standard pressure sensing data stream set and displacement data stream, the pressure distribution and displacement changes of the resin tank under different pressure conditions can be accurately extracted, providing a quantitative basis for the evaluation of sealing performance. Noise identification and filtering preprocessing are performed on the deformation image set to ensure image quality and improve the accuracy of deformation analysis. Based on the available deformation image set, the deformation analysis of the tank cover can accurately identify and quantify the deformation of the tank cover, further improving the comprehensiveness and accuracy of the sealing performance evaluation.

[0071] Furthermore, the obtaining of the deformation parameters of the resin tank cover includes: determining a preset resin tank cover image according to the production design information of the target resin tank; building a feature twin network, and performing feature extraction processing on the preset resin tank cover image and the available resin tank cover deformation image set based on the feature twin network, and outputting a preset resin tank cover feature set and a current resin tank cover feature set; performing a comparison loss analysis on the current resin tank cover feature set based on the preset resin tank cover feature set to obtain resin tank cover loss feature data; performing a tank cover deformation analysis based on the resin tank cover loss feature data to obtain the deformation parameters of the resin tank cover.

[0072] Specifically, the feature twin network is a deep learning model composed of two or more neural networks with shared weights, which is used to compare the similarity between two inputs. It is often used for tasks such as image matching and similarity measurement. The current resin can lid feature set refers to the feature vector set obtained by extracting features from the available resin can lid deformation image set obtained during the actual detection process through the feature twin network, which reflects the actual state of the resin can lid during the detection process.

[0073] First, the preset resin tank cover image is determined based on the production design information of the target resin tank. This process usually uses computer-aided design (CAD) software to generate a standard 3D model of the resin tank cover based on the design drawings and render it into a high-resolution 2D image. These preset images represent the appearance and structure of the resin tank cover in an ideal state, serving as a benchmark for subsequent comparative analysis.

[0074] Next, a feature twin network is built, and based on the network, feature extraction processing is performed on the preset resin tank cover image and the available resin tank cover deformation image set, and the preset resin tank cover feature set and the current resin tank cover feature set are output. The feature twin network consists of two convolutional neural networks with shared weights, which can extract high-dimensional feature vectors from the input image. Specifically, first, data preparation is performed to collect and organize the preset resin tank cover image and the deformation image obtained during the actual detection process to ensure that each pair of images has a corresponding relationship. Next, the network architecture is designed, and two convolutional neural network branches with shared weights are constructed using a deep learning framework such as TensorFlow or PyTorch. Each branch processes an input image and extracts a high-dimensional feature vector. Then, an appropriate loss function is selected, and contrastive loss or triplet loss is commonly used to measure the similarity or difference between the features of two images. Subsequently, the network is trained, and the network parameters are optimized by back propagation using the labeled image pairs, so that the feature vectors of similar image pairs are closer and the distances of non-similar images are farther. During the training process, hyperparameters such as learning rate, batch size, and network depth are tuned to improve model performance. Finally, verification and testing are performed to evaluate the accuracy and robustness of the network through an independent validation set to ensure that it can accurately extract and compare the deformation features of the resin can lid in practical applications. Through this process, the feature twin network can effectively identify and quantify the slight deformation of the resin can lid, supporting the accurate evaluation of sealing performance.

[0075] Then, a comparison loss analysis is performed on the current resin tank cap feature set based on the preset resin tank cap feature set to obtain the resin tank cap loss feature data. Comparison loss analysis usually uses a loss function such as Euclidean distance or cosine similarity to quantify the difference between the preset feature and the current feature. For example, the Euclidean distance between two feature vectors is calculated to measure the degree of deformation of the resin tank cap.

[0076] Finally, the deformation analysis of the resin tank cover is performed based on the loss characteristic data to obtain the deformation parameters of the resin tank cover. This analysis step converts the loss characteristic data into specific deformation parameters, such as deformation amplitude and deformation area distribution, by setting a threshold or using a regression model. For example, if the loss characteristic data exceeds the set threshold, it indicates that the tank cover has significant deformation and requires further inspection and maintenance.

[0077] By building a feature twin network and deep learning technology, the deformation features of the resin can cover are accurately extracted and compared, achieving high-precision acquisition of the can cover deformation parameters.

[0078] Furthermore, the resin tank sealing performance test result is obtained, including: collecting and acquiring a resin tank sealing monitoring database, using a support vector machine to perform identification training on the resin tank sealing monitoring database, and obtaining a sealing performance standard discriminator; using a deep neural network to perform identification training on the resin tank sealing monitoring database, and obtaining a sealing performance grade analysis network; merging and connecting the sealing performance standard discriminator and the sealing performance grade analysis network to generate a resin tank sealing performance evaluation model; based on the resin tank sealing performance evaluation model, the resin tank pressure distribution parameters, the resin tank cover displacement change parameters, and the resin tank cover deformation parameters are evaluated for sealing performance, and the resin tank sealing performance test result of the target resin tank is obtained.

