An intermediate frequency pipe bending process management system and management method

Abstract

An intermediate frequency pipe bending process management system and management method. In this method, first, a finite element model of intermediate frequency pipe bending is established. Based on the finite element model, simulated pipe bending is completed to obtain simulation experiment data. After analysis and optimization, pipe bending process parameters are obtained and used for production trial. Then, during the trial production process, the reference state of the pipe bending is evaluated to obtain experimental data of the pipe bending in healthy and abnormal states. Using the simulation experiment data and trial production experiment data as samples, the prediction model is strengthened. Finally, in the actual pipe bending process, real-time data is obtained and predicted using the trained prediction model to determine whether abnormalities occur in the pipe bending process. At the same time, based on the abnormal state, the pipe bending parameters are optimized until the real-time data is predicted to be normal by the model. The present invention implements intelligent management of intermediate frequency pipe bending, simplifies the process R & D process, realizes online monitoring of the bending process and automatic optimization in case of abnormalities, and improves the quality and production efficiency of pipe bending products.

Description

Technical Field

[0001] The present invention relates to the field of medium-frequency pipe bending, and in particular to a medium-frequency pipe bending process management system and a management method. Background Art

[0002] Medium-frequency pipe bending is to place a medium-frequency induction coil around the pipe, and rely on the medium-frequency induction current to locally heat the pipe to the required high temperature, and then bend the heated part to obtain the required bent pipe fitting. Medium-frequency pipe bending is one of the most advanced process methods in pipe bending processing recognized internationally today, and has been widely used in pipe prefabrication in engineering fields such as electric power, petroleum, chemical industry, navigation, and nuclear industry.

[0003] Medium-frequency pipe bending equipment is the guarantee for realizing the pipe bending process. At present, the process control of medium-frequency pipe bending mostly focuses on the accurate and stable control of process parameters such as heating power, pushing bending speed, and cooling method. However, medium-frequency pipe bending is a complex plastic deformation process in which heating, deformation, and cooling occur simultaneously. During the pipe bending process, due to process conditions or improper operation, various degrees of defects may occur in the pipe fittings, and the occurrence of these defects will directly affect the final appearance quality, safety, and reliability of the product. Therefore, in order to prevent or reduce the generation of pipe bending defects and obtain satisfactory pipe bending quality, it is particularly important to monitor the pipe bending process online and take corresponding countermeasures in a timely manner during the pipe bending process.

[0004] Therefore, a medium-frequency pipe bending process management system and a management method are needed, which can realize the online monitoring of the pipe bending process and automatic optimization under abnormal conditions, improve the quality and production efficiency of the pipe bending products, and solve the problems existing in the above prior art. Summary of the Invention

[0005] The present invention provides a medium-frequency pipe bending process management system and a management method in view of the above deficiencies existing in the prior art.

[0006] The object of the present invention is achieved through the following technical solutions:

[0007] A medium-frequency pipe bending process management system, characterized in that it comprises: a finite element analysis module configured to configure a three-dimensional model and a finite element model of medium-frequency induction local heating pipe bending;

[0008] A data acquisition module configured to evaluate the reference state of the pipe bending during the trial production process and obtain the experimental data of the pipe bending in healthy and abnormal states;

[0009] A data transmission module configured to transmit the simulation experimental data and the experimental data during the trial production process;

[0010] The model generation module is configured to analyze the relationship between the reference state of the bent pipe and the real-time data during the actual bending process, and use the simulation experiment data, the experiment data in the healthy state, and the experiment data in the abnormal state as samples, which are divided into three parts for strengthening the prediction model;

[0011] The parameter tuning module is configured to optimize the bending parameters of the bent pipe based on the abnormal state until the real-time data is predicted to be normal by the model;

[0012] The data processing module is configured to store the finite element analysis model, the simulation experiment data, the trial production experiment data, and the prediction model;

[0013] The finite element analysis module, the data acquisition module, the data transmission module, the model generation module, and the parameter tuning module are respectively connected to the data processing module.

