Dissimilar metal welded joint temperature field optimization control method and system

Abstract

The invention relates to the technical field of welding processing, and discloses a dissimilar metal welding joint temperature field optimization control method and system. Real-time temperature field scanning is conducted on the dissimilar metal welded joint, image segmentation processing is conducted on a temperature field image obtained through scanning, temperature data of a weld joint area, a heat affected area and a base metal area are extracted, and whether the temperature gradient exceeds a preset threshold value or not is judged; aiming at an area with non-uniform distribution of heat input quantity, analyzing the performance of the welding joint under the combined action of thermal stress and mechanical stress, obtaining the difference between residual stress distribution and plastic deformation capacity, and analyzing a stress concentration area by combining with the yield strength of the material; and a mapping relation between the welding process parameters and temperature field distribution is established, and a welding process parameter adjustment prediction model is obtained through historical welding process parameters and corresponding temperature field data training in combination with heat flux density distribution. The problems of uneven heat input, stress concentration and the like in dissimilar metal welding are effectively solved, and the welding quality and efficiency are improved.

Description

Technical Field

[0001] The invention relates to the technical field of welding processing, and in particular to a method and system for optimizing the temperature field of a dissimilar metal welding joint. Background Art

[0002] In the process of dissimilar metal welding, due to the significant differences in the thermophysical properties of different materials, the temperature field distribution of the welded joint is uneven. This unevenness will cause a series of problems: first, excessive temperature gradient will cause local stress concentration and increase the risk of welding cracks; second, the difference in thermal expansion coefficient will cause joint deformation and affect product precision; third, the inconsistency of cooling rate will cause uneven organization and reduce joint strength.

[0003] However, the current analysis methods for the temperature field of dissimilar metal welding joints are still imperfect, and it is difficult to accurately capture the transient change characteristics of the temperature field. At the same time, the existing optimization methods for process parameters such as heat source power and welding speed often rely on experience and lack theoretical guidance. In addition, the accuracy and real-time performance of temperature field control are insufficient, making it difficult to cope with complex welding conditions.

[0004] How to establish a systematic temperature field analysis method for dissimilar metal welding, how to develop a process parameter optimization algorithm based on the temperature field distribution law, and how to achieve high-precision real-time control of the temperature field are technical problems that need to be solved urgently. Summary of the invention

[0005] In view of this, the present invention provides a method and system for optimizing the temperature field control of dissimilar metal welding joints to solve the problems of uneven heat input, stress concentration, poor welding quality, etc. existing in dissimilar metal welding joints.

[0006] In a first aspect, the present invention provides a method for optimizing the temperature field of a dissimilar metal welding joint, comprising:

[0007] Scan the temperature field of dissimilar metal welding joints in real time to obtain temperature field images;

[0008] Perform image segmentation processing on the temperature field image to distinguish the weld area, heat-affected zone and base metal area, and extract the temperature data of each area;

[0009] Determine whether the temperature gradient between the temperature data of each zone exceeds the preset threshold; perform numerical simulation on the temperature data when the temperature gradient exceeds the threshold, establish a three-dimensional temperature field model of the dissimilar metal welding joint that combines the difference in thermal properties between the weld and the base material, and analyze the distribution of heat input during welding based on the three-dimensional temperature field model to determine whether there is an area with uneven heat input distribution;

[0010] For areas with uneven heat input distribution, analyze the performance of welded joints under the combined action of thermal stress and mechanical stress, obtain the residual stress distribution and plastic deformation capacity differences, and analyze the stress concentration area in combination with the yield strength of the material;

[0011] According to the stress concentration area, the mapping relationship between welding process parameters and three-dimensional temperature field model is established. Through the training of historical welding process parameters and corresponding three-dimensional temperature field model, the welding process parameter adjustment prediction model is obtained.

[0012] The welding process parameters obtained from the welding process parameter adjustment prediction model are applied to the welding process control to make the real-time temperature field distribution uniform and reduce the temperature gradient between dissimilar metals. During the welding process, the real-time monitoring signal of welding deformation is obtained, and a real-time temperature field and tissue evolution coupling model is established to dynamically predict the weld cooling rate, heat input distribution and temperature gradient. According to the prediction results, the welding process parameters are adjusted in real time to optimize the overall quality of the dissimilar metal welding joints.

[0013] In an optional embodiment, the temperature field image is obtained by performing real-time temperature field scanning on the dissimilar metal welding joint using a temperature field scanning device, and the temperature field scanning device includes an infrared thermal imager;

[0014] The three-dimensional temperature field model of the dissimilar metal welding joint that combines the difference in thermal properties between the weld and the base material is constructed based on the temperature data using an interpolation algorithm;

[0015] The temperature gradient of each zone is calculated based on the three-dimensional temperature field model. The temperature gradient is obtained by calculating the temperature difference between adjacent zones and dividing it by the distance.

[0016] Based on the three-dimensional temperature field model, a thermal map showing the temperature anomaly area is generated through data visualization tools;

[0017] For the abnormal temperature area shown in the thermal map, the welding current, voltage and welding speed are adjusted in real time to control the heat input. During the welding process, the real-time monitoring signal of welding deformation is obtained, and a real-time temperature field and organization evolution coupling model is established to dynamically predict the weld cooling rate, heat input distribution and temperature gradient. According to the prediction results, the welding process parameters are adjusted in real time to optimize the overall quality of the dissimilar metal welding joint, including:

[0018] If the predicted temperature is higher than a preset threshold, the current is reduced or the welding speed is increased;

[0019] If the predicted temperature is below a preset threshold, the voltage is increased or the welding speed is decreased;

[0020] The welding process parameters are continuously optimized through real-time monitoring and dynamic adjustment to keep the weld cooling rate and temperature gradient within the preset range.

[0021] In an optional embodiment, the determining whether the temperature gradient between the temperature data of each zone exceeds a preset threshold value; performing numerical simulation on the temperature data when the temperature gradient exceeds the threshold value, establishing a three-dimensional temperature field model of a dissimilar metal welded joint combining the difference in thermal properties between the weld and the base material, analyzing the distribution of heat input during welding according to the three-dimensional temperature field model, and determining whether there is an area with uneven heat input distribution includes:

[0022] The Fourier heat conduction equation is discretized using the finite difference method;

[0023] According to the discretization results, the thermophysical parameters of the weld and the base material are obtained from the dissimilar metal material property database, and the thermophysical parameters are used as the attributes of the grid unit to construct a numerical calculation grid that reflects the difference in thermophysical properties between the weld and the base material.

[0024] The temperature field data is numerically simulated using a numerical calculation grid to obtain numerical simulation results;

[0025] A three-dimensional temperature field model of dissimilar metal welding joints is established based on the numerical simulation results, and the three-dimensional temperature field model is refined using the three-dimensional Kriging interpolation method.

[0026] Calculate the temperature gradient of each grid node in the three-dimensional temperature field model, and use Fourier's law to calculate the heat flux density;

[0027] According to the heat flux density, the average heat input of the weld zone and the heat affected zone is calculated to determine whether the difference in heat input between the weld zone and the heat affected zone exceeds a preset threshold. If it exceeds the preset threshold, it is determined that there is an uneven distribution of heat input.

[0028] In an optional embodiment, the analysis of the performance of the welded joint under the combined action of thermal stress and mechanical stress for the area with uneven heat input distribution, obtaining the residual stress distribution and the difference in plastic deformation capacity, and analyzing the stress concentration area in combination with the yield strength of the material includes:

[0029] Based on the temperature field data of the area with uneven heat input distribution, the finite element method is used to perform thermal-mechanical coupling analysis, including calculating the temperature distribution through the heat conduction equation;

[0030] The thermal stress distribution is obtained by using the temperature distribution and the thermal expansion coefficient of the material. The mechanical stress distribution is calculated by combining the mechanical constraints during the welding process.

