A method for optimizing individual process parameters for the ageing of aluminium alloy profiles

Abstract

The application discloses an individualized process parameter optimization method for aging treatment of aluminum alloy profiles, aiming at solving the problems of uneven performance caused by wall thickness difference in industrial heat treatment of complex cross-section profiles and low precision of traditional experience process. The method first determines the precipitation characteristic temperature of the material by differential scanning calorimetry, and sets the aging temperature boundary in combination with the solid solution supersaturation; a multi-objective optimization model is established based on the response surface method, and the user demand weight coefficient is introduced to balance the strength, toughness and corrosion resistance and other indexes; the actual furnace and profile cross-section temperature difference data are obtained through industrial grade temperature field coupling test; the precipitation activation energy parameter of the material is introduced, the equivalent compensation time is calculated by using the Arrhenius equation, and the hysteresis compensation strategy of "reducing temperature and prolonging holding time" is implemented. The application realizes the accurate conversion from laboratory theoretical parameters to industrial production, and significantly improves the uniformity of profile cross-section performance and the scientificity of process customization.

Description

Technical Field

[0001] This invention relates to the field of aluminum alloy profile parameter optimization technology, and in particular to a personalized process parameter optimization method for aging treatment of aluminum alloy profiles. Background Technology

[0002] In actual production, many high-performance aluminum alloy profiles often have cross-sections containing extremely thin ribs and extremely thick walls. During traditional aging heat treatment, the thin-walled sections heat up very quickly, while the thick-walled core heats up very slowly due to lagging heat conduction. If the process is set according to the temperature of the thin-walled sections, the thick-walled core will experience "under-aging" (insufficient strength) because the temperature does not meet requirements; conversely, if the holding time is blindly extended or the furnace temperature is increased to accommodate the thick-walled sections, the thin-walled sections will experience "over-aging" (grain boundary coarsening, leading to a decrease in strength, toughness, and corrosion resistance) due to overheating. Many existing process optimization methods (such as simple response surface methodology) often remain at the laboratory small-scale stage. Laboratory samples are heated very uniformly in precision electric furnaces, and the resulting "theoretical optimal parameters" often become invalid when placed in industrial aging furnaces with large loading capacities and complex hot air circulation. Ignoring the non-uniformity of the spatial temperature field within industrial furnaces (such as temperature differences at furnace doors and corners) leads to large fluctuations in product performance and unstable yield rates during pilot or mass production. Faced with the challenges of quenching sensitivity and thermal hysteresis in thick-walled profiles, traditional factories often rely on the experience of veteran craftsmen to adjust processes (e.g., "bake the thicker material for an extra hour or two"). This experience lacks theoretical support from physical metallurgy and cannot precisely quantify "how much temperature to lower" and "how much time to extend" to ensure a consistent volume fraction of precipitated phases. For high-value-added new aluminum alloys, this blind trial-and-error approach is extremely costly and risky. Summary of the Invention

[0003] A method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles includes the following steps: S1. Multidimensional feature acquisition: Acquire the pre-extrusion quenching state parameters of the target aluminum alloy profile, the maximum and minimum wall thickness parameters of the profile section, and the personalized performance requirements set for the profile, including at least two mechanical and physical performance indicators and their weighting coefficients. S2. Thermodynamic boundary definition: Differential scanning calorimetry is used to scan the quenched sample of the profile to determine its characteristic peak temperature of the precipitate phase. Based on the characteristic peak temperature of the precipitate phase and combined with the solid solution supersaturation in the previous extrusion quenching state parameters, the dynamic upper and lower limits of the aging temperature are set. S3. Response surface modeling: Within the set dynamic upper and lower limits of the effective temperature, a response surface experiment is designed and a small-scale experiment is performed with the aging temperature and holding time as decision variables. Based on the experimental results, a second-order polynomial prediction model is fitted and constructed with the decision variables as independent variables and the performance index as dependent variables. S4. Multi-objective optimization solution of the model: Construct a fitness function based on the weight coefficients in the personalized performance requirements, use a multi-objective heuristic optimization algorithm to iteratively solve the second-order polynomial prediction model, generate a Pareto front solution set, and select the theoretically optimal combination of time-sensitive parameters from the Pareto front solution set; S5. Industrial-grade temperature field coupling test: The theoretically optimal combination of aging parameters is applied to the industrial aging furnace for pilot production. Thermocouples are installed in the central and edge areas of the material frame, as well as at the maximum and minimum wall thicknesses, to collect furnace gas temperature and material body temperature in real time, and to obtain spatial temperature difference distribution data and wall thickness temperature difference distribution data during the heating and heat preservation stages. S6. Kinetic-based thickness compensation correction: Extract the real-time temperature difference between the maximum wall thickness and the thinnest wall from the wall thickness temperature difference distribution data. When the real-time temperature difference exceeds a preset threshold and the pilot performance test results show that the thinnest wall area meets the standard while the maximum wall thickness area is under-aged, introduce the precipitation activation energy parameter of the target profile, reduce the aging temperature in the theoretically optimal aging parameter combination, and perform parameter iterative correction based on the hysteresis compensation strategy of equivalently extending the heat preservation time according to the Arrhenius equation until the corrected parameters meet the performance standards of both the thick and thin areas, and output the final personalized aging process parameters.

