Method for adjusting correction parameters in real time in process of correcting welding deformation through induction heating

Abstract

The invention relates to the technical field of metal welding, in particular to a method for adjusting correction parameters in real time in the process of correcting welding deformation through induction heating. Comprising the steps of determining the correction amount of a welding part, establishing a database, designing a correction scheme, performing segmented correction, storing segmented correction data in the database, designing a next segment of correction scheme, continuing to store the correction data in the database, repeating the operation until all machining is completed, and the like. According to the method, the welding process is adjusted by detecting the welding deformation data in real time in the induction heating correction welding deformation machining process, the induction heating construction scheme of current correction construction can be continuously obtained according to the real-time deformation amount of the deformation structure, the correction procedure times are effectively reduced, and the machining precision is improved.

Description

Technical Field

[0001] This invention relates to the field of metal welding technology, and in particular to a method for real-time adjustment of correction parameters during induction heating to correct welding deformation. Background Technology

[0002] In shipbuilding, welding deformation affects the construction precision of the hull structure, threatening its stability and safety. Residual welding stress affects the hull's strength and material toughness, reducing its resistance to brittle fracture and corrosion. Both can lead to hull plate instability, ultimately impacting the overall quality of the ship's construction. Currently, due to factors such as equipment costs and worker skill levels, flame straightening methods in China primarily involve shipyard workers using combustible gases to heat specific areas of the deformed hull structure, generating a counter-deformation to offset the welding deformation. However, flame straightening methods heavily rely on operator experience and judgment, generally suffering from low efficiency, high labor intensity, and unstable production quality, making them increasingly unsuitable for modern shipbuilding technology. Summary of the Invention

[0003] To address the aforementioned issues, this patent relates to a heating scheme design method that considers real-time deformation during induction heating correction of welding deformation. The correction process has low dependence on worker operating skills and experience, high process controllability, and simple heat source parameters. By continuously and timely adjusting the equipment parameters during processing, precise control of deformation correction of plates of different thicknesses can be achieved, effectively solving the above problems.

[0004] The real-time adjustment method of correction parameters during the induction heating correction of welding deformation of the present invention includes the following steps: 1) Constructing a three-dimensional model of the welded deformed component using a 3D scanning method, and extracting the size specifications and deformation correction amount of the welded component; 2) Import the dimensional model of the deformable component into ANSYS finite element software, and based on the software simulation calculation, preliminarily determine the current welding deformation correction construction plan, and store the plan in the database; 3) Construct a construction model for the induction heating correction scheme based on a BP neural network model optimized by a genetic algorithm, and train it to be accurate based on training samples to determine the current deformation parameters such as heating wire length, moving speed, heating height, coil current, voltage and frequency, and design the correction construction scheme A0. 4) Divide the current heating wire into N equal parts according to its length, and number them sequentially as N1, N2, ..., N n Based on the point cloud data of the heating surface near the N1 heating line, the attitude path M1 of the induction heating coil moving parallel to the N1 path is designed. Based on the coil heating path M1 and the correction construction plan A0, the deformed structure is corrected. 5) When the induction coil moves to the end of M1, rescan the three-dimensional model of the currently welded deformed component, output the amount to be corrected after the current stage correction, and store the correction data of stage N1 in the database. 6) Set the current deformation state to be corrected as the input quantity. Based on the relevant training samples, use the neural network intelligent model to calculate the induction heating correction construction scheme A1 of the remaining N-1 heating wires of the current deformed component, including the moving speed, heating height, coil current, voltage and frequency parameters. At the same time, select the point cloud data of the heating wire attachment in the N2 stage, and design the attitude path M2 of the coil moving parallel to the N2 path based on this data. 7) Perform induction correction on segment N2 according to coil movement attitude path M2 and correction construction plan A1, and store the construction plan and deformation results after construction in the database; 8) Repeat the previous steps until all the heating wires that have been divided into N equal parts have been processed.

[0005] Specifically, in step 1), a laser 3D scanning method is used to collect real-time data on the overall deformation and geometric dimensions of the current hull welded components. This data is then compared with the designed geometric model to determine whether the current welded components meet the shipbuilding precision requirements. To ensure the accuracy and completeness of the data, the laser 3D scanning equipment must operate according to the preset scanning path and parameters, and the scanning environment should be kept stable to avoid interference from external factors. The collected data will undergo preliminary processing using professional data processing software to remove noise and outliers, thereby improving data quality. The processed data will then be precisely compared with the designed geometric model. If the comparison results show that the deformation and geometric dimensions of the current welded components exceed the shipbuilding precision requirements, the component is deemed unqualified and requires subsequent correction. The deformation data measured in this step can represent either the initial post-weld deformation state of the current structural component or the deformation distribution state after correction. The coordinate system of the real-time deformation point cloud data is referenced to the coordinate system of the scanning equipment.