[0031] Based on the superposition results of thermal stress and mechanical stress, the VonMises yield criterion is used to calculate the equivalent stress, where VonMises equivalent stress is:

[0032]

[0033] σ v represents the VonMises equivalent stress, σ 1 , σ 2 and σ 3 represent the principal stresses respectively;

[0034] The plastic strain distribution is solved by the elastic-plastic constitutive equation to obtain the residual stress distribution;

[0035] If the ratio of equivalent stress to yield strength exceeds a preset threshold, it is marked as a high stress area;

[0036] Adaptive mesh refinement for high stress areas: Calculate the stress gradient inside each mesh unit. If the stress gradient exceeds the preset threshold, divide the unit into four sub-units. Repeat this process until the preset maximum number of refinement layers is reached or the stress gradients of all units are below the threshold.

[0037] For the refined mesh, the thermal-mechanical coupling analysis is repeated and the stress and strain distribution results are updated.

[0038] In an optional embodiment, the mapping relationship between the welding process parameters and the three-dimensional temperature field model is established according to the stress concentration area, and the welding process parameter adjustment prediction model is obtained by training the historical welding process parameters and the corresponding three-dimensional temperature field model, including:

[0039] Using numerical simulation software, a three-dimensional mapping relationship between welding heat source power distribution, welding speed and temperature field distribution data is constructed to obtain a continuous function relationship in the parameter space;

[0040] Obtain historical welding process parameters and the corresponding three-dimensional temperature field model from the welding database, and perform standardization on the temperature field data, including selecting the principal component whose cumulative contribution rate reaches a preset threshold if the data dimension is high, and projecting the original data into the principal component space;

[0041] Using the historical data after dimension reduction and combined with the heat flux density distribution information, the support vector regression algorithm is used to train the correlation model between welding process parameters and temperature field distribution. The radial basis function is used as the kernel function in the correlation model.

[0042] Based on the correlation model, combined with the temperature field data of the real-time welding process, the trained correlation model is regarded as a function, and the target temperature field distribution is taken as the expected output;

[0043] By calculating the gradient of the loss function with respect to the welding process parameters, the welding heat source power and welding speed are iteratively updated until the preset optimization target is reached, and the prediction results of welding process parameter adjustment are obtained.

[0044] In an optional embodiment, the welding process parameters obtained by the welding process parameter adjustment prediction model are applied to the welding process control to make the real-time temperature field distribution uniform and reduce the temperature gradient between dissimilar metals, including:

[0045] Obtain welding heat source power distribution parameters and welding speed parameters obtained by the welding process parameter adjustment prediction model, and transmit them to the welding control unit through the data interface;

[0046] According to the welding heat source power distribution parameters and welding speed parameters, the proportional-integral-differential control method is used to adjust the welding power output power and wire feeding speed in real time, so that the deviation between the welding heat input and the predicted value is kept within the preset range;

[0047] The ideal heat input at each point on the welding trajectory is calculated using an interpolation algorithm, and the basic path is constructed by combining the workpiece geometric features and weld shape;

[0048] The temperature field data of the welding process is collected in real time by infrared thermal imager, and the temperature field data is divided into zones;

[0049] According to the temperature field data and the adjustment results of welding process parameters, the fuzzy controller with triangular membership function is used to dynamically adjust the swing amplitude and frequency of the welding gun to achieve uniform temperature field distribution and precise control of cooling rate. The input variables of the fuzzy controller are temperature gradient and cooling rate, and the output variables are the adjustment amount of the swing amplitude and frequency of the welding gun.

[0050] In an optional embodiment, during the welding process, real-time monitoring signals of welding deformation are obtained, a real-time coupled model of temperature field and microstructure evolution is established, the cooling rate, heat input distribution and temperature gradient of the weld are dynamically predicted, and welding process parameters are adjusted in real time according to the prediction results to optimize the overall quality of the dissimilar metal welded joint, including:

[0051] The data acquisition card is used to synchronously collect data from the infrared thermal imager, displacement sensor and strain gauge to obtain the temperature field, displacement and strain information during the welding process;

[0052] Establish a state space model based on the collected data, use different sensor data as observations, estimate the welding deformation state through the Kalman filter algorithm, and eliminate noise and redundant information;

[0053] According to the estimated welding deformation state, a coupled model of temperature field and microstructure evolution based on finite element method is established. The coupled model is used to describe the temperature distribution and metal microstructure evolution during welding.

[0054] Analyze temperature distribution and metal structure evolution, and dynamically predict weld cooling rate, heat input distribution, and temperature gradient;

[0055] According to the dynamic prediction results, a fuzzy adaptive controller is applied. If the deviation between the prediction result and the target value is large and the deviation change rate is positive, the welding power is increased;

[0056] If the deviation between the predicted result and the target value is small and the deviation change rate is negative, the welding power is reduced and the welding process parameters are adjusted in real time.

[0057] In a second aspect, the present invention also provides a temperature field optimization control system for a dissimilar metal welding joint, comprising:

[0058] Real-time temperature field scanning module, used to perform real-time temperature field scanning on dissimilar metal welding joints and obtain temperature field images;

[0059] The image segmentation processing module is used to perform image segmentation processing on the temperature field image, distinguish the weld area, heat-affected zone and base material area, and extract the temperature data of each area;

[0060] The numerical simulation module is used to determine whether the temperature gradient between the temperature data of each zone exceeds the preset threshold value; when the temperature gradient exceeds the threshold value, the temperature data is numerically simulated to establish a three-dimensional temperature field model of the dissimilar metal welding joint that combines the difference in thermal properties between the weld and the base material; based on the three-dimensional temperature field model, the heat input distribution during the welding process is analyzed to determine whether there is an area with uneven heat input distribution;

[0061] Stress analysis module, which is used to analyze the performance of welded joints under the combined action of thermal stress and mechanical stress in areas with uneven heat input distribution, obtain the residual stress distribution and plastic deformation capacity differences, and analyze the stress concentration area in combination with the yield strength of the material;

[0062] The welding process parameter adjustment prediction module is used to establish a mapping relationship between the welding process parameters and the three-dimensional temperature field model according to the stress concentration area, and obtain the welding process parameter adjustment prediction model through training of historical welding process parameters and the corresponding three-dimensional temperature field model;

[0063] The real-time process control module is used to apply the welding process parameters obtained from the welding process parameter adjustment prediction model to the welding process control, so as to make the real-time temperature field distribution uniform and reduce the temperature gradient between dissimilar metals; during the welding process, the real-time monitoring signal of welding deformation is obtained, and a real-time temperature field and tissue evolution coupling model is established to dynamically predict the weld cooling rate, heat input distribution and temperature gradient. According to the prediction results, the welding process parameters are adjusted in real time to optimize the overall quality of the dissimilar metal welding joints.

[0064] In a third aspect, the present invention further provides an electronic device, comprising: one or more processors, and a memory; the processor and the memory are coupled; the memory is used to store computer program codes, the computer program codes comprising computer instructions;

[0065] When the processor executes the computer instructions, the electronic device executes the method for optimizing the temperature field of a dissimilar metal welding joint as described in any one of the above technical solutions.

[0066] In a fourth aspect, the present invention also provides a computer-readable storage medium, which stores computer instructions or programs. When the computer instructions or programs are run on a computer, the computer executes the temperature field optimization control method for dissimilar metal welding joints described in any one of the above technical solutions.

[0067] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0068] The present invention discloses a method for optimizing and controlling the temperature field of a dissimilar metal welding joint. The method extracts the temperature data of each region by real-time scanning of the welding temperature field and combines image segmentation technology to establish a three-dimensional temperature field model. Based on the temperature gradient and heat flux density distribution, the stress concentration area is analyzed, and a mapping relationship between the welding process parameters and the temperature field distribution is established. By real-time monitoring of welding deformation, a coupling model of temperature field and organizational evolution is constructed to dynamically predict key parameters of the welding process. The present invention achieves uniform distribution of the temperature field and reduces the temperature gradient between dissimilar metals by real-time adjustment of process parameters such as welding heat source power distribution and welding speed, thereby optimizing the overall quality of the welding joint. The method can effectively solve the problems of uneven heat input and stress concentration in dissimilar metal welding, and improve welding quality and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0070] Figure 1 The present invention is a flow chart of a method for optimizing the temperature field of a dissimilar metal welding joint according to an embodiment of the present invention. DETAILED DESCRIPTION

[0071] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0072] Combine the following Figure 1 , describing an embodiment of the present invention.