[0079] Specifically, the resin tank sealing monitoring database refers to the collection of all relevant data collected and stored during the sealing detection process, including pressure sensing data, tank cover displacement data, deformation images and other relevant parameters. These data are used to train and verify machine learning models to evaluate the sealing performance of resin tanks. Support vector machine is a supervised learning algorithm mainly used for classification and regression analysis. SVM distinguishes data points of different categories by finding the best segmentation hyperplane, which is suitable for classification tasks of high-dimensional data. Identification training refers to the use of labeled data sets to train machine learning models so that they can recognize and classify new data. During the identification training process, the model learns patterns and features in the data to achieve accurate classification and prediction. The sealing performance standard discriminator is a model trained based on the support vector machine, which is used to judge whether the sealing performance of the resin tank meets the predetermined standard. The discriminator can judge whether the sealing is qualified based on the input data. Deep neural network is a complex neural network composed of multiple layers of neurons, which can automatically extract and learn high-level features in data. DNN is widely used in image recognition, speech processing and complex pattern recognition. The sealing performance grade analysis network refers to a model trained based on the deep neural network, which is used to grade the sealing performance of the resin tank. The network can classify the sealing performance in detail according to the input data and provide more detailed evaluation results. The sealing performance evaluation model refers to a comprehensive evaluation model formed by merging and connecting the sealing performance standard discriminator and the sealing performance level analysis network. This model combines the advantages of SVM and DNN and can comprehensively and accurately evaluate the sealing performance of the resin tank. The pressure distribution parameters refer to the parameters of the pressure distribution of each key point of the resin tank under different pressure conditions, such as the maximum pressure value, pressure gradient, etc.

[0080] First, the resin tank sealing monitoring database is acquired. This process involves collecting all relevant data generated during the sealing detection process, including pressure sensing data streams, tank cover displacement data streams, and tank cover deformation image sets. Through sensors and image acquisition devices, these data are systematically stored in the database, providing a basis for subsequent model training.

[0081] Next, the support vector machine (SVM) is used to train the resin tank sealing monitoring database to obtain a sealing performance standard discriminator. In this stage, the SVM model is trained with a labeled data set (e.g., samples with qualified and unqualified sealing) so that it can judge whether the sealing performance of the resin tank meets the standard based on the input data. Specifically, the pressure sensing data stream set and displacement data stream in the resin tank sealing monitoring database are collected and sorted to ensure that the data is complete and clearly labeled (such as "meeting the standard" or "not meeting the standard"). Next, the data is preprocessed, including normalization, denoising, and missing value filling, to improve the quality of model training. Then, the support vector machine (SVM) algorithm is selected, and the preprocessed data set is divided into a training set and a test set. The SVM model is trained with the training set, and the hyperparameters (such as kernel function type and penalty coefficient) are optimized through cross-validation to improve the classification accuracy of the model. After training, the model performance is verified on the test set to ensure that it has good generalization ability. Finally, the trained SVM model is saved as a sealing performance standard discriminator, which can accurately judge the new test data. SVM finds the best segmentation hyperplane to distinguish data points of different categories and realizes binary classification judgment of sealing performance.

[0082] Subsequently, a deep neural network (DNN) is used to train the resin tank sealing monitoring database to obtain a sealing performance level analysis network. Specifically, the multidimensional data in the resin tank sealing monitoring database, including the pressure sensing data stream, the tank cover displacement data stream and the deformation image, are collected and sorted, and the data are annotated and classified according to the sealing performance level (such as excellent, good, medium, and poor). Next, the data is preprocessed, including normalizing the pressure and displacement data, enhancing and cleaning the deformation image to improve the data quality and consistency. Then, the deep neural network (DNN) architecture is designed, such as using a convolutional neural network (CNN) for image feature extraction, and combining it with a fully connected layer for classification. Subsequently, the data set is divided into a training set, a validation set, and a test set. The training set is used to train the DNN model, the network weights are optimized by back propagation, and hyperparameters (such as learning rate and batch size) are tuned on the validation set to prevent overfitting. After training, the model performance is evaluated, and the accuracy and classification ability of the network are verified using the test set to ensure that it can effectively distinguish different sealing performance levels. Finally, the trained DNN model is saved and deployed, and integrated into the sealing performance evaluation system to achieve automated and intelligent sealing performance grade analysis. DNN automatically extracts and learns high-level features in the data through multiple layers of neurons, and can perform detailed grading of sealing performance.