[0006] Specifically, in step 2), the accuracy of the correction construction scheme is determined based on the electromagnetic-thermal-structural multiphysics conversion method of ANSYS. Point cloud data obtained from 3D point cloud coordinate identification is imported into the ANSYS simulation software, and a deformation correction simulation model is constructed based on the thermo-elastic-plastic finite element theory. A preliminary correction scheme for the current deformed structure is obtained, and the process parameters included in this scheme are stored in a sample database. Historical correction schemes similar to the current deformed structure are retrieved from the sample database. Through comparative analysis and combined with an expert experience knowledge base, the preliminary correction scheme is optimized and adjusted to ensure that it more accurately meets the correction requirements of the current deformed structure. The optimized correction scheme will clearly define key process parameters such as the specific heating area, heating temperature, and heating time, providing precise guidance for subsequent correction construction. This construction scheme is the first processing scheme, and all simulation calculation results in this step and the correction construction measurement results in subsequent steps are samples from the database, continuously expanding the data sample.

[0007] Furthermore, in step 4), based on the currently determined electromagnetic induction heating path coordinates, the position and orientation of the induction coil during the heating process are calculated to ensure that the working surface of the heating terminal remains parallel and at a constant height to the heating surface of the current weldment during the heating process. During heating, the position and orientation information of the induction coil is monitored in real time and compared with the preset ideal position and orientation. Once a deviation in position and orientation is detected, an automatic adjustment mechanism is immediately activated. A servo control system precisely controls the movement of the induction coil, quickly restoring it to the correct position and orientation, ensuring the uniformity and stability of heating, thereby effectively improving the correction effect and preventing deformation or other quality problems in the weldment due to uneven heating.

[0008] Furthermore, in the heating steps, each time a heating scheme is obtained, the heating scheme must be imported into the sample database, and special parameters must be used to mark the heating stage to which the sample belongs, in order to improve the sample quality of the neural network database.

[0009] Furthermore, in steps 2) to 6), during each stage of the correction process, the coil's movement attitude is recalculated based on the point cloud coordinates of the steel plate near the current heating wire. This ensures that the coil remains parallel and equidistant from the heating surface of the welded deformation structure throughout the heating process. An advanced algorithm model is introduced to calculate the coil's movement attitude. This model comprehensively considers factors such as the complex shape, degree of deformation, and material properties of the steel plate near the heating wire, achieving more accurate attitude calculations. Simultaneously, to ensure the accuracy and reliability of the calculation results, the calculated coil movement attitude is repeatedly verified and optimized. The effect is evaluated through simulated heating processes, and adjustments are made based on the evaluation results. In addition, a real-time feedback mechanism is established during the correction process to monitor the actual coil movement attitude and compare it with the calculated theoretical attitude. This allows for timely detection and correction of potential deviations, ensuring the smooth progress of the entire correction process and a significant improvement in the correction effect.

[0010] This invention uses real-time detection of welding deformation data during the induction heating correction process to adjust the welding process. It can continuously derive the induction heating construction plan for the current correction construction based on the real-time deformation of the deformed structure, effectively reducing the number of correction steps and improving processing accuracy.

[0011] This invention calculates the motion posture of the induction heating coil terminal during movement based on real-time deformation measurement data, making the processing path more precise and effectively improving the quality of deformation correction. By collecting the shape of the welded deformed part during the deformation correction process in real time, and based on the deformation result after correction, a neural network intelligent algorithm is used to continuously design and update the correction scheme. Furthermore, the correction scheme continuously expands the database samples, effectively improving the welding deformation correction effect and increasing the efficiency of correction construction. Attached Figure Description

[0012] Figure 1 This is a flowchart of the method for real-time adjustment of correction parameters during the induction heating correction of welding deformation according to the present invention.

[0013] Figure 2 This is a schematic diagram illustrating the overall principle of the T-joint deformation correction process after welding in the example of this invention.