[0073] According to an embodiment of the present invention, in a first aspect, a method for optimizing and controlling a temperature field of a dissimilar metal welding joint is provided, comprising:

[0074] Step S10, performing real-time temperature field scanning on the dissimilar metal welding joint to obtain a temperature field image;

[0075] Step S20, performing image segmentation processing on the temperature field image, distinguishing the weld area, the heat affected zone and the base material area, and extracting the temperature data of each area.

[0076] An infrared thermal imager is used to scan the temperature field of dissimilar metal welding joints in real time to obtain temperature data of the weld area, heat-affected zone and base metal area; based on the temperature data, the weld cooling rate and heat input distribution are calculated, and the temperature field image is segmented to extract the temperature data of each area; for the extracted temperature data, a three-dimensional temperature distribution model is constructed using an interpolation algorithm to calculate the temperature gradient between each area. The temperature gradient is obtained by dividing the temperature difference between adjacent areas by the distance; based on the three-dimensional temperature distribution model, a thermal map is generated using a data visualization tool to display the temperature abnormality area; for the temperature abnormality area displayed on the thermal map, the welding current, voltage and welding speed are adjusted in real time to control the heat input. The real-time control of the welding process includes reducing the current or increasing the welding speed if the predicted temperature is higher than the preset threshold; increasing the voltage or reducing the welding speed if the predicted temperature is lower than the preset threshold; through real-time control of the welding process, the welding process parameters are continuously optimized to keep the weld cooling rate and temperature gradient within the preset range.

[0077] Specifically, the calculated temperature gradient value is compared with the preset threshold, and the position and value exceeding the threshold are recorded. A thermal map is generated by a data visualization tool to intuitively display the temperature abnormality area. According to the thermal map data, the K-means clustering algorithm is used to classify the temperature abnormality area, determine the spatial range of the abnormal temperature distribution and the temperature change trend, and establish the association model between welding process parameters and temperature distribution through multivariate linear regression to predict the temperature field distribution under different welding conditions. According to the association model, the expected temperature distribution under different welding process parameter combinations is calculated, and the parameter combination closest to the target temperature distribution is selected as the adjustment suggestion. According to the predicted temperature field distribution, the temperature control target value is set, and the rule-based control strategy is used to adjust the welding current, voltage and welding speed in real time to control the heat input. According to the adjustment suggestion, the specific adjustment rules of welding current, voltage and speed are formulated as follows: if the predicted temperature is too high, the current is reduced or the welding speed is increased; if the predicted temperature is too low, the voltage is increased or the welding speed is reduced. The welding process parameters are continuously optimized through the feedback control mechanism to keep the weld cooling rate and temperature gradient within a reasonable range. The infrared thermal imager scans the temperature field of the dissimilar metal welding joint in real time at a frequency of 30 frames per second, and obtains the temperature distribution data with a resolution of 1280×960 pixels. The data processing algorithm calculates the cooling rate of the weld, such as the time required to drop from 800℃ to 500℃ is 10 seconds, and the heat input distribution is 1.5 kilojoules per millimeter. The image segmentation processing uses the threshold segmentation method to divide the temperature field image into the weld area, whose temperature is greater than 800℃, the heat affected zone, whose temperature is 500℃-800℃, and the parent metal area, whose temperature is less than 500℃. The three-dimensional temperature distribution model is constructed by the cubic spline interpolation algorithm, and the temperature gradient from the center of the weld to the edge of the parent metal is calculated to be 100℃ per millimeter. The preset temperature gradient threshold is 120℃ per millimeter, and the position coordinates and specific values ​​exceeding the threshold are recorded. The thermal map is color-coded, red represents the high temperature zone, whose temperature is above 800℃, yellow represents the medium temperature zone, whose temperature is 500℃-800℃, and blue represents the low temperature zone, whose temperature is below 500℃. The K-means clustering algorithm sets three cluster centers to classify the abnormal temperature areas, and determines that the abnormal temperature distribution range is ±5 mm from the center of the weld, and the temperature change trend is exponential decay. The multivariate linear regression model uses welding current 100-300 amperes, voltage 20-40 volts, and speed 2-10 mm per second as independent variables, and temperature field distribution as the dependent variable to establish an association model. The temperature field distribution under different welding conditions is predicted, such as when the current is 200 amperes, the voltage is 30 volts, and the speed is 5 mm per second, the maximum temperature in the center of the weld is 1200°C. The parameter combination closest to the target temperature distribution of 1000°C is selected as current 180 amperes, voltage 28 volts, and speed 6 mm per second.The rule-based control strategy adjusts the welding process parameters according to the deviation between the predicted temperature and the target temperature: if the predicted temperature exceeds the target temperature by 100°C, the current is reduced by 10 amps or the welding speed is increased by 0.5 mm / s; if the predicted temperature is lower than the target temperature by 100°C, the voltage is increased by 2 volts or the welding speed is reduced by 0.5 mm / s. The feedback control mechanism optimizes the parameters every 0.1 seconds to keep the weld cooling rate at 50°C per second and the temperature gradient within 100°C per millimeter.

[0078] Step S30, determining whether the temperature gradient between the temperature data of each zone exceeds a preset threshold value; when the temperature gradient exceeds the threshold value, numerical simulation is performed on the temperature data to establish a three-dimensional temperature field model of the dissimilar metal welding joint that combines the difference in thermal properties between the weld and the base material; based on the three-dimensional temperature field model, the distribution of heat input during the welding process is analyzed to determine whether there is an area with uneven heat input distribution.

[0079] Specifically, in some embodiments, the Fourier heat conduction equation is discretized using a finite difference method;

[0080] According to the discretization results, the thermophysical parameters of the weld and the base material are obtained from the dissimilar metal material property database, and the thermophysical parameters are used as the attributes of the grid unit to construct a numerical calculation grid that reflects the difference in thermophysical properties between the weld and the base material.

[0081] The temperature field data is numerically simulated using a numerical calculation grid to obtain numerical simulation results;

[0082] A three-dimensional temperature field model of dissimilar metal welding joints is established based on the numerical simulation results, and the three-dimensional temperature field model is refined using the three-dimensional Kriging interpolation method.

[0083] Calculate the temperature gradient of each grid node in the three-dimensional temperature field model, and use Fourier's law to calculate the heat flux density;

[0084] According to the heat flux density, the average heat input of the weld zone and the heat affected zone is calculated to determine whether the difference in heat input between the weld zone and the heat affected zone exceeds a preset threshold. If it exceeds the preset threshold, it is determined that there is an uneven distribution of heat input.