[0083] Then, the sealing performance standard discriminator and the sealing performance grade analysis network are merged and connected to generate a sealing performance evaluation model. Specifically, the sealing performance standard discriminator (SVM model) and the sealing performance grade analysis network (DNN model) are integrated on the same platform, and programming languages ​​such as Python and deep learning frameworks such as TensorFlow or PyTorch are used to implement the joint operation of the model. Then, an integrated interface is designed so that the outputs from SVM and DNN can be uniformly received and processed. For example, the input data is first judged by SVM whether it meets the standard. If it meets the standard, the data is passed to DNN for detailed grade analysis. Use API or middleware to connect the two models to ensure smooth data flow and synchronous output of results. Finally, test and optimize the integrated model to ensure its consistency and accuracy on different data sets.

[0084] The resin tank sealing performance evaluation model combines the high efficiency of SVM in binary classification tasks and the meticulousness of DNN in multi-classification tasks, and can comprehensively and accurately evaluate the sealing performance of the resin tank. Finally, based on the sealing performance evaluation model, the pressure distribution parameters, tank cover displacement change parameters and tank cover deformation parameters are used to evaluate the sealing performance, and the final resin tank sealing performance test results are generated.

[0085] By collecting the resin tank sealing monitoring database and using support vector machines and deep neural networks for identification training, a sealing performance standard discriminator and a sealing performance grade analysis network are generated, and then the two are combined to form a sealing performance evaluation model, which can achieve a comprehensive and accurate evaluation of the resin tank sealing performance.

[0086] In summary, the resin tank sealing detection method provided in the embodiment of the present application has the following technical effects: 1. By placing pressure sensor groups at key points of the resin tank, real-time pressure data streams are collected, and industrial cameras and displacement sensors are combined to monitor the deformation and displacement of the tank cover to achieve multi-dimensional sealing status perception. The mechanical sensor directly detects the pressure changes inside the tank body and dynamically reflects the mechanical response of the weak sealing area, while the visual and displacement data supplement the deformation information to form a pressure-deformation-displacement collaborative analysis. This method significantly improves the comprehensiveness of detection through multi-modal sensor fusion, avoids missed detection by a single sensor, and adapts to the sealing condition parameters of different application scenarios to ensure detection accuracy and scene adaptability.

[0087] 2. By defining the sealing boundary conditions and pressure test loads, finite element sealing simulation analysis is performed to accurately obtain stress distribution and deformation parameters. This process improves the accuracy and reliability of the test, can effectively identify key areas with stress concentration and significant deformation, ensures the scientific nature of the sealing test results, and enhances the safety of the resin tank and the structural stability.

[0088] 3. By integrating support vector machines and deep neural networks, a comprehensive sealing performance evaluation model was constructed to achieve a comprehensive and accurate evaluation of the sealing performance of the resin tank. This model combines the efficient classification capability of SVM and the detailed hierarchical analysis of DNN to improve the accuracy and reliability of the test results. The automated evaluation process optimizes the detection efficiency, enhances decision support and preventive maintenance capabilities, and ensures the long-term safety and stable operation of the resin tank.

[0089] Embodiment 2: Figure 2 The schematic diagram of the structure of the electronic device provided in the third embodiment of the present invention shows a block diagram of an exemplary electronic device suitable for implementing the implementation mode of the present invention. Figure 2 The electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention. Figure 2 As shown, the electronic device includes a processor 21, a memory 22, an input device 23 and an output device 24; the number of processors 21 in the electronic device can be one or more. Figure 2 Taking a processor 21 as an example, the processor 21, memory 22, input device 23 and output device 24 in the electronic device can be connected through a bus or other means. Figure 2 The example of connecting through bus is taken in the following.

[0090] In the third embodiment, the memory 22 is a computer-readable storage medium that can be used to store software programs, computer executable programs and modules, such as program instructions / modules corresponding to the resin tank sealing detection method in the embodiment of the present application. The processor 21 executes various functional applications and data processing of the computer device by running the software programs, instructions and modules stored in the memory 22, that is, realizing the above-mentioned resin tank sealing detection method.

[0091] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0092] Furthermore, the first or second mentioned above not only represents an order relationship, but also represents a specific concept, and / or refers to the selection of multiple elements individually or in whole. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application intends to include these modifications and variations.

Claims

1. A method for detecting the sealing performance of a resin tank, characterized in that: The method comprises: The target resin tank is fixedly placed on a sealing detection platform, and key point analysis is performed on the target resin tank to obtain a key point set of the resin tank; Acquire application scenario information of the target resin tank, perform usage condition analysis on the application scenario information, and determine resin tank sealing detection condition parameters; A pressure sensor group is arranged on the key point set of the resin tank, and a resin tank sealing state detection is performed based on the resin tank sealing detection condition parameters and the pressure sensor group to obtain a resin tank pressure sensing data stream set; An industrial camera and a displacement sensor are installed on the sealing detection platform, and the state of the resin tank cover of the target resin tank is detected by the industrial camera and the displacement sensor to obtain a deformation image set of the resin tank cover and a displacement data stream of the resin tank cover; The sealing performance of the target resin tank is evaluated based on the resin tank pressure sensing data stream set, the resin tank cover deformation image set and the resin tank cover displacement data stream to generate a resin tank sealing performance detection result.