[0014] Figure 3 This is a schematic diagram of the measurement points for extracting 3D scanning data of the T-joint deformation after welding in an example of the present invention.

[0015] Figure 4 This is a curve showing the deformation of the measurement points after the construction of each section of the heating line in the example of this invention. Detailed Implementation

[0016] The present invention will be described below with reference to examples. These examples are only used to explain the present invention and are not intended to limit the scope of the present invention.

[0017] A method for real-time adjustment of correction parameters during induction heating to correct welding deformation, such as... Figure 1 As shown, it includes the following steps: a) Real-time deformation data of the hull welding deformation structure is collected based on the 3D scanning measurement method. A three-dimensional model of the current deformed structure is established based on the measured point cloud data, and the size specifications and deformation correction amount of the welded parts are extracted. b) Import the three-dimensional model of the deformable component into ANSYS finite element software, and analyze the influence of different induction heating correction process parameters on the amount of deformation based on the thermo-elastic-plastic finite element theory; according to the requirements of the ship hull construction accuracy standard for the amount of deformation, preliminarily determine the current welding deformation correction construction scheme, and store the scheme in the sample database; c) Based on the sample database, a prediction model for the induction heating correction deformation construction scheme based on the BP neural network model algorithm is constructed, and the model parameters are optimized based on the genetic algorithm, and finally trained to be accurate; then, the optimal correction heating wire length L, moving speed, heating height, coil current, voltage and frequency parameters of the current deformed component are determined by the prediction model, and the correction construction scheme A0 is designed. d) Divide the current straightening heating wire into N equal parts according to its length. When the induction coil moves to the end of the first segment, rescan the three-dimensional model of the current welded deformed component, output the amount to be straightened after the current stage of straightening, and store the data in the database. e) In step (d), when performing induction heating correction on the deformed structure, a 3D scanning device is required to measure the coordinate data of the working surface of the induction coil and the target surface coordinate data of the corresponding deformed steel plate in real time, and match the vertical coordinate distance between the two. When the vertical distance between the two coordinate surfaces is not uniform, the position and attitude of the induction coil need to be adjusted in real time according to the difference. f) After step (d) is completed, the deformation state of the current component to be corrected, the correction construction parameters that have been completed, and the remaining length of the correction heating wire [L*((N-1) / N)] are set as input quantities. Based on the sample database, a neural network intelligent model is used to construct an intelligent prediction model of the induction heating correction welding deformation construction scheme under the current heating state. The model is used to calculate the induction heating correction construction scheme A1 of the remaining N-1 heating wires of the current deformed component, including the moving speed, heating height, coil current, voltage and frequency parameters. g) Perform induction correction construction on the second heating line according to the correction model output scheme in step (e), and store the construction scheme and the deformation results after construction in the sample database. h) In step (g), a 3D scanning device is also required to measure the vertical distance distribution between the working surface of the induction coil and the target surface of the correction steel plate in real time, and adjust the position and attitude of the induction coil in real time according to the difference. i) Repeat steps (c) to (h) until all heating wires currently divided into N equal parts have been processed. After all processing is completed, determine whether the current structural deformation meets the shipbuilding accuracy requirements. If the structure needs to be corrected again, proceed to step (b).

[0018] It should be noted that: in step (a), the deformation data of the hull welding structure measured by the laser 3D scanning method needs to be reconstructed in three dimensions to transform the set of three-dimensional coordinate points into a visualized three-dimensional structure and obtain the three-dimensional coordinates of the current outer plate shape point cloud; in step (b), the magnetic-thermal-mechanical-structural coupling method is used to simulate the induction heating correction deformation construction process to ensure the fidelity of the deformation correction process and obtain a relatively accurate initial correction construction plan; in step (c), the induction heating correction deformation construction plan prediction model uses improved artificial neural networks and other intelligent algorithms as the preferred algorithm; in step (d), based on the currently determined correction heating path coordinates, the initial position and attitude of the induction coil are calculated to ensure that the heating terminal remains perpendicular to the current outer plate; in step (e), based on the currently determined correction heating path coordinates, the position and attitude of the induction coil during the heating process are calculated to realize the real-time positioning of the induction coil on the correction processing trajectory, ensuring that the working surface of the coil always remains parallel and at a constant height with the target surface of the current plate during the heating process; after the above correction construction, if the residual deformation of the welded structure still needs to be corrected, this cycle can be repeated.