[0085] Specifically, the temperature gradient is calculated based on the acquired temperature field data. If the temperature gradient exceeds the preset threshold, the finite difference method is used to discretize the Fourier heat conduction equation. The thermal physical properties of the weld and the base material, such as thermal conductivity, specific heat capacity and density, are obtained from the material property database of dissimilar metals. These parameters are used as the attributes of the grid unit to construct a numerical calculation grid containing the difference in thermal properties between the weld and the base material, and the temperature field data is numerically simulated. A three-dimensional temperature field model of the dissimilar metal welding joint is established based on the numerical simulation results, and the temperature field is refined using the three-dimensional Kriging interpolation method to improve the model accuracy. First, the numerical simulation results are used as known data points to define the spatial covariance function, such as the spherical model or the exponential model, and then the Kriging interpolation formula is used to calculate the temperature value of the unknown point. Finally, the interpolation accuracy is evaluated through cross-validation to obtain the continuous temperature distribution of each area during the welding process. The temperature gradient of each grid node in the three-dimensional temperature field model is calculated using the central difference method. The temperature gradient calculation formula is: Where h is the grid spacing, T is the temperature field, which is a three-dimensional function T(x,y,z), which represents the temperature value at the spatial coordinate (x,y,z). The heat flux density q is calculated using Fourier's law. The heat flux density calculation formula is: Where k is the thermal conductivity, and the three-dimensional distribution diagram of temperature gradient and heat flux density is drawn to intuitively display the heat transfer during welding. According to the heat flux density distribution diagram, the average heat input of the weld area and the heat affected zone is calculated. By statistically analyzing the historical welding data, the mean and standard deviation of the difference in heat input between the weld area and the heat affected zone are calculated, and the mean plus two times the standard deviation is used as the preset threshold. If the difference between the two exceeds this preset threshold, it is determined that there is an uneven distribution of heat input, and the position and heat input value of the uneven area are recorded to provide a basis for the subsequent optimization of welding process parameters. During the welding process of dissimilar metals, the infrared thermal imager collects temperature field data at a frequency of 50Hz with a resolution of 640×480 pixels. The temperature gradient calculation results show that the maximum gradient from the center to the edge of the weld is 150℃ / mm, which exceeds the preset threshold of 120℃ / mm. Then the finite difference method is started to discretize the Fourier heat conduction equation, and the grid size is set to 0.1mm×0.1mm×0.1mm. From the material property database, the thermal conductivity of the weld (austenitic stainless steel) is 16.3 W / (m·K), the specific heat capacity is 500 J / (kg·K), and the density is 7900 kg / m 3 ; The thermal conductivity of the parent material (low carbon steel) is 51.9W / (m·K), the specific heat capacity is 486J / (kg·K), and the density is 7850kg / m 3. Assign these parameters to the corresponding grid cells and construct a computational grid that contains the differences in thermophysical properties. The numerical simulation adopts the alternating direction implicit method, with a time step of 0.01s and a total simulation time of 10s. Based on the simulation results, the temperature field model is established using the three-dimensional Kriging interpolation method. The spherical model is selected as the spatial covariance function, and the effective range is set to 5mm. The resolution of the temperature field after interpolation is increased to 0.05mm×0.05mm×0.05mm. The temperature gradient is calculated using the central difference method, and the maximum gradient of 145℃ / mm is observed at the interface between the weld and the base material. The heat flux calculation shows that the maximum value at the center of the weld is 3.2×10W / m 2 . The drawn three-dimensional distribution diagram clearly shows the diffusion process of heat transfer from the weld to the base material. The average heat input in the weld area is calculated to be 1.8 kJ / mm, and that in the heat-affected zone is 1.2 kJ / mm. Analysis of historical data shows that the mean of the heat input difference is 0.5 kJ / mm, and the standard deviation is 0.1 kJ / mm. Based on this, the preset threshold is set to 0.7 kJ / mm. During the current welding process, the difference in heat input between the two areas is 0.6 kJ / mm, which does not exceed the threshold, but is close to the critical value. The coordinates of the area with relatively uneven heat input distribution (x=5 mm, y=2 mm, z=1 mm) and the heat input at this location of 1.5 kJ / mm are recorded, providing an accurate basis for the subsequent optimization of welding process parameters.

[0087] Step S40, for the area with uneven heat input distribution, analyze the performance of the weld joint under the combined action of thermal stress and mechanical stress, obtain the residual stress distribution and plastic deformation capacity difference, and analyze the stress concentration area in combination with the yield strength of the material.

[0088] Specifically, in some embodiments, a finite element method is used to perform a thermal-mechanical coupling analysis based on temperature field data of a region where heat input is unevenly distributed, including calculating temperature distribution through a heat conduction equation;

[0089] The thermal stress distribution is obtained by using the temperature distribution and the thermal expansion coefficient of the material. The mechanical stress distribution is calculated by combining the mechanical constraints during the welding process.

[0090] Based on the superposition results of thermal stress and mechanical stress, the VonMises yield criterion is used to calculate the equivalent stress, where VonMises equivalent stress is:

[0091]

[0092] σ v represents the VonMises equivalent stress, σ 1 , σ 2 and σ 3 represent the principal stresses respectively;

[0093] The plastic strain distribution is solved by the elastic-plastic constitutive equation to obtain the residual stress distribution;

[0094] If the ratio of equivalent stress to yield strength exceeds a preset threshold, it is marked as a high stress area;

[0095] Adaptive mesh refinement for high stress areas: Calculate the stress gradient inside each mesh unit. If the stress gradient exceeds the preset threshold, divide the unit into four sub-units. Repeat this process until the preset maximum number of refinement layers is reached or the stress gradients of all units are below the threshold.

[0096] For the refined mesh, the thermal-mechanical coupling analysis is repeated and the stress and strain distribution results are updated.

[0097] Specifically, based on the temperature field data in the area where the heat input is unevenly distributed, the finite element method is used to discretize the solution area, and the solution area is divided into a finite number of units and nodes. Through the heat conduction equation, combined with the temperature field data in the area where the heat input is unevenly distributed, the temperature value of each node is calculated to obtain the temperature distribution of the entire area. According to the thermal expansion coefficient of the material, the temperature distribution is used to calculate the thermal strain of each node to obtain the thermal stress distribution. According to the mechanical constraints in the welding process, the force balance equation is used to calculate the displacement of each node to obtain the mechanical stress distribution. The thermal stress distribution and the mechanical stress distribution are superimposed to obtain the comprehensive stress distribution of the welding process. Based on the superposition results of thermal stress and mechanical stress, the VonMises yield criterion is used to calculate the equivalent stress, where the VonMises equivalent stress:

[0098]

[0099] σ 1 , σ 2 , σ 3The three principal stresses for each grid node refer to the stress components in three mutually perpendicular directions inside the material at a certain point. The plastic strain distribution is solved by the elastic-plastic constitutive equation to obtain the residual stress distribution, and the equivalent plastic strain of each region is calculated. By comparing the equivalent plastic strain values ​​of different regions, the plastic strain ratio is calculated, which is the ratio of the equivalent plastic strain of different regions to the maximum equivalent plastic strain to quantify the difference in plastic deformation capacity. The stress field and strain field are smoothed by the interpolation algorithm. The yield strength data of each region of the weld joint are obtained from the material database. By comparing the ratio of equivalent stress to yield strength, the stress concentration area is judged. If the ratio exceeds the preset threshold, it is marked as a high stress area. The quadtree algorithm is used to adaptively refine the high stress area, including first dividing the entire calculation domain into initial grids, calculating the stress gradient inside each grid unit, and if the stress gradient exceeds the preset threshold, the unit is divided into four sub-units. This process is repeated until the preset maximum number of refinement layers is reached or the stress gradient of all units is lower than the threshold. For the refined mesh, the thermal-mechanical coupling analysis was re-performed to update the stress and strain distribution results. The singularity index method was used to evaluate the stress singularity point, and the characteristic equation method was used to solve the stress distribution characteristics near the stress singularity point. The stress intensity factor was calculated, and the displacement extrapolation method was used to evaluate the stress intensity factor by calculating the displacement of the node near the crack tip. The local failure risk of the welded joint was determined to provide a basis for the subsequent optimization of the welding process. During the welding process of dissimilar metals, the temperature field data in the area with uneven heat input distribution showed that the highest temperature was 1500℃ and the lowest temperature was 200℃. The thermal-mechanical coupling analysis model was established using ANSYS finite element software, and the area was discretized and divided into 10,000 tetrahedral units and 5,000 nodes with a mesh size of 0.1mm. According to the heat conduction equation, combined with the temperature field data in the area with uneven heat input distribution, the temperature distribution range was calculated to be 200-1500℃. Using the thermal expansion coefficient, where austenitic stainless steel is 16.5×10 -6 / ℃, low carbon steel is 11.7×10 -6 / ℃ to obtain the thermal stress distribution, and the maximum thermal stress is 500MPa. Considering the welding fixture constraint, the maximum mechanical stress is calculated to be 300MPa. The VonMises yield criterion is applied to calculate the equivalent stress, and the maximum value is 750MPa. The plastic strain distribution is solved by the Ramberg-Osgood model, and the maximum equivalent plastic strain is 0.05. The plastic strain ratio is calculated, which is 1.0 in the weld area, 0.8 in the heat-affected zone, and 0.2 in the parent material area, and the difference in plastic deformation capacity is quantified. The stress field is smoothed using cubic spline interpolation. The yield strength is obtained from the material database, where austenitic stainless steel is 210MPa and low carbon steel is 250MPa. The stress ratio is calculated, and the maximum value is 3.57, which exceeds the preset threshold of 2.5 and is marked as a high stress area. The quadtree algorithm is used for mesh refinement, with an initial mesh size of 0.1mm, a stress gradient threshold of 50MPa / mm, a maximum number of refinement layers of 4, and a minimum mesh size of 0.00625mm. After refinement and reanalysis, the maximum stress dropped to 720MPa. The singularity index was calculated using the characteristic equation method, and the value was 0.3, indicating the presence of a stress singularity point. The stress intensity factor was calculated using the displacement extrapolation method, KⅠ=25MPa·m^0.5, which is lower than the material fracture toughness of 30MPa·m^0.5. It is judged that the risk of local failure of the welded joint is low, but the welding process still needs to be optimized to reduce residual stress.