[0004] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. The personalized performance requirements set for this profile in step S1, which include at least two mechanical and physical performance indicators and their weighting coefficients, are as follows: Mechanical and physical properties include at least two of the following: tensile strength, yield strength, elongation after fracture, microhardness, electrical conductivity, and corrosion resistance. Personalized performance requirements include at least two of the following: target tensile strength, target yield strength, target elongation, and target electrical conductivity, as well as weighting coefficients for each performance indicator.

[0005] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. In step S2, the dynamic upper and lower limits of the aging temperature are set. The specific setting method is as follows: Upper limit of aging temperature = peak temperature of precipitated phase + Lower limit of aging temperature = characteristic peak temperature of precipitated phase - ,in, and This is a correction value determined based on the aforementioned solution supersaturation, and , The values ​​range from 10℃ to 30℃.

[0006] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. In step S3, response surface experimental design is carried out with aging temperature and heat preservation time as decision variables, and a small-sample experiment is performed. Based on the experimental results, a second-order polynomial prediction model with decision variables as independent variables and performance indicators as dependent variables is fitted and constructed. The specific steps are as follows. S31. Select the response surface experimental design type: Use a central composite design with two factors, including axial points and central point replications. The specific structure includes: Factorial part: 2² = 4 corner points ((-1, -1), (+1, -1), (-1, +1), (+1, +1)). Axial portion: 2×2=4 star points ((±α, 0), (0, ±α)), where α = (Rotational design) or α = 1 (face-centered composite design, if the range of variables cannot be exceeded). Center point: at least 3 repetitions (for estimating pure error and model bending test); Total number of experiments ,recommend ,Right now Second-rate; S32. Preparation of small-scale experimental samples: Cut small samples from the target aluminum alloy profile. The sample size is designed according to the corresponding mechanical property testing standards. Non-standard micro samples can be used to ensure that all samples come from the same batch and the same quenching state (i.e., the parameters of the previous extrusion quenching state in step S1 are consistent). For each experimental condition (temperature-time combination), at least three parallel samples should be prepared for subsequent performance testing and averaging. S33. Perform small-scale aging experiments: Use a small laboratory air-circulating aging furnace (temperature uniformity better than ±2℃), perform aging treatment according to the conditions designed in S31, record the actual furnace temperature curve, and ensure that the heating rate is similar to that of the industrial furnace (it can be set to raise the temperature from room temperature to the set temperature ≤ 15 min to reduce the aging effect during the heating stage, or use the same heating rate and record it). After the heat preservation is completed, immediately remove the furnace and air cool it to room temperature. S34. Performance Index Testing: Based on the performance indexes set in step S1, each aged sample is tested, and the arithmetic mean of the test results of parallel samples under each experimental condition is taken to obtain the response value. (j=1,…,M, where M is the number of performance metrics); S35. Establish a second-order polynomial prediction model: for each performance index Fit the following second-order polynomial models (response surfaces): ,in: Encoded values ​​for aging temperature and holding time: ranging from -1 to +1, not the original values, regression coefficients. Estimated using the least squares method, based on experimental data. , i = 1 ~ N; S36. Model quality evaluation and verification: Under the center point condition, conduct 2 to 3 additional independent experiments and compare the measured values ​​with the model predictions. If the relative error is within ±5%, the model is considered effective. If the model quality does not meet the requirements (e.g., significant lack of fit), consider adding higher-order terms, changing the response variable, narrowing the variable range, and redesigning the model. S37. Output usable prediction models: Finally, obtain a set of second-order response surface models of aging parameters and performance indices applicable to the target aluminum alloy profile within its dynamic temperature boundary.