[0014] The management method of the medium-frequency bent pipe process management system includes the following steps:

[0015] Step 1: Establish a three-dimensional model and a finite element model of medium-frequency induction local heating of the bent pipe;

[0016] Step 2: Complete the simulation of the bent pipe based on the finite element model to obtain simulation experiment data;

[0017] Step 3: Analyze the simulation experiment data and optimize to obtain the bending process parameters;

[0018] Step 4: Apply the optimized process parameters to production trial;

[0019] Step 5: Evaluate the reference state of the bent pipe during the trial production process;

[0020] Step 6: Obtain the experiment data of the bent pipe in the healthy state and the experiment data of the bent pipe in the abnormal state;

[0021] Step 7: Use the simulation experiment data, the experiment data in the healthy state, and the experiment data in the abnormal state as samples, which are divided into three parts for strengthening the prediction model;

[0022] Step 8: Obtain the real-time data during the actual bending process, and predict the real-time data using the prediction model trained in Step 7;

[0023] Step 9: If the result of the model prediction of the real-time data is normal, it indicates that the bending state of the bent pipe is normal; if the result of the model prediction of the real-time data is abnormal, it indicates that the bending state of the bent pipe is abnormal;

[0024] Step 10: Optimize the bending parameters of the bent pipe based on the abnormal state until the real-time data is predicted to be normal by the model.

[0025] The management method of the medium-frequency pipe bending process management system described above. The simulation experiment data in step (2) are multiple spatial geometric feature points and multiple temperature field feature points of the pipe bending changing with time, as well as the operating stresses of the clamping and rotating device and the guide roller device changing with time.

[0026] The management method of the medium-frequency pipe bending process management system described above. The specific steps for obtaining the simulation experiment data in step (2) include:

[0027] (1) Perform digital modeling based on the pipe material, pipe pushing device, medium-frequency induction local heating device, clamping and rotating device, and guide roller device to obtain a three-dimensional model of medium-frequency induction local heating pipe bending;

[0028] (2) Complete finite element analysis based on the three-dimensional model of medium-frequency induction local heating pipe bending, and through processing, obtain a three-dimensional space model and a three-dimensional temperature field model of the pipe bending changing with time, as well as the operating stresses of the clamping and rotating device and the guide roller device changing with time;

[0029] (3) Extract multiple feature points from the three-dimensional space model and the three-dimensional temperature field model. The relevant feature points can fully characterize the three-dimensional space model and the three-dimensional temperature field model of the pipe bending.

[0030] The management method of the medium-frequency pipe bending process management system described above. The specific steps for evaluating the reference state of the pipe bending during the trial production process in step (5) include:

[0031] (1) Obtain data such as the surface state, wall thickness, outer diameter, bending angle, and bending radius of the pipe bending through means such as non-destructive testing and dimensional measurement;

[0032] (2) Obtain real-time quality data such as angle deviation, thinning rate, and out-of-roundness according to the preset conversion formula, and compare them with the standard requirements to evaluate the reference state of the pipe bending as a healthy state or an abnormal state.

[0033] The management method of the medium-frequency pipe bending process management system described above. The experimental data of the pipe bending in the healthy state and the experimental data of the pipe bending in the abnormal state in step (6) are multiple spatial geometric feature points and multiple temperature field feature points of the pipe bending changing with time, as well as the operating stresses of the clamping and rotating device and the guide roller device changing with time.

[0034] The management method of the medium-frequency pipe bending process management system described above. The specific steps for obtaining the experimental data of the pipe bending in the healthy state and the experimental data of the pipe bending in the abnormal state in step (6) include:

[0035] (1) Set up more than two high-speed cameras with thermal imaging functions to obtain high-definition images of the pipe bending;

[0036] (2)Based on the spatial positions of the elbow pipe, pipe bender, camera, etc., combined with high-definition images, perform three-dimensional construction to establish the mapping relationship between the image coordinates and the spatial coordinates of the survey area;

[0037] (3)Obtain the three-dimensional spatial model and three-dimensional temperature field model of the elbow pipe through the feature matching algorithm;

[0038] (4)Extract multiple feature points from the three-dimensional spatial model and three-dimensional temperature field model, and the relevant feature points can fully represent the three-dimensional spatial model and three-dimensional temperature field model of the elbow pipe;

[0039] (5)Add stress acquisition devices to the chuck of the clamping and rotating device and the guide wheel of the guide roller device respectively to obtain the operating stresses of the clamping and rotating device and the guide roller device under the operating state.