[0100] Step S50, establishing a mapping relationship between welding process parameters and a three-dimensional temperature field model according to the stress concentration area, and obtaining a welding process parameter adjustment prediction model by training historical welding process parameters and the corresponding three-dimensional temperature field model.

[0101] Specifically, in some embodiments, numerical simulation software is used to construct a three-dimensional mapping relationship between welding heat source power distribution, welding speed and temperature field distribution data to obtain a continuous function relationship in the parameter space;

[0102] Obtain historical welding process parameters and the corresponding three-dimensional temperature field model from the welding database, and perform standardization on the temperature field data, including selecting the principal component whose cumulative contribution rate reaches a preset threshold if the data dimension is high, and projecting the original data into the principal component space;

[0103] Using the historical data after dimension reduction and combined with the heat flux density distribution information, the support vector regression algorithm is used to train the correlation model between welding process parameters and temperature field distribution. The radial basis function is used as the kernel function in the correlation model.

[0104] Based on the correlation model, combined with the temperature field data of the real-time welding process, the trained correlation model is regarded as a function, and the target temperature field distribution is taken as the expected output;

[0105] By calculating the gradient of the loss function with respect to the welding process parameters, the welding heat source power and welding speed are iteratively updated until the preset optimization target is reached, and the prediction results of welding process parameter adjustment are obtained.

[0106] Specifically, according to the stress concentration analysis results, numerical simulation software is used to construct the three-dimensional mapping relationship between welding heat source power distribution, welding speed and temperature field distribution, and a three-dimensional scatter plot is used to represent the relationship between the three. Then, a three-dimensional spline interpolation method is used to create a continuous mapping function. The discrete data points are processed continuously through the interpolation algorithm to obtain a continuous function relationship in the parameter space. The historical welding process parameters and the corresponding temperature field data are extracted from the welding database, the data are standardized, and the principal component analysis method is used to reduce the data dimension. The specific steps include calculating the covariance matrix, solving the eigenvalues ​​and eigenvectors, selecting the principal components with a cumulative contribution rate of 95%, and projecting the original data into the principal component space, so as to screen out the key characteristic parameters that affect the temperature field distribution. Using the historical data after dimensionality reduction and combining with the heat flux density distribution information, the support vector regression algorithm is used to train the association model between the welding process parameters and the temperature field distribution. The radial basis function is selected as the kernel function, and the optimal penalty coefficient and kernel function parameters are determined by grid search. The heat flux density distribution information is incorporated into the model training process as an additional feature, and the optimal hyperparameter configuration is determined by cross-validation. Based on the trained association model and the temperature field data in the real-time welding process, the trained model is regarded as a function, and the target temperature field distribution is taken as the expected output. By calculating the gradient of the loss function to the welding process parameters, the welding heat source power and welding speed are iteratively updated until the preset optimization target or number of iterations is reached, thereby obtaining the prediction results of welding process parameter adjustment and providing parameter guidance for adaptive welding control. In the process of dissimilar metal welding, the stress concentration analysis results show that the maximum stress is 750MPa, which is located at the junction of the weld and the heat affected zone. The three-dimensional mapping relationship is constructed using COMSOL Multiphysics software. The welding heat source power range is 2000-5000W, the welding speed range is 5-15mm / s, and the temperature field distribution range is 200-1800℃. A three-dimensional scatter plot is used to represent 1000 data points, and a continuous mapping function is created by the cubic spline interpolation method, with an interpolation accuracy of 99.5%. 10,000 sets of historical data are extracted from the welding database, and principal component analysis is performed after standardization. The top five principal components are selected, and the cumulative contribution rate reaches 96.8%. The support vector regression algorithm uses a radial basis function kernel, and the optimal penalty coefficient C = 10 and the kernel function parameter γ = 0.01 are determined by grid search. The heat flux density distribution information is used as an additional feature, and the total input feature dimension is 8. The cross-validation adopts a 5-fold method, and the average relative error of the model is 3.2%. During the real-time welding process, the temperature field data is collected every 0.1 second, and the target temperature field distribution is set to 1500℃ at the center of the weld and 800℃ at the edge of the heat-affected zone. The welding process parameters are optimized by the gradient descent method, the learning rate is set to 0.01, and the maximum number of iterations is 100.The final optimized welding heat source power is 3500W, the welding speed is 8mm / s, and the predicted temperature field distribution has an error of less than 5% from the target, providing precise parameter guidance for the adaptive welding control system.

[0107] Step S60, applying the welding process parameters obtained by the welding process parameter adjustment prediction model to the welding process control, so as to make the real-time temperature field distribution uniform and reduce the temperature gradient between dissimilar metals.

[0108] Specifically, in some embodiments, applying the welding process parameters obtained by the welding process parameter adjustment prediction model to the welding process control to make the real-time temperature field distribution uniform and reduce the temperature gradient between dissimilar metals includes:

[0109] Obtain welding heat source power distribution parameters and welding speed parameters obtained by the welding process parameter adjustment prediction model, and transmit them to the welding control unit through the data interface;

[0110] According to the welding heat source power distribution parameters and welding speed parameters, the proportional-integral-differential control method is used to adjust the welding power output power and wire feeding speed in real time, so that the deviation between the welding heat input and the predicted value is kept within the preset range;

[0111] The ideal heat input at each point on the welding trajectory is calculated using an interpolation algorithm, and the basic path is constructed by combining the workpiece geometric features and weld shape;

[0112] The temperature field data of the welding process is collected in real time by infrared thermal imager, and the temperature field data is divided into zones;

[0113] According to the temperature field data and the adjustment results of welding process parameters, the fuzzy controller with triangular membership function is used to dynamically adjust the swing amplitude and frequency of the welding gun to achieve uniform temperature field distribution and precise control of cooling rate. The input variables of the fuzzy controller are temperature gradient and cooling rate, and the output variables are the adjustment amount of the swing amplitude and frequency of the welding gun.