[0007] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. In step S4, a multi-objective heuristic optimization algorithm is used to iteratively solve the second-order polynomial prediction model to generate a Pareto front solution set. The specific steps are as follows: S41. Define a multi-objective optimization problem: the decision variables are aging temperature and holding time, the objective function is the prediction model of each performance index obtained in step S3, and the optimization direction is set as follows: maximizing strength and ensuring that the elongation is not lower than the lower limit. The constraints are that the temperature is between the dynamic upper and lower limits and the time is between the set upper and lower limits. S42. The solution is obtained using a non-dominated sorting genetic algorithm (NSGA-II) with an elitist strategy: First, the algorithm parameters are initialized: population size is 50 to 200, maximum number of generations is 100 to 500, crossover probability is 0.8 to 0.9, and mutation probability is 0.5. An initial population is randomly generated in the decision space, and the predicted values ​​of all performance indicators for each individual are calculated. Then, the iterative process begins: non-dominated sorting is performed on each generation of the population, and all individuals are divided into different levels according to Pareto dominance (the first level is the non-dominated solution). Individuals within the same level are then sorted. Calculate the crowding degree (representing the sparsity around an individual); select the parent generation through a binary tournament selection (prioritizing individuals with smaller non-dominated layer numbers and higher crowding degree in the same layer), and generate offspring populations of equal size through crossover and mutation according to probability; merge the parent and offspring generations, perform non-dominated sorting and crowding degree calculation again, and select the individuals with the optimal population size as the new generation population. Repeat the above iterations until the maximum number of generations is reached or the Pareto front changes minimally over multiple generations. After the iterations are completed, extract the first layer of non-dominated solutions from the final population to form the Pareto front solution set; S43. Based on the weight coefficients of each performance index set by the user in step S1, select the theoretically optimal solution from the Pareto front solution set, find the maximum and minimum values ​​of each performance index on the Pareto front, normalize each index of each solution (maximize the index by subtracting the minimum value from the actual value and dividing by the range, minimize the index by subtracting the actual value from the maximum value and dividing by the range), and then calculate the weighted sum. The solution with the highest score is the theoretically optimal combination of time-sensitive parameters. If no weight is specified, the weights are equal by default. S44. Output the theoretically optimal aging temperature and holding time, as well as the corresponding predicted performance values, and record the Pareto front solution set for future reference.

[0008] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. In step S5, thermocouples are installed in the central and edge areas of the molded material frame, as well as at the locations of maximum and minimum wall thickness, according to the following rules: Temperature field distribution rules in the furnace space (center and edge areas): monitor the temperature uniformity in the aging furnace and identify cold and hot spots in the furnace; Layout dimensions: A three-dimensional layout method is adopted, covering the length (from the feed end to the discharge end), width (from the left to the right side), and height (from the top to the bottom) of the furnace. Central area placement: 1-2 K-type and N-type thermocouples are placed at the geometric center of the material loading area as reference points for the furnace reference temperature; Points of contact in the edge areas: Four-corner placement: Install one thermocouple at each of the four corners (top left, bottom left, top right, and bottom right) of the loading area of ​​the material frame, because corners are usually dead zones for heat circulation, where the temperature difference is most obvious; Tubular outlets and furnace doors: One tube is installed near the hot air inlet (air inlet) and the furnace door seal, respectively, to monitor the extreme temperatures in the strong convection zone and the heat dissipation zone; Temperature field layout rules for profile solid (at maximum and minimum wall thickness): Monitor the cross-sectional temperature difference of the profile during the heating and heat preservation stages, and evaluate the thermal hysteresis effect at thick walls; Monitoring points at the point of maximum wall thickness (core monitoring points): Heart area distribution The sensing end of the thermocouple is pre-embedded or drilled into the geometric center of the profile's maximum wall thickness section: at 1 / 2 wall thickness, which is the area where heat conduction is slowest and most prone to "under-aging". Surface dot distribution A thermocouple is installed close to the surface of the profile at the point of maximum wall thickness to calculate the radial temperature difference at that cross-section. ); Points of observation at minimum wall thickness (for comparison with monitoring points): Surface dot distribution A thermocouple is placed on the surface of the profile with the minimum wall thickness (thinnest rib). Since the temperature rises very quickly at the thin wall, it is very easy to over-age. The data at this point is used to evaluate the synchronicity of the thick and thin areas.