[0040] For the management method of the intermediate frequency elbow pipe process management system described above, the multiple feature points are simultaneously assisted and corrected by using the laser ranging and contact temperature measurement methods, and an adaptive algorithm is added to reduce the interference of light and water vapor changes on the feature point recognition until the recognized feature points meet the tracking requirements.

[0041] For the management method of the intermediate frequency elbow pipe process management system described above, the prediction model aims at multi-feature point recognition, and the weight of the experimental data in the trial production process is greater than the weight of the simulation experimental data during the model training process.

[0042] For the management method of the intermediate frequency elbow pipe process management system described above, the specific steps for optimizing the elbow pipe bending parameters in step ten include:

[0043] (1)Perform spline fitting on the data obtained by finite element analysis of the intermediate frequency induction local heating elbow pipe;

[0044] (2)Obtain the influence laws of process parameters such as outer wall temperature, temperature gradient, thermal deformation width, and pushing speed on the quality data;

[0045] (3)Perform process optimization according to the influence laws of relevant parameters on the bending quality of the elbow pipe until the real-time data and the model prediction results are normal. Beneficial effects

[0046] 1. The present invention establishes a finite element model of the intermediate frequency elbow pipe, completes the simulation elbow pipe based on the finite element model, obtains the simulation experimental data, analyzes and optimizes to obtain the elbow pipe process parameters and uses them for production trial, simplifies the process R & D process, shortens the product process R & D cycle, and reduces the production cost of the enterprise;

[0047] 2. The present invention obtains real-time data during the actual bending process through the data acquisition module, and stores it through the data transmission module and data storage module within the system, achieving online monitoring to a great extent, reducing the frequency of manual inspections, and simultaneously realizing the full-information playback of the pipe bending process.

[0048] 3. For the real-time data obtained during the actual bending process, the present invention uses the trained prediction model to make predictions to determine whether abnormalities occur during the pipe bending process, and records and feeds back in real time through the data processing module, thereby realizing the intelligent management of medium-frequency pipe bending.

[0049] 4. Through the analysis and prediction of the real-time data of the pipe bending, combined with the relevant influence laws obtained from the finite element analysis, corresponding countermeasures are taken in a timely manner during the pipe bending process to prevent or reduce the generation of bending defects, improving the quality and production efficiency of the pipe bending products. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic diagram of the modules of the present invention

[0051] Figure 2 It is a schematic diagram of the steps of the simulated pipe bending of the present invention;

[0052] Figure 3 It is a schematic diagram of the steps of the model prediction and process optimization of the present invention;

[0053] Reference numerals: 101, finite element analysis module; 102, data acquisition module; 103, data transmission module; 104, model generation module; 105, parameter optimization module; 106, data processing module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] Referring to Figure 1 , a medium-frequency pipe bending process management system, which comprises:

[0055] The finite element analysis module 101 is configured to configure a three-dimensional model and a finite element model of medium-frequency induction local heating pipe bending, complete the simulated pipe bending based on the finite element model, obtain the simulated experimental data, and analyze and optimize to obtain the pipe bending process parameters for production trial;

[0056] The data acquisition module 102 is configured to evaluate the reference state of the pipe bending during the trial production process and obtain the experimental data of the pipe bending in healthy and abnormal states;

[0057] The data transmission module 103 is configured to transmit the simulated experimental data and the experimental data during the trial production process;

[0058] The model generation module 104 is configured to analyze the relationship between the reference state of the bent pipe and the real-time data during the actual bending process, and use the simulation experiment data, the experiment data in the healthy state, and the experiment data in the abnormal state as samples, which are divided into three parts (training set, test set, validation set) for strengthening the prediction model.

[0059] The parameter tuning module 105 is configured to optimize the bending parameters of the bent pipe based on the abnormal state until the real-time data and the model prediction result are normal.

[0060] The data processing module 106 is configured to store the finite element analysis model, the simulation experiment data, the trial production experiment data, and the prediction model.

[0061] The finite element analysis module, the data acquisition module, the data transmission module, the model generation module, and the parameter tuning module are respectively connected to the data processing module.