[0019] Based on the principles of the above method, a typical T-joint, the most common type of welded structural component in shipbuilding, is selected for demonstration. The base plate of the T-joint is 600*400*5mm, and the stiffening plate is 600*60*5mm. The overall principle of its post-weld deformation correction process is as follows: Figure 2 As shown, the specific correction steps are as follows: 1) For T-joints that deform after welding, 3D scanning equipment is used to measure the deformation of the current structure and obtain the deformation field distribution of the current welded joint. Figure 3 These are the deformation measurement points selected in the scanned cloud image data. The welding deformation data for each measurement point can be found in [reference needed]. Figure 4 According to the scanning results, the maximum vertical deformation of the bottom plate edge is 2.8mm, which does not meet the shipbuilding accuracy requirements and needs to be corrected. For this deformation, the shipyard construction personnel need to use their long-term work experience to perform flame calcination correction multiple times until the deformation of this structure meets the shipyard's accuracy requirements. 2) Import the 3D scan data of this T-joint into ANSYS finite element software. Based on the expert experience database of previous T-joint correction schemes, determine the correction construction scheme for the current deformed structure, and then conduct a simulation of the induction heating correction process. According to the plate size and maximum deformation, the preliminary correction scheme can be determined as follows (heating wire length, coil height, induction frequency, current, speed: 600 mm, 3 mm, 26.8 kHz, 2480 A, 5 mm / s). Using the preliminary correction scheme, a thermo-elastic-plastic simulation is performed, and the maximum deformation of the current T-profile after correction is found to be 1.2 mm, which meets the shipbuilding accuracy requirements for this size structure. The deformation correction scheme of this structure is then stored in the expert experience database. 3) Based on the current T-joint size, select training samples, use the current deformation data and heating wire length (600mm) as input, construct an induction heating correction scheme prediction model based on small sample BP neural network algorithm, and use genetic algorithm to optimize the model parameters to determine the correction scheme of the current welded joint as A0; 4) Based on the accuracy of the correction equipment parameters, the model prediction scheme data is rounded, and the actual processing parameters are set as A0. The coil height, induction frequency, current, and speed are 3 mm, 28.3 kHz, 2660 A, and 5 mm / s, respectively. 5) Divide the heating wire into 6 equal parts, select the point cloud data of the board near the first heating wire, calculate the normal coordinates of the heating area of ​​the first heating wire, and then make it perpendicular to the horizontal section of the coil according to the normal coordinate data, and make the vertical distance between the heating wire and the coil 3mm. Thus, the coil movement posture Z1 of the first heating wire can be determined. 6) Using posture Z1 and correction scheme A0, with parameters such as coil height, induction frequency, current, and speed of 3 mm, 28.3 kHz, 2660 A, and 5 mm / s respectively, the first section of the heating wire of the T-shaped welding head is subjected to induction heating correction operation. 7) After the first heating wire completes heating, rescan the deformation distribution data of the current welded joint and store this deformation data in the database. Figure 4 This refers to the deformation data of the measurement points after the construction of each section of the heating line. The "Line 1st" curve represents the deformation data of the measurement points after the construction of the first section of the heating line. According to the scanning measurement, the maximum deformation after correction at this stage is 2.5mm. 8) Using the current deformation data and heating wire length (500mm) as input, a new deformation correction scheme prediction model is constructed based on the database and neural network algorithm. The correction construction scheme A1 for the remaining 500mm heating wire is calculated. The scheme after rounding is as follows: coil height, induction frequency, current, and speed are 3 mm, 32.1 kHz, 2680 A, and 5 mm / s, respectively. 9) Select the point cloud data of the second heating wire attachment, calculate the normal coordinates of the heating area of ​​the second plate, and then make it perpendicular to the horizontal section of the coil according to the normal coordinate data, and make the vertical distance between the heating wire and the coil 3mm. Thus, the coil movement posture Z2 of the second heating wire can be determined. 10) Based on the coil movement path Z2 and the correction scheme A1, with the coil height, induction frequency, current, and speed set to 3 mm, 32.1 kHz, 2680 A, and 5 mm / s respectively, perform induction correction on the second heating wire section; after the second heating wire section is heated, scan the deformation data of this T-shaped welded joint. Figure 4 The "Line 2st" curve represents the deformation data of the measurement points after the construction of the second heating line. Based on the measurement results, the maximum deformation after heating and correction of the second section is 2.1 mm. The construction plan and deformation results of this second heating line are stored in the database. 11) Repeat the operation based on the calculation process of the second heating wire until the calculation of the correction scheme, coil correction attitude coordinates, and correction construction of the subsequent third to sixth heating wires are completed, and the data of each stage is stored in the database. 12) After the sixth heating wire is processed, scan and measure the deformation data of this T-joint. The final deformation data of this T-welded joint after correction will be used in the process. Figure 4 According to the measurement results, the maximum deformation of the plate edge after correction is 0.9mm, which fully meets the shipyard's processing accuracy requirements.