[0009] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. In step S6, the aging temperature in the theoretically optimal combination of aging parameters is reduced, and the parameters are iteratively corrected based on the hysteresis compensation strategy of equivalently extending the heat preservation time according to the Arrhenius equation, as follows; S61. Determine the triggering conditions and objectives for compensation correction: Input data: Theoretical optimal time-efficiency parameters: The global optimal solution obtained by the S4 multi-objective optimization algorithm, denoted as... ; Measured temperature difference data: The maximum temperature difference between the core at the thickest wall section and the furnace gas / thin wall section, as monitored by the S5 thermocouple. ; Intrinsic parameters of materials: precipitation activation energy of the target profile Unit: J / mol, determined by differential scanning calorimetry; When the measured temperature difference If the temperature exceeds the preset process tolerance threshold by 10°C, the hysteresis compensation strategy will be activated. S62. Reduce aging temperature To ensure that the thin-walled sections do not undergo aging (avoiding grain boundary coarsening), the overall aging temperature must be reduced, and the reduction should be proportional to the maximum temperature difference measured in the experiment. Related; Calculation formula: ,in, This is the corrected aging temperature (i.e., the target temperature that the core of the thick-walled section needs to catch up with). The theoretically optimal aging temperature, For correction factor ( This indicates that the target temperature will be directly lowered to a level close to the actual temperature achievable in the thick-walled core. S63. Equivalent extension of holding time based on the Arrhenius equation: According to the Arrhenius equation, the reaction rate constant... With temperature It exhibits an exponential relationship, especially at lower temperatures. To obtain high temperature For the same degree of reaction, extending the heat preservation time, the time correction formula is as follows: ,in This is the corrected heat preservation time (output result). The theoretically optimal heat preservation time (usually taken as...) ), It is the ideal gas constant (8.314 J / mol·K). The theoretical aging temperature used as a reference (usually taken as...) ); S64. Convergence Criterion: Condition 1: The relative deviation between the measured core hardness at the maximum wall thickness of the profile and the target hardness is less than 3%; Condition 2: The hardness difference between the maximum and minimum wall thickness of the profile is less than 5 HV; Condition 3: The difference in the heat preservation time calculated in two consecutive iterations is less than 5% (indicating that the parameters have stabilized).

[0010] The beneficial effects of this invention are as follows: By employing industrial-grade temperature field coupling testing and kinetic compensation correction based on precipitation activation energy, the long-standing technical contradiction of "over-aging of thin-walled sections and under-aging of thick-walled sections" in traditional aging processes is precisely resolved. By scientifically "reducing the aging temperature and equivalently extending the holding time," while ensuring sufficient precipitation strengthening in the thick-walled core, grain boundary coarsening and performance degradation in thin-walled regions caused by excessively high temperatures are effectively avoided. This significantly reduces the hardness and strength deviations on the profile cross-section, and substantially improves the overall quality consistency of large and complex aluminum alloy profiles. The invention abandons the traditional, crude method of blindly adjusting holding time based on manual experience, and creatively introduces the Arrhenius equation and intrinsic precipitation activation energy parameters of the material to establish a precise equivalent compensation model for temperature and time. This quantitative calculation method based on physical metallurgical principles greatly reduces the number of industrial-scale trial-and-error (pilot-scale) tests, shortens the development cycle of new processes, and significantly reduces the energy consumption and scrap rate in the production of high-value-added aluminum alloy profiles. Attached Figure Description