[0114] Specifically, the optimized welding heat source power distribution and welding speed parameters are obtained from the prediction model, and the parameters are transmitted to the welding control unit through the data interface. The proportional-integral-differential control algorithm is used to adjust the welding power output and wire feeding speed in real time to keep the deviation between the welding heat input and the predicted value within the range of ±5%. The PID controller parameters are determined by the Ziegler-Nichols tuning method, and the proportional gain Kp, integral time Ti and differential time Td are obtained by the critical oscillation method. Then, the PID parameters are optimized in real time using the online self-tuning algorithm. According to the welding heat source power distribution data, the interpolation algorithm is used to calculate the ideal heat input at each point on the welding trajectory. Combined with the workpiece geometric characteristics and weld shape, the A* algorithm is used to construct the basic path. The dynamic window method is combined to adjust the local path in real time, and the optimal welding gun movement speed, swing amplitude and frequency are dynamically calculated to generate an optimized welding trajectory. The temperature field data of the welding process is collected in real time by infrared thermal imager. The temperature field image is partitioned by Otsu threshold segmentation method. The contours of weld, heat affected zone and parent metal zone are extracted by morphological operation. Then, the average temperature and temperature gradient of each zone are calculated by regional growing algorithm. If the temperature gradient exceeds the preset threshold, the welding process parameter fine-tuning mechanism is triggered. Based on the temperature field data and the fine-tuning results of welding process parameters, the fuzzy controller of triangular membership function is used. The input variables are temperature gradient and cooling rate, and the output variables are the adjustment amount of welding torch swing amplitude and frequency. The welding torch swing amplitude and frequency are dynamically adjusted. At the same time, a three-layer feedforward network structure is used to predict the cooling rate of the weld. The input layer includes the current temperature, heat input and material properties, and the output layer is the predicted cooling rate. By adjusting the welding heat source power distribution and welding speed, the temperature field distribution is homogenized and the cooling rate is precisely controlled. Especially for the dissimilar metal joint area, the temperature gradient is reduced by adjusting the local heat input. In the process of dissimilar metal welding, the optimized welding heat source power distribution output by the prediction model is 2500-3500W, and the welding speed is 8-12mm / s. The data interface transmits the parameters to the welding control unit at a frequency of 100Hz. The initial parameters of the PID controller are set to Kp=0.6, Ti=0.5s, Td=0.125s by the Ziegler-Nichols method. The online self-tuning algorithm adjusts the parameters every 5 seconds to control the welding heat input deviation within ±3%. The A* algorithm divides the workpiece surface into 0.1mm grids, and combines the dynamic window method to update the local path every 0.5s. The optimal moving speed of the welding gun is calculated to be 10mm / s, the swing amplitude is 3mm, and the frequency is 2Hz. The infrared thermal imager collects temperature field images with a resolution of 640×480 and a frame rate of 60fps. The Otsu threshold method divides the temperature field into three areas: weld (>1000℃), heat-affected zone (600-1000℃) and parent metal area (<600℃).The morphological operation uses 3×3 structural elements to extract the regional contour. The regional growing algorithm calculates the average temperature of the weld zone to be 1200℃, the heat affected zone to be 750℃, and the maximum temperature gradient to be 150℃ / mm. The fuzzy controller designs 5 fuzzy sets, the input variable temperature gradient range is 0-200℃ / mm, the cooling rate range is 0-100℃ / s, the output variable welding gun swing amplitude adjustment is -1mm to +1mm, and the frequency adjustment is -0.5Hz to +0.5Hz. The three-layer feedforward neural network structure is 10-20-1, with 5000 sets of training samples, and the prediction accuracy of the weld cooling rate reaches ±5℃ / s. Through these precise control measures, the temperature field distribution is uniform, the temperature gradient in the weld area is reduced to below 120℃ / mm, and the cooling rate is controlled in the range of 40-60℃ / s, which effectively reduces the temperature gradient in the dissimilar metal joint area and improves the welding quality.

[0115] The present invention provides a method for optimizing and controlling the temperature field of a dissimilar metal welding joint. The method extracts the temperature data of each region by real-time scanning of the welding temperature field and combines image segmentation technology to establish a three-dimensional temperature field model. Based on the temperature gradient and heat flux density distribution, the stress concentration area is analyzed, and a mapping relationship between the welding process parameters and the temperature field distribution is established. By real-time monitoring of welding deformation, a coupling model of temperature field and organizational evolution is constructed to dynamically predict key parameters of the welding process. The present invention achieves uniform distribution of the temperature field and reduces the temperature gradient between dissimilar metals by real-time adjustment of process parameters such as welding heat source power distribution and welding speed, thereby optimizing the overall quality of the welding joint. This method can effectively solve problems such as uneven heat input and stress concentration in dissimilar metal welding, and improve welding quality and efficiency.

[0116] Step S70, during the welding process, obtain the real-time monitoring signal of welding deformation, establish a real-time temperature field and tissue evolution coupling model, dynamically predict the weld cooling rate, heat input distribution and temperature gradient, and adjust the welding process parameters in real time according to the prediction results to optimize the overall quality of the dissimilar metal welding joint.

[0117] Specifically, in some embodiments, a data acquisition card is used to synchronously collect data from an infrared thermal imager, a displacement sensor, and a strain gauge to obtain temperature field, displacement, and strain information during welding;

[0118] Establish a state space model based on the collected data, use different sensor data as observations, estimate the welding deformation state through the Kalman filter algorithm, and eliminate noise and redundant information;

[0119] According to the estimated welding deformation state, a coupled model of temperature field and microstructure evolution based on finite element method is established. The coupled model is used to describe the temperature distribution and metal microstructure evolution during welding.

[0120] Analyze temperature distribution and metal structure evolution, and dynamically predict weld cooling rate, heat input distribution, and temperature gradient;

[0121] According to the dynamic prediction results, a fuzzy adaptive controller is applied. If the deviation between the prediction result and the target value is large and the deviation change rate is positive, the welding power is increased;

[0122] If the deviation between the predicted result and the target value is small and the deviation change rate is negative, the welding power is reduced and the welding process parameters are adjusted in real time.

[0123] Specifically, a data acquisition card is used to synchronously collect data from infrared thermal imagers, displacement sensors and strain gauges. Image registration is used to align the temperature fields at different times for infrared thermal imager data, and time synchronization and data normalization are performed on displacement sensor and strain gauge data. A state space model is designed, and different sensor data are used as observations. The Kalman filter algorithm is used to estimate the welding deformation state, eliminate noise and redundant information, and obtain a real-time monitoring signal of welding deformation. According to the fused data, a coupled model of temperature field and microstructure evolution based on the finite element method is established, and a three-dimensional finite element mesh is established. The unit type is selected as hexahedral quadratic element, and the mesh size is refined to 0.5 mm in the weld area. The heat conduction equation uses the Crank-Nicolson implicit difference format, and the phase change kinetic equation uses the Johnson-Mehl-Avrami-Kolmogorov model. The coupled equations are solved by the Newton-Raphson iterative method to describe the temperature distribution and metal microstructure evolution during welding. The long short-term memory network algorithm is used to dynamically predict the cooling rate, heat input distribution and temperature gradient of the weld. A three-layer LSTM structure is designed. The number of nodes in the input layer is the number of sensors, the hidden layer has 100 neurons, and the three nodes in the output layer correspond to the cooling rate, heat input and temperature gradient respectively. The data of the past hour is used as input to predict the trend of change in the next 10 minutes. The root mean square error is used as the loss function, and the Adam optimizer is used for training, and the confidence interval of the prediction result is calculated. Based on the prediction results and confidence intervals, a fuzzy adaptive controller is designed. Five fuzzy sets are designed. The input variables are the deviation and deviation change rate between the prediction result and the target value, and the output variable is the adjustment amount of the welding process parameters. A rule base is formulated, such as "if the deviation is large and the deviation change rate is positive, the welding power is greatly increased." The welding heat source power, welding speed, cooling rate, welding temperature and holding time are used as control variables to adjust the welding process parameters in real time. For the holding time, it is dynamically adjusted according to the cooling rate prediction results. The faster the cooling rate, the longer the holding time. The overall quality of the dissimilar metal welding joint is continuously optimized through the feedback control mechanism. In the process of dissimilar metal welding, NIPXIe-6363 data acquisition card is used to synchronously collect data with a sampling frequency of 1kHz, including an infrared thermal imager with a resolution of 640×480 and a temperature accuracy of ±2℃, a displacement sensor with a range of 0-10mm and an accuracy of 0.01mm, and a strain gauge with a sensitivity factor of 2.08±1%. The infrared thermal imager data is image-aligned using the phase correlation method, and the alignment accuracy reaches the sub-pixel level; the displacement and strain data are denoised by a 5th-order Butterworth low-pass filter (cutoff frequency 100Hz). The state vector of the Kalman filter algorithm contains 6 variables, which are the displacement and velocity in the x, y, and z directions. The observation matrix is ​​constructed by the sensor data, and the welding deformation state estimation is obtained by iterative calculation.The finite element model was constructed using ANSYS software, with a total grid size of 500,000 units. The grid size in the weld area was 0.5 mm, and gradually transitioned to 5 mm in the far field area. The boundary conditions of the heat conduction equation consider convection and radiation heat transfer, and the material properties vary with temperature. The JMAK model parameters were obtained through isothermal phase transformation experiments, such as the austenitization activation energy of 280 kJ / mol. The LSTM network was implemented using the TensorFlow framework, and the training data set contained 1,000 sets of welding process data, 200 sets of validation sets, and 100 sets of test sets. The learning rate of the Adam optimizer was set to 0.001, the batch size was 64, and the number of training rounds was 200. The input variables "bias" and "bias change rate" of the fuzzy controller were each divided into 5 fuzzy sets, and the output variable "adjustment amount" was divided into 7 fuzzy sets, with a total of 25 fuzzy rules. The welding heat source power adjustment range is ±500W, the welding speed adjustment range is ±2mm / s, the cooling rate adjustment range is ±10℃ / s, the welding temperature adjustment range is ±50℃, and the holding time adjustment range is ±30s. Through multi-sensor fusion and intelligent control, the temperature field uniformity of the welded joint is improved by 15%, the organizational consistency is improved by 20%, and the quality of dissimilar metal welding is significantly optimized.