[0062] The implementation principle of the present invention is as follows: First, a finite element model of the intermediate frequency bent pipe is established through the finite element analysis module. Based on the finite element model, a simulated bent pipe is completed to obtain simulation experiment data. The bending process parameters are analyzed and optimized and used for production trial. Then, during the trial production process, the reference state of the bent pipe is evaluated through the data acquisition module to obtain the experiment data of the bent pipe in the healthy and abnormal states. The prediction model is strengthened through the model generation module using the simulation experiment data and the trial production experiment data as samples. Finally, during the actual bending process, real-time data is obtained through the data acquisition module and predicted using the trained prediction model to determine whether an abnormality occurs during the bending process. At the same time, based on the abnormal state, the parameter tuning module is used to optimize the bending parameters of the bent pipe until the real-time data and the model prediction result are normal. The data transmission module plays a role in data transfer between each module, and the data processing module is responsible for managing the data processing tasks within the system. The present invention implements intelligent management of the intermediate frequency bent pipe, simplifies the process research and development process, realizes online monitoring of the bending process and automatic optimization in case of abnormalities, and improves the quality and production efficiency of the bent pipe products.

[0063] Referring to Figures 2 - 3 , the management method of the intermediate frequency bent pipe process management system described above, the method includes the following steps:

[0064] Step 1: Establish a three-dimensional model and a finite element model of the intermediate frequency induction local heating bent pipe.

[0065] As one of the implementation manners, the specific steps of establishing a three-dimensional model and a finite element model of the intermediate frequency induction local heating bent pipe include:

[0066] Digital modeling is carried out based on the geometric characteristics of the pipe, the pipe pushing device, the intermediate frequency induction local heating device, the clamping and rotating device, and the guide roller device, that is, the modeling design of components and parts is completed using an interactive drawing system CAD drawing software (such as Pro / ENGINEER, Unigraphics, SolidEdge, SolidWorks, IDEAS, Bentley, AutoCAD, etc.), and a three-dimensional model of the intermediate frequency induction local heating elbow pipe is obtained.

[0067] Based on the three-dimensional model of the intermediate frequency induction local heating elbow pipe, finite element analysis is completed. First, determine the geometric parameters and material properties of the elbow pipe, that is, clarify the geometric parameter characteristics such as the diameter, wall thickness, bending radius, and bending angle of the elbow pipe, and determine the physical properties such as the elastic modulus, yield strength, and Poisson's ratio of the elbow pipe material. Select a suitable finite element software for structural analysis (such as ANSYS, Abaqus, SolidWorks Simulation, etc.). Import the three-dimensional model created by the CAD drawing software into the finite element software, and select a suitable mesh type and size according to the geometric characteristics and calculation accuracy requirements of the elbow pipe, and assign the previously determined material properties to the model.

[0068] As another implementation method, directly use a finite element software integrated with CAD drawing modeling technology (such as ADINA, etc.) to complete the creation of the three-dimensional model and the finite element model within the software;

[0069] Step 2: Based on the finite element model, complete the simulation of the elbow pipe to obtain simulation experiment data;

[0070] The simulation experiment data are multiple spatial geometric feature points and multiple temperature field feature points of the elbow pipe changing with time, as well as the operating stress of the clamping and rotating device and the guide roller device changing with time. The specific steps to obtain the relevant multiple feature points and operating stress include:

[0071] According to the elbow pipe process, set appropriate boundary conditions, such as fixed constraints, symmetry constraints, etc., apply loads, simulate the temperature field and force field during intermediate frequency heating and bending, and at the same time collect the real-time data of the three-dimensional space and three-dimensional temperature field of the elbow pipe during heating and bending, as well as the operating stress of the clamping and rotating device and the guide roller device changing with time. Extract multiple feature points from the three-dimensional space model and the three-dimensional temperature field model. The relevant feature points can fully characterize the three-dimensional space model and the three-dimensional temperature field model of the elbow pipe.