[0011] Figure 1 A flowchart illustrating a method for optimizing personalized process parameters in the aging treatment of aluminum alloy profiles; Detailed Implementation A method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles, such as... Figure 1 As shown, it includes the following steps: S1. Multidimensional feature acquisition: Acquire the pre-extrusion quenching state parameters of the target aluminum alloy profile, the maximum and minimum wall thickness parameters of the profile section, and the personalized performance requirements set for the profile, including at least two mechanical and physical performance indicators and their weighting coefficients. S2. Thermodynamic boundary definition: Differential scanning calorimetry is used to scan the quenched sample of the profile to determine its characteristic peak temperature of the precipitate phase. Based on the characteristic peak temperature of the precipitate phase and combined with the solid solution supersaturation in the previous extrusion quenching state parameters, the dynamic upper and lower limits of the aging temperature are set. S3. Response surface modeling: Within the set dynamic upper and lower limits of the effective temperature, a response surface experiment is designed and a small-scale experiment is performed with the aging temperature and holding time as decision variables. Based on the experimental results, a second-order polynomial prediction model is fitted and constructed with the decision variables as independent variables and the performance index as dependent variables. S4. Multi-objective optimization solution of the model: Construct a fitness function based on the weight coefficients in the personalized performance requirements, use a multi-objective heuristic optimization algorithm to iteratively solve the second-order polynomial prediction model, generate a Pareto front solution set, and select the theoretically optimal combination of time-sensitive parameters from the Pareto front solution set; S5. Industrial-grade temperature field coupling test: The theoretically optimal combination of aging parameters is applied to the industrial aging furnace for pilot production. Thermocouples are installed in the central and edge areas of the material frame, as well as at the maximum and minimum wall thicknesses, to collect furnace gas temperature and material body temperature in real time, and to obtain spatial temperature difference distribution data and wall thickness temperature difference distribution data during the heating and heat preservation stages. S6. Kinetic-based thickness compensation correction: Extract the real-time temperature difference between the maximum wall thickness and the thinnest wall from the wall thickness temperature difference distribution data. When the real-time temperature difference exceeds a preset threshold and the pilot performance test results show that the thinnest wall area meets the standard while the maximum wall thickness area is under-aged, introduce the precipitation activation energy parameter of the target profile, reduce the aging temperature in the theoretically optimal aging parameter combination, and perform parameter iterative correction based on the hysteresis compensation strategy of equivalently extending the heat preservation time according to the Arrhenius equation until the corrected parameters meet the performance standards of both the thick and thin areas, and output the final personalized aging process parameters.

[0012] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. The personalized performance requirements set for this profile in step S1, which include at least two mechanical and physical performance indicators and their weighting coefficients, are as follows: Mechanical and physical properties include at least two of the following: tensile strength, yield strength, elongation after fracture, microhardness, electrical conductivity, and corrosion resistance. Personalized performance requirements include at least two of the following: target tensile strength, target yield strength, target elongation, and target electrical conductivity, as well as weighting coefficients for each performance indicator.

[0013] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. In step S2, the dynamic upper and lower limits of the aging temperature are set. The specific setting method is as follows: Upper limit of aging temperature = peak temperature of precipitated phase + Lower limit of aging temperature = characteristic peak temperature of precipitated phase - ,in, and This is a correction value determined based on the aforementioned solution supersaturation, and , The values ​​range from 10℃ to 30℃.