[0124] According to an embodiment of the present invention, in a second aspect, a temperature field optimization control system for a dissimilar metal welding joint is provided, comprising:

[0125] Real-time temperature field scanning module, used to perform real-time temperature field scanning on dissimilar metal welding joints and obtain temperature field images;

[0126] The image segmentation processing module is used to perform image segmentation processing on the temperature field image, distinguish the weld area, heat-affected zone and base material area, and extract the temperature data of each area;

[0127] The numerical simulation module is used to determine whether the temperature gradient between the temperature data of each zone exceeds the preset threshold value; when the temperature gradient exceeds the threshold value, the temperature data is numerically simulated to establish a three-dimensional temperature field model of the dissimilar metal welding joint that combines the difference in thermal properties between the weld and the base material; based on the three-dimensional temperature field model, the heat input distribution during the welding process is analyzed to determine whether there is an area with uneven heat input distribution;

[0128] Stress analysis module, which is used to analyze the performance of welded joints under the combined action of thermal stress and mechanical stress in areas with uneven heat input distribution, obtain the residual stress distribution and plastic deformation capacity differences, and analyze the stress concentration area in combination with the yield strength of the material;

[0129] The welding process parameter adjustment prediction module is used to establish a mapping relationship between the welding process parameters and the three-dimensional temperature field model according to the stress concentration area, and obtain the welding process parameter adjustment prediction model through training of historical welding process parameters and the corresponding three-dimensional temperature field model;

[0130] The real-time process control module is used to apply the welding process parameters obtained from the welding process parameter adjustment prediction model to the welding process control, so as to make the real-time temperature field distribution uniform and reduce the temperature gradient between dissimilar metals; during the welding process, the real-time monitoring signal of welding deformation is obtained, and a real-time temperature field and tissue evolution coupling model is established to dynamically predict the weld cooling rate, heat input distribution and temperature gradient. According to the prediction results, the welding process parameters are adjusted in real time to optimize the overall quality of the dissimilar metal welding joints.

[0131] According to an embodiment of the present invention, in a third aspect, there is further provided an electronic device, comprising: one or more processors, and a memory; the processor and the memory are coupled; the memory is used to store a computer program code, wherein the computer program code comprises a computer instruction;

[0132] When the processor executes the computer instructions, the electronic device executes the temperature field optimization control method for dissimilar metal welding joints described in any one of the above technical solutions.

[0133] According to an embodiment of the present invention, in a fourth aspect, a computer-readable storage medium is also provided, wherein the computer-readable storage medium stores computer instructions or programs. When the computer instructions or programs are run on a computer, the computer executes the method for optimizing the temperature field of dissimilar metal welding joints described in any one of the above technical solutions.

[0134] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for optimizing the temperature field of a dissimilar metal welding joint, characterized in that: include: Scan the temperature field of dissimilar metal welding joints in real time to obtain temperature field images; Perform image segmentation processing on the temperature field image to distinguish the weld area, heat-affected zone and base metal area, and extract the temperature data of each area; Determine whether the temperature gradient between the temperature data of each zone exceeds the preset threshold; perform numerical simulation on the temperature data when the temperature gradient exceeds the threshold, establish a three-dimensional temperature field model of the dissimilar metal welding joint that combines the difference in thermal properties between the weld and the base material, and analyze the distribution of heat input during welding based on the three-dimensional temperature field model to determine whether there is an area with uneven heat input distribution; For areas with uneven heat input distribution, analyze the performance of welded joints under the combined action of thermal stress and mechanical stress, obtain the residual stress distribution and plastic deformation capacity differences, and analyze the stress concentration area in combination with the yield strength of the material; According to the stress concentration area, the mapping relationship between welding process parameters and three-dimensional temperature field model is established. Through the training of historical welding process parameters and corresponding three-dimensional temperature field model, the welding process parameter adjustment prediction model is obtained. Apply the welding process parameters obtained from the welding process parameter adjustment prediction model to the welding process control to make the real-time temperature field distribution uniform and reduce the temperature gradient between dissimilar metals; During the welding process, real-time monitoring signals of welding deformation are obtained, and a real-time coupling model of temperature field and microstructure evolution is established. The weld cooling rate, heat input distribution and temperature gradient are dynamically predicted. Based on the prediction results, the welding process parameters are adjusted in real time to optimize the overall quality of dissimilar metal welding joints.

2. The method for optimizing the temperature field of a dissimilar metal welding joint according to claim 1, characterized in that: The temperature field image is obtained by performing real-time temperature field scanning on the dissimilar metal welding joint using a temperature field scanning device, wherein the temperature field scanning device includes an infrared thermal imager; The three-dimensional temperature field model of the dissimilar metal welding joint that combines the difference in thermal properties between the weld and the base material is constructed based on the temperature data using an interpolation algorithm; The temperature gradient of each zone is calculated based on the three-dimensional temperature field model. The temperature gradient is obtained by calculating the temperature difference between adjacent zones and dividing it by the distance. Based on the three-dimensional temperature field model, a thermal map showing the temperature anomaly area is generated through data visualization tools; For the abnormal temperature area shown in the thermal map, the welding current, voltage and welding speed are adjusted in real time to control the heat input. During the welding process, the real-time monitoring signal of welding deformation is obtained, and a real-time temperature field and organization evolution coupling model is established to dynamically predict the weld cooling rate, heat input distribution and temperature gradient. According to the prediction results, the welding process parameters are adjusted in real time to optimize the overall quality of the dissimilar metal welding joint, including: If the predicted temperature is higher than a preset threshold, the current is reduced or the welding speed is increased; If the predicted temperature is below a preset threshold, the voltage is increased or the welding speed is decreased; The welding process parameters are continuously optimized through real-time monitoring and dynamic adjustment to keep the weld cooling rate and temperature gradient within the preset range.

3. The method for optimizing the temperature field of a dissimilar metal welding joint according to claim 1 or 2, characterized in that: The method of determining whether the temperature gradient between the temperature data of each zone exceeds a preset threshold value; performing numerical simulation on the temperature data when the temperature gradient exceeds the threshold value, establishing a three-dimensional temperature field model of a dissimilar metal welding joint combining the difference in thermal properties between the weld and the base material, analyzing the distribution of heat input during welding according to the three-dimensional temperature field model, and determining whether there is an area with uneven heat input distribution includes: The Fourier heat conduction equation is discretized using the finite difference method; According to the discretization results, the thermophysical parameters of the weld and the base material are obtained from the dissimilar metal material property database, and the thermophysical parameters are used as the attributes of the grid unit to construct a numerical calculation grid that reflects the difference in thermophysical properties between the weld and the base material. The temperature field data is numerically simulated using a numerical calculation grid to obtain numerical simulation results; A three-dimensional temperature field model of dissimilar metal welding joints is established based on the numerical simulation results, and the three-dimensional temperature field model is refined using the three-dimensional Kriging interpolation method. Calculate the temperature gradient of each grid node in the three-dimensional temperature field model, and use Fourier's law to calculate the heat flux density; According to the heat flux density, the average heat input of the weld zone and the heat affected zone is calculated to determine whether the difference in heat input between the weld zone and the heat affected zone exceeds a preset threshold. If it exceeds the preset threshold, it is determined that there is an uneven distribution of heat input.