[0072] Step 3: Analyze the simulation experiment data and optimize to obtain the elbow pipe process parameters;

[0073] According to the results of the simulation experiment data, adjust the process parameters of the bent pipe, such as the outer wall temperature, temperature gradient, thermal deformation width, pushing speed, etc. Repeat the simulation analysis to verify whether the optimized process parameters are effective. Finally, obtain the process parameters of the bent pipe. At the same time, perform spline fitting on the data obtained by finite element analysis of the intermediate frequency induction local heating bent pipe to obtain the influence laws of process parameters such as outer wall temperature, temperature gradient, thermal deformation width, and pushing speed on the quality data of the bent pipe process.

[0074] Step Four: Apply the optimized process parameters to production trial;

[0075] Apply the optimized process parameters to actual production trial, compare the simulation results and the actual trial results, evaluate the accuracy and reliability of the process parameters, and further adjust and optimize the process parameters according to the actual trial results. At the same time, continuously collect simulation experiment data during the process.

[0076] Step Five: Evaluate the reference state of the bent pipe during the trial;

[0077] Use non-destructive testing methods such as ultrasonic and magnetic particle testing to evaluate whether there are defects such as cracks and inclusions on the surface and inside of the bent pipe. Use an ultrasonic thickness gauge to perform wall thickness inspection on the outer arc surface, inner arc surface, and neutral layer center line of the bent pipe, and use measuring tools such as external calipers, snap gauges, and straight rulers to measure to obtain data such as the outer diameter, bending angle, and bending radius of the bent pipe.

[0078] Input the collected data into a computer or special software for processing. According to the preset conversion formula, calculate real-time quality data such as the bending angle deviation, thinning rate, and out-of-roundness of the bent pipe, and compare it with the standard requirements to evaluate whether the reference state of the bent pipe is a healthy state or an abnormal state.

[0079] Step Six: Obtain the experimental data of the bent pipe in the healthy state and the experimental data of the bent pipe in the abnormal state;

[0080] Set up more than two high-speed cameras with thermal imaging functions to obtain high-definition images of the bent pipe. According to the spatial positions of the bent pipe, the pipe bender, and the cameras, etc., perform 3D construction in combination with the high-definition images, establish the mapping relationship between the image coordinates and the spatial coordinates of the measurement area, and obtain the 3D spatial model and 3D temperature field model of the bent pipe through the feature matching algorithm. The relevant feature points can fully represent the 3D spatial model and 3D temperature field model of the bent pipe. Add stress acquisition devices to the chucks of the clamping and rotating device and the guide wheels of the guide roller device respectively to obtain the operating stresses of the clamping and rotating device and the guide roller device under the operating state. When the reference state of the bent pipe evaluated in step five is the healthy state, the relevant feature point data and operating stress are the experimental data of the bent pipe in the healthy state. Preferably, multiple feature points are simultaneously assisted and verified and corrected by methods such as laser ranging and contact temperature measurement, and an adaptive algorithm is added to reduce the interference of changes in light, water vapor, etc. on the feature point recognition until the recognized feature points meet the tracking requirements.

[0081] According to the relevant method steps in step six, when the reference state of the bent pipe evaluated in step five is the abnormal state, the relevant feature point data and operating stress are the experimental data of the bent pipe in the abnormal state. Preferably, multiple feature points are simultaneously assisted and verified and corrected by methods such as laser ranging and contact temperature measurement, and an adaptive algorithm is added to reduce the interference of changes in light, water vapor, etc. on the feature point recognition until the recognized feature points meet the tracking requirements;

[0082] Step seven: Use the simulation experiment data, the experimental data in the healthy state, and the experimental data in the abnormal state as samples, and divide them into three parts (training set, test set, validation set) for strengthening the prediction model;

[0083] First, collect and organize all available data, including simulation experiment data, experimental data in the healthy state, and experimental data in the abnormal state, and ensure that the data has been preprocessed, including steps such as cleaning, normalization, and label encoding, so that the model can better learn the features in the data. Subsequently, divide the data set into a training set (Training Set), a test set (Test Set), and a validation set (Validation Set) to effectively train and evaluate the performance of the strengthening prediction model. When dividing the data set, it is necessary to ensure that the time sequences of the training set, the validation set, and the test set do not overlap. Use the training set to train the model, and use the validation set to evaluate the performance of the model during the training process. According to the performance on the validation set, adjust the hyperparameters and structure of the model. When the performance of the model on the validation set reaches the best, use the test set to finally evaluate the model to obtain the performance of the model in actual applications.