2. The resin tank sealing detection method according to claim 1, characterized in that: The step of obtaining a key point set of the resin tank includes: Acquiring production design information of the target resin tank, performing three-dimensional modeling based on the production design information, and generating a resin tank space model; Determine the mesh type and mesh density based on the sealing analysis requirements and model complexity; Performing finite element division on the resin tank space model according to the mesh type and mesh density to obtain a finite element mesh model of the resin tank; Based on the resin tank finite element mesh model, a sealing simulation test and key point analysis are performed to obtain a resin tank key point set.

3. The method for detecting the sealing performance of a resin tank according to claim 2, wherein: The step of obtaining a key point set of the resin tank includes: Defining sealing boundary conditions and setting pressure test load information according to the working pressure range of the target resin tank; Applying the sealing boundary conditions and the pressure test load information to the resin tank finite element mesh model to perform sealing simulation analysis to obtain the stress distribution parameters of the resin tank; Based on the stress distribution parameters of the resin tank, stress area identification and deformation assessment are performed to obtain a resin tank stress concentration area set and a resin tank deformation parameter set; Based on the stress concentration area set of the resin tank and the deformation parameter set of the resin tank, a key point analysis is performed on the finite element mesh model of the resin tank to obtain the key point set of the resin tank.

4. The method for detecting the sealing performance of a resin tank according to claim 2, wherein: The generating of the resin tank sealing performance test result includes: Setting an abnormal value determination condition, and performing abnormal value identification on the resin tank pressure sensing data stream set and the resin tank cover displacement data stream according to the abnormal value determination condition to obtain an abnormal resin tank detection data set; Determining the abnormal data preprocessing process according to the abnormal value determination condition; Preprocessing the abnormal resin tank detection data set based on the abnormal data preprocessing process to obtain a standard resin tank pressure sensing data stream set and a standard resin tank cover displacement data stream; Based on the standard resin tank pressure sensing data stream set, the standard resin tank cover displacement data stream, and the resin tank cover deformation image set, the sealing performance of the target resin tank is evaluated to generate a resin tank sealing performance test result.

5. The method for detecting the sealing performance of a resin tank according to claim 4, characterized in that: The generating of the resin tank sealing performance test result includes: Analyze and process the standard resin tank pressure sensing data stream set and the standard resin tank cover displacement data stream to obtain the resin tank pressure distribution parameters and the resin tank cover displacement change parameters; Performing noise identification and filtering preprocessing on the resin can cover deformation image set to obtain a usable resin can cover deformation image set; Performing a can cover deformation analysis based on the available resin can cover deformation image set to obtain a resin can cover deformation parameter; The sealing performance of the target resin tank is evaluated based on the pressure distribution parameters of the resin tank, the displacement change parameters of the resin tank cover, and the deformation parameters of the resin tank cover to obtain a sealing performance test result of the resin tank.

6. The method for detecting the sealing performance of a resin tank according to claim 5, wherein: The step of obtaining the deformation parameters of the resin tank cover comprises: Determining a preset resin tank cover image according to the production design information of the target resin tank; Building a feature twin network, performing feature extraction processing on the preset resin tank cover image and the available resin tank cover deformation image set based on the feature twin network, and outputting a preset resin tank cover feature set and a current resin tank cover feature set; Performing a comparison loss analysis on the current resin tank cover feature set based on the preset resin tank cover feature set to obtain resin tank cover loss feature data; Based on the resin tank cover loss characteristic data, a tank cover deformation analysis is performed to obtain the resin tank cover deformation parameters.

7. The method for detecting the sealing performance of a resin tank according to claim 5, wherein: The obtaining of the resin tank sealing performance test result comprises: Acquire a resin tank sealing monitoring database, use a support vector machine to perform identification training on the resin tank sealing monitoring database, and obtain a sealing performance standard discriminator; Using a deep neural network to perform identification training on the resin tank sealing monitoring database to obtain a sealing performance level analysis network; The sealing performance standard discriminator and the sealing performance grade analysis network are combined and connected to generate a resin tank sealing performance evaluation model; Based on the resin tank sealing performance evaluation model, the pressure distribution parameters of the resin tank, the displacement change parameters of the resin tank cover, and the deformation parameters of the resin tank cover are evaluated for sealing performance, so as to obtain a resin tank sealing performance test result of the target resin tank.

8. An electronic device, characterized in that: The electronic device comprises: processor; a memory for storing instructions executable by the processor; Wherein, the processor is used to execute the resin tank sealing detection method described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, so that the computer program is used to execute the resin tank sealing detection method according to any one of claims 1 to 7.

Images