[0014] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. In step S3, response surface experimental design is carried out with aging temperature and heat preservation time as decision variables, and a small-sample experiment is performed. Based on the experimental results, a second-order polynomial prediction model with decision variables as independent variables and performance indicators as dependent variables is fitted and constructed. The specific steps are as follows. S31. Select the response surface experimental design type: Use a central composite design with two factors, including axial points and central point replications. The specific structure includes: Factorial part: 2² = 4 corner points ((-1, -1), (+1, -1), (-1, +1), (+1, +1)). Axial portion: 2×2=4 star points ((±α, 0), (0, ±α)), where α = (Rotational design) or α = 1 (face-centered composite design, if the range of variables cannot be exceeded). Center point: at least 3 repetitions (for estimating pure error and model bending test); Total number of experiments ,recommend ,Right now Second-rate; S32. Preparation of small-scale experimental samples: Cut small samples from the target aluminum alloy profile. The sample size is designed according to the corresponding mechanical property testing standards. Non-standard micro samples can be used to ensure that all samples come from the same batch and the same quenching state (i.e., the parameters of the previous extrusion quenching state in step S1 are consistent). For each experimental condition (temperature-time combination), at least three parallel samples should be prepared for subsequent performance testing and averaging. S33. Perform small-scale aging experiments: Use a small laboratory air-circulating aging furnace (temperature uniformity better than ±2℃), perform aging treatment according to the conditions designed in S31, record the actual furnace temperature curve, and ensure that the heating rate is similar to that of the industrial furnace (it can be set to raise the temperature from room temperature to the set temperature ≤ 15 min to reduce the aging effect during the heating stage, or use the same heating rate and record it). After the heat preservation is completed, immediately remove the furnace and air cool it to room temperature. S34. Performance Index Testing: Based on the performance indexes set in step S1, each aged sample is tested, and the arithmetic mean of the test results of parallel samples under each experimental condition is taken to obtain the response value. (j=1,…,M, where M is the number of performance metrics); S35. Establish a second-order polynomial prediction model: for each performance index Fit the following second-order polynomial models (response surfaces): ,in: Encoded values ​​for aging temperature and holding time: ranging from -1 to +1, not the original values, regression coefficients. Estimated using the least squares method, based on experimental data. , i = 1 ~ N; S36. Model quality evaluation and verification: Under the center point condition, conduct 2 to 3 additional independent experiments and compare the measured values ​​with the model predictions. If the relative error is within ±5%, the model is considered effective. If the model quality does not meet the requirements (e.g., significant lack of fit), consider adding higher-order terms, changing the response variable, narrowing the variable range, and redesigning the model. S37. Output usable prediction models: Finally, obtain a set of second-order response surface models of aging parameters and performance indices applicable to the target aluminum alloy profile within its dynamic temperature boundary.

[0015] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. In step S4, a multi-objective heuristic optimization algorithm is used to iteratively solve the second-order polynomial prediction model to generate a Pareto front solution set. The specific steps are as follows: S41. Define a multi-objective optimization problem: the decision variables are aging temperature and holding time, the objective function is the prediction model of each performance index obtained in step S3, and the optimization direction is set as follows: maximizing strength and ensuring that the elongation is not lower than the lower limit. The constraints are that the temperature is between the dynamic upper and lower limits and the time is between the set upper and lower limits. S42. The solution is obtained using a non-dominated sorting genetic algorithm (NSGA-II) with an elitist strategy: First, the algorithm parameters are initialized: population size is 50 to 200, maximum number of generations is 100 to 500, crossover probability is 0.8 to 0.9, and mutation probability is 0.5. An initial population is randomly generated in the decision space, and the predicted values ​​of all performance indicators for each individual are calculated. Then, the iterative process begins: non-dominated sorting is performed on each generation of the population, and all individuals are divided into different levels according to Pareto dominance (the first level is the non-dominated solution). Individuals within the same level are then sorted. Calculate the crowding degree (representing the sparsity around an individual); select the parent generation through a binary tournament selection (prioritizing individuals with smaller non-dominated layer numbers and higher crowding degree in the same layer), and generate offspring populations of equal size through crossover and mutation according to probability; merge the parent and offspring generations, perform non-dominated sorting and crowding degree calculation again, and select the individuals with the optimal population size as the new generation population. Repeat the above iterations until the maximum number of generations is reached or the Pareto front changes minimally over multiple generations. After the iterations are completed, extract the first layer of non-dominated solutions from the final population to form the Pareto front solution set; S43. Based on the weight coefficients of each performance index set by the user in step S1, select the theoretically optimal solution from the Pareto front solution set, find the maximum and minimum values ​​of each performance index on the Pareto front, normalize each index of each solution (maximize the index by subtracting the minimum value from the actual value and dividing by the range, minimize the index by subtracting the actual value from the maximum value and dividing by the range), and then calculate the weighted sum. The solution with the highest score is the theoretically optimal combination of time-sensitive parameters. If no weight is specified, the weights are equal by default. S44. Output the theoretically optimal aging temperature and holding time, as well as the corresponding predicted performance values, and record the Pareto front solution set for future reference.