[0084] At the same time, the prediction model aims at multi-feature point recognition, and the weight of the experimental data in the trial production process is greater than the weight of the simulation experiment data during the model training process.

[0085] Step Eight: Obtain the real-time data during the actual bending process and use the prediction model trained in Step Seven to predict the real-time data.

[0086] According to the relevant method steps in Step Six, obtain the real-time data during the actual bending process, that is, the real-time data of multiple spatial geometric feature points and multiple temperature field feature points of the bent pipe, as well as the operating stress of the clamping and rotating device and the guide roller device changing with time. First, preprocess the real-time data and input it into the prediction model trained in Step Seven for prediction. The model will generate a prediction result based on the input data, which includes possible quality problems during the bent pipe bending process.

[0087] Step Nine: If the result of predicting the real-time data by the model is normal, it indicates that the bending state of the bent pipe is normal; if the result of predicting the real-time data by the model is abnormal, it indicates that the bending state of the bent pipe is abnormal.

[0088] If the result of predicting the real-time data by the model is normal, this indeed indicates that under the current operating conditions and the bending state of the bent pipe, the state of the bent pipe is normal and no unexpected problems or deviations have occurred.

[0089] If the result of predicting the real-time data by the model is abnormal, this indeed indicates that under the current operating conditions and the bending state of the bent pipe, the state of the bent pipe is abnormal and certain problems or deviations have occurred during the bending process.

[0090] Step Ten: Based on the abnormal state, optimize the bending parameters of the bent pipe until the result of predicting the real-time data by the model is normal.

[0091] If the result of predicting the real-time data by the model is abnormal, combine the influence law of the relevant parameters obtained by finite element analysis in Step Six on the bending quality of the bent pipe, take corresponding countermeasures in a timely manner during the bending process of the bent pipe, and then execute Step Eight to obtain the real-time data during the actual bending process, and use the prediction model trained in Step Seven to predict the real-time data until the result of predicting the real-time data by the model is normal.

Claims

1. A medium-frequency pipe bending process management system, characterized in that: It consists of: A finite element analysis module, configured to establish a three-dimensional model and a finite element model of an intermediate frequency induction local heating elbow; complete a simulated elbow based on the finite element model to obtain simulated experimental data; analyze the simulated experimental data, optimize to obtain elbow processing parameters; and use the optimized processing parameters for production trial manufacturing; A data acquisition module, configured to evaluate the reference state of the elbow during the trial manufacturing process, obtain experimental data of the elbow in healthy and abnormal states, and generate trial manufacturing experimental data; A data transmission module, configured to transmit the simulated experimental data and the experimental data during the trial manufacturing process; A model generation module, configured to analyze the relationship between the reference state of the elbow and the real-time data during the actual bending process, and use the simulated experimental data, the experimental data in the healthy state, and the experimental data in the abnormal state as samples, which are divided into three parts for strengthening the prediction model; the simulated experimental data are multiple spatial geometric feature points and multiple temperature field feature points of the elbow changing with time, as well as the operating stresses of the clamping and rotating device and the guide roller device changing with time; the experimental data of the elbow in the healthy state and the experimental data of the elbow in the abnormal state are multiple spatial geometric feature points and multiple temperature field feature points of the elbow changing with time, as well as the operating stresses of the clamping and rotating device and the guide roller device changing with time; A parameter optimization module, configured to optimize the elbow bending parameters based on the abnormal state until the real-time data is predicted to be normal by the model; A data processing module, configured to store the finite element analysis model, the simulated experimental data, the trial manufacturing experimental data, and the prediction model; The finite element analysis module, the data acquisition module, the data transmission module, the model generation module, and the parameter optimization module are respectively connected to the data processing module.