[0020] The above embodiments illustrate in detail a design method for a correction scheme that considers real-time deformation during induction heating to correct welding deformation, as claimed in this invention. This method provides real-time correction parameter adjustment and construction coil positioning basis for the automatic processing technology of induction heating to correct welding deformation considering real-time welding deformation through a real-time measurement method of deformation data of hull welding components based on 3D scanning, an intelligent prediction method for induction heating to correct deformation construction scheme considering real-time deformation, and a real-time attitude control method for the induction coil movement path.

[0021] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for real-time adjustment of correction parameters during induction heating to correct welding deformation, characterized in that, Includes the following steps: 1) A three-dimensional model of the welded deformed component was constructed using a 3D scanning method, and the dimensions and deformation correction amount of the welded component were extracted; 2) Import the dimensional model of the deformable component into ANSYS finite element software, and based on the software simulation calculation, preliminarily determine the current welding deformation correction construction plan, and store the plan in the database; 3) Construct a construction model for the induction heating correction scheme based on a BP neural network model optimized by a genetic algorithm, and train it to be accurate based on training samples to determine the current deformation parameters such as heating wire length, moving speed, heating height, coil current, voltage and frequency, and design the correction construction scheme A0. 4) Divide the current heating wire into N equal parts according to its length, and number them sequentially as N1, N2, ..., N n Based on the point cloud data of the heating surface near the N1 heating line, the attitude path M1 of the induction heating coil moving parallel to the N1 path is designed. Based on the coil heating path M1 and the correction construction plan A0, the deformed structure is corrected. 5) When the induction coil moves to the end of M1, rescan the three-dimensional model of the currently welded deformed component, output the amount to be corrected after the current stage correction, and store the correction data of stage N1 in the database. 6) Set the current deformation state to be corrected as the input quantity. Based on the relevant training samples, use the neural network intelligent model to calculate the induction heating correction construction scheme A1 of the remaining N-1 heating wires of the current deformed component. Select the point cloud data of the heating wire attachment in stage N2 and design the attitude path M2 of the coil moving parallel to the N2 path based on this data. 7) Perform induction correction on segment N2 according to coil movement attitude path M2 and correction construction plan A1, and store the construction plan and deformation results after construction in the database; 8) Repeat the previous steps until all the heating wires that have been divided into N equal parts have been processed.

2. The method according to claim 1, characterized in that, In step 1), the laser 3D scanning method is used to collect the overall deformation data and geometric dimension data of the current hull welding components in real time, and compare them with the designed geometric model to determine whether the current welding components meet the shipbuilding precision requirements.

3. The method of claim 1, wherein, In step 2), the accuracy of the correction construction scheme is determined based on the electromagnetic-thermal-structural multiphysics conversion method of ANSYS.

4. The method of claim 3, wherein, The point cloud data obtained from the three-dimensional point cloud coordinate identification is imported into the ANSYS simulation software, and a deformation correction simulation model is constructed based on the thermo-elastic-plastic finite element theory. The correction scheme for the current deformed structure is initially obtained, and the process parameters are stored in the sample database.

5. The method of claim 1, wherein, In step 4), based on the currently determined electromagnetic induction heating path coordinates, the position and attitude of the induction coil during the heating process are calculated to ensure that the working surface of the heating terminal remains parallel and at a constant height to the heating surface of the current weldment during the heating process.

6. The method of claim 1, wherein, In the heating process, each time a heating scheme is obtained, it must be imported into the sample database, and a special parameter must be used to mark the heating stage to which the sample belongs.

7. The method of claim 1, wherein, In steps 2) to 6), during each stage of correction construction, the movement posture of the coil is recalculated based on the point cloud coordinates of the steel plate in the area near the current heating line, so that the coil remains parallel and equidistant from the heating surface of the welded deformed structure during the heating process.

Images