4. The method for optimizing the temperature field of a dissimilar metal welding joint according to claim 1 or 2, characterized in that: The above analysis is aimed at the area with uneven heat input distribution, and the performance of the welded joint under the combined action of thermal stress and mechanical stress is analyzed to obtain the difference in residual stress distribution and plastic deformation capacity. The stress concentration areas are analyzed in combination with the yield strength of the material, including: Based on the temperature field data of the area with uneven heat input distribution, the finite element method is used to perform thermal-mechanical coupling analysis, including calculating the temperature distribution through the heat conduction equation; The thermal stress distribution is obtained by using the temperature distribution and the thermal expansion coefficient of the material. The mechanical stress distribution is calculated by combining the mechanical constraints during the welding process. Based on the superposition results of thermal stress and mechanical stress, the VonMises yield criterion is used to calculate the equivalent stress, where VonMises equivalent stress is: σ v represents VonMises equivalent stress, σ1, σ2 and σ3 represent principal stresses respectively; The plastic strain distribution is solved by the elastic-plastic constitutive equation to obtain the residual stress distribution; If the ratio of equivalent stress to yield strength exceeds a preset threshold, it is marked as a high stress area; Adaptive mesh refinement for high stress areas: Calculate the stress gradient inside each mesh unit. If the stress gradient exceeds the preset threshold, divide the unit into four sub-units. Repeat this process until the preset maximum number of refinement layers is reached or the stress gradients of all units are below the threshold. For the refined mesh, the thermal-mechanical coupling analysis is repeated and the stress and strain distribution results are updated.

5. The method for optimizing the temperature field of a dissimilar metal welding joint according to claim 1 or 2, characterized in that: The mapping relationship between welding process parameters and three-dimensional temperature field models is established according to the stress concentration area, and the welding process parameter adjustment prediction model is obtained by training the historical welding process parameters and the corresponding three-dimensional temperature field model, including: Using numerical simulation software, a three-dimensional mapping relationship between welding heat source power distribution, welding speed and temperature field distribution data is constructed to obtain a continuous function relationship in the parameter space; Obtain historical welding process parameters and the corresponding three-dimensional temperature field model from the welding database, and perform standardization on the temperature field data, including selecting the principal component whose cumulative contribution rate reaches a preset threshold if the data dimension is high, and projecting the original data into the principal component space; Using the historical data after dimension reduction and combined with the heat flux density distribution information, the support vector regression algorithm is used to train the correlation model between welding process parameters and temperature field distribution. The radial basis function is used as the kernel function in the correlation model. Based on the correlation model, combined with the temperature field data of the real-time welding process, the trained correlation model is regarded as a function, and the target temperature field distribution is taken as the expected output; By calculating the gradient of the loss function with respect to the welding process parameters, the welding heat source power and welding speed are iteratively updated until the preset optimization target is reached, and the prediction results of welding process parameter adjustment are obtained.

6. The method for optimizing the temperature field of a dissimilar metal welding joint according to claim 1 or 2, characterized in that: The welding process parameters obtained by the welding process parameter adjustment prediction model are applied to the welding process control to make the real-time temperature field distribution uniform and reduce the temperature gradient between dissimilar metals, including: Obtain welding heat source power distribution parameters and welding speed parameters obtained by the welding process parameter adjustment prediction model, and transmit them to the welding control unit through the data interface; According to the welding heat source power distribution parameters and welding speed parameters, the proportional-integral-differential control method is used to adjust the welding power output power and wire feeding speed in real time, so that the deviation between the welding heat input and the predicted value is kept within the preset range; The ideal heat input at each point on the welding trajectory is calculated using an interpolation algorithm, and the basic path is constructed by combining the workpiece geometric features and weld shape; The temperature field data of the welding process is collected in real time by infrared thermal imager, and the temperature field data is divided into zones; According to the temperature field data and the adjustment results of welding process parameters, the fuzzy controller with triangular membership function is used to dynamically adjust the swing amplitude and frequency of the welding gun to achieve uniform temperature field distribution and precise control of cooling rate. The input variables of the fuzzy controller are temperature gradient and cooling rate, and the output variables are the adjustment amount of the swing amplitude and frequency of the welding gun.

7. The method for optimizing the temperature field of a dissimilar metal welding joint according to claim 1 or 2, characterized in that: In the welding process, real-time monitoring signals of welding deformation are obtained, a real-time coupling model of temperature field and microstructure evolution is established, and the cooling rate, heat input distribution and temperature gradient of the weld are dynamically predicted. According to the prediction results, the welding process parameters are adjusted in real time to optimize the overall quality of the dissimilar metal welding joint, including: The data acquisition card is used to synchronously collect data from the infrared thermal imager, displacement sensor and strain gauge to obtain the temperature field, displacement and strain information during the welding process; Establish a state space model based on the collected data, use different sensor data as observations, estimate the welding deformation state through the Kalman filter algorithm, and eliminate noise and redundant information; According to the estimated welding deformation state, a coupled model of temperature field and microstructure evolution based on finite element method is established. The coupled model is used to describe the temperature distribution and metal microstructure evolution during welding. Analyze temperature distribution and metal structure evolution, and dynamically predict weld cooling rate, heat input distribution, and temperature gradient; According to the dynamic prediction results, a fuzzy adaptive controller is applied. If the deviation between the prediction result and the target value is large and the deviation change rate is positive, the welding power is increased; If the deviation between the predicted result and the target value is small and the deviation change rate is negative, the welding power is reduced and the welding process parameters are adjusted in real time.

8. A temperature field optimization control system for a dissimilar metal welding joint, characterized in that: include: Real-time temperature field scanning module, used to perform real-time temperature field scanning on dissimilar metal welding joints and obtain temperature field images; The image segmentation processing module is used to perform image segmentation processing on the temperature field image, distinguish the weld area, heat-affected zone and base material area, and extract the temperature data of each area; The numerical simulation module is used to determine whether the temperature gradient between the temperature data of each zone exceeds the preset threshold value; when the temperature gradient exceeds the threshold value, the temperature data is numerically simulated to establish a three-dimensional temperature field model of the dissimilar metal welding joint that combines the difference in thermal properties between the weld and the base material; based on the three-dimensional temperature field model, the heat input distribution during the welding process is analyzed to determine whether there is an area with uneven heat input distribution; Stress analysis module, which is used to analyze the performance of welded joints under the combined action of thermal stress and mechanical stress in areas with uneven heat input distribution, obtain the residual stress distribution and plastic deformation capacity differences, and analyze the stress concentration area in combination with the yield strength of the material; The welding process parameter adjustment prediction module is used to establish a mapping relationship between the welding process parameters and the three-dimensional temperature field model according to the stress concentration area, and obtain the welding process parameter adjustment prediction model through training of historical welding process parameters and the corresponding three-dimensional temperature field model; The real-time process control module is used to apply the welding process parameters obtained from the welding process parameter adjustment prediction model to the welding process control, so as to make the real-time temperature field distribution uniform and reduce the temperature gradient between dissimilar metals; during the welding process, the real-time monitoring signal of welding deformation is obtained, and a real-time temperature field and tissue evolution coupling model is established to dynamically predict the weld cooling rate, heat input distribution and temperature gradient. According to the prediction results, the welding process parameters are adjusted in real time to optimize the overall quality of the dissimilar metal welding joints.

9. An electronic device, characterized in that: include: One or more processors, and a memory; the processor and the memory are coupled; the memory is used to store computer program code, the computer program code includes computer instructions; When the processor executes the computer instructions, the electronic device executes the temperature field optimization control method for dissimilar metal welding joints as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions or programs. When the computer instructions or programs are executed on a computer, the computer executes the method for optimizing the temperature field of a dissimilar metal welding joint as described in any one of claims 1 to 7.

Images