2. The management method of an intermediate frequency pipe bending process management system according to claim 1, characterized in that: The method includes the following steps: Step 1: Establish a three-dimensional model and a finite element model of an intermediate frequency induction local heating elbow; Step 2: Complete a simulated elbow based on the finite element model to obtain simulated experimental data; Step 3: Analyze the simulated experimental data and optimize to obtain elbow processing parameters; Step 4: Use the optimized processing parameters for production trial manufacturing; Step 5: Evaluate the reference state of the elbow during the trial manufacturing process; Step 6: Obtain the experimental data of the elbow in the healthy state and the experimental data of the elbow in the abnormal state; Step 7: Use the simulated experimental data, the experimental data in the healthy state, and the experimental data in the abnormal state as samples, which are divided into three parts for strengthening the prediction model; Step 8: Obtain the real-time data during the actual bending process, and predict the real-time data using the prediction model trained in Step 7; Step 9: If the result of predicting the real-time data by the model is normal, it indicates that the bending state of the elbow is normal; if the result of predicting the real-time data by the model is abnormal, it indicates that the bending state of the elbow is abnormal; Step 10: Optimize the elbow bending parameters based on the abnormal state until the real-time data is predicted to be normal by the model.

3. The management method of an intermediate frequency pipe bending process management system according to claim 2, characterized in that: The specific steps for obtaining the simulated experimental data in Step 2 include: (1)Digitally model according to the pipe material, push tube device, intermediate frequency induction local heating device, clamping and rotating device, and guide roller device to obtain a three-dimensional model of the intermediate frequency induction local heating elbow. (2)Complete finite element analysis based on the three-dimensional model of the intermediate frequency induction local heating elbow, and through processing, obtain a three-dimensional space model and a three-dimensional temperature field model of the elbow changing with time, as well as the operating stress of the clamping and rotating device and the guide roller device changing with time. (3)Extract multiple feature points from the three-dimensional space model and the three-dimensional temperature field model, and the multiple feature points can fully represent the three-dimensional space model and the three-dimensional temperature field model of the elbow.

4. The management method of an intermediate frequency pipe bending process management system according to claim 2, characterized in that: The specific steps of evaluating the reference state of the elbow in the trial production process in step five include: (1)Obtain the surface state, wall thickness, outer diameter, bending angle, and bending radius data of the elbow through non-destructive testing and dimensional measurement means. (2)Obtain the real-time quality data of angle deviation, thinning rate, and out-of-roundness according to the preset conversion formula, and compare it with the standard requirements to evaluate the reference state of the elbow as a healthy state or an abnormal state.

5. The management method of an intermediate frequency pipe bending process management system according to claim 2, characterized in that: The specific steps of obtaining the experimental data of the elbow in the healthy state and the experimental data of the elbow in the abnormal state in step six include: (1)Set up more than two high-speed cameras with thermal imaging functions to obtain high-definition images of the elbow. (2)According to the spatial positions of the elbow, the pipe bender, and the cameras, combine the high-definition images for three-dimensional construction to establish the mapping relationship between the image coordinates and the spatial coordinates of the measurement area. (3)Obtain the three-dimensional space model and the three-dimensional temperature field model of the elbow through the feature matching algorithm. (4)Extract multiple feature points from the three-dimensional space model and the three-dimensional temperature field model, and the multiple feature points can fully represent the three-dimensional space model and the three-dimensional temperature field model of the elbow. (5)Add stress acquisition devices to the chucks of the clamping and rotating device and the guide wheels of the guide roller device respectively to obtain the operating stress of the clamping and rotating device and the guide roller device in the operating state.

6. The management method of an intermediate frequency pipe bending process management system according to claim 3, characterized in that: The multiple feature points are simultaneously assisted and corrected by using the laser ranging and contact temperature measurement methods, and an adaptive algorithm is added to reduce the interference of light and water vapor changes on the feature point recognition until the recognized feature points meet the tracking requirements.

7. The management method of an intermediate frequency pipe bending process management system according to claim 2, characterized in that: The prediction model aims at multi-feature point recognition, and the weight of the experimental data in the trial production process is greater than the weight of the simulation experimental data in the model training process.

8. The management method of an intermediate frequency pipe bending process management system according to claim 2, characterized in that: The specific steps of optimizing the bending parameters of the elbow in step ten include: (1)Perform spline fitting on the data obtained by finite element analysis of the intermediate frequency induction local heating elbow. (2)Obtain the influence laws of process parameters such as outer wall temperature, temperature gradient, thermal deformation width, and pushing speed on the quality data. (3)Conduct process optimization according to the influence laws of relevant parameters on the bending quality of the elbow until the real-time data and the model prediction results are normal.

Images