[0016] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. In step S5, thermocouples are installed in the central and edge areas of the molded material frame, as well as at the locations of maximum and minimum wall thickness, according to the following rules: Temperature field distribution rules in the furnace space (center and edge areas): monitor the temperature uniformity in the aging furnace and identify cold and hot spots in the furnace; Layout dimensions: A three-dimensional layout method is adopted, covering the length (from the feed end to the discharge end), width (from the left to the right side), and height (from the top to the bottom) of the furnace. Central area placement: 1-2 K-type and N-type thermocouples are placed at the geometric center of the material loading area as reference points for the furnace reference temperature; Points of contact in the edge areas: Four-corner placement: Install one thermocouple at each of the four corners (top left, bottom left, top right, and bottom right) of the loading area of ​​the material frame, because corners are usually dead zones for heat circulation, where the temperature difference is most obvious; Tubular outlets and furnace doors: One tube is installed near the hot air inlet (air inlet) and the furnace door seal, respectively, to monitor the extreme temperatures in the strong convection zone and the heat dissipation zone; Temperature field layout rules for profile solid (at maximum and minimum wall thickness): Monitor the cross-sectional temperature difference of the profile during the heating and heat preservation stages, and evaluate the thermal hysteresis effect at thick walls; Monitoring points at the point of maximum wall thickness (core monitoring points): Heart area distribution The sensing end of the thermocouple is pre-embedded or drilled into the geometric center of the profile's maximum wall thickness section: at 1 / 2 wall thickness, which is the area where heat conduction is slowest and most prone to "under-aging". Surface dot distribution A thermocouple is installed close to the surface of the profile at the point of maximum wall thickness to calculate the radial temperature difference at that cross-section. ); Points of observation at minimum wall thickness (for comparison with monitoring points): Surface dot distribution A thermocouple is placed on the surface of the profile with the minimum wall thickness (thinnest rib). Since the temperature rises very quickly at the thin wall, it is very easy to over-age. The data at this point is used to evaluate the synchronicity of the thick and thin areas.

[0017] Furthermore, a method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles is proposed. In step S6, the aging temperature in the theoretically optimal combination of aging parameters is reduced, and the parameters are iteratively corrected based on the hysteresis compensation strategy of equivalently extending the heat preservation time according to the Arrhenius equation, as follows; S61. Determine the triggering conditions and objectives for compensation correction: Input data: Theoretical optimal time-efficiency parameters: The global optimal solution obtained by the S4 multi-objective optimization algorithm, denoted as... ; Measured temperature difference data: The maximum temperature difference between the core at the thickest wall section and the furnace gas / thin wall section, as monitored by the S5 thermocouple. ; Intrinsic parameters of materials: precipitation activation energy of the target profile Unit: J / mol, determined by differential scanning calorimetry; When the measured temperature difference If the temperature exceeds the preset process tolerance threshold by 10°C, the hysteresis compensation strategy will be activated. S62. Reduce aging temperature To ensure that the thin-walled sections do not undergo aging (avoiding grain boundary coarsening), the overall aging temperature must be reduced, and the reduction should be proportional to the maximum temperature difference measured in the experiment. Related; Calculation formula: ,in, This is the corrected aging temperature (i.e., the target temperature that the core of the thick-walled section needs to catch up with). The theoretically optimal aging temperature, For correction factor ( This indicates that the target temperature will be directly lowered to a level close to the actual temperature achievable in the thick-walled core. S63. Equivalent extension of holding time based on the Arrhenius equation: According to the Arrhenius equation, the reaction rate constant... With temperature It exhibits an exponential relationship, especially at lower temperatures. To obtain high temperature For the same degree of reaction, extending the heat preservation time, the time correction formula is as follows: ,in This is the corrected heat preservation time (output result). The theoretically optimal heat preservation time (usually taken as...) ), It is the ideal gas constant (8.314 J / mol·K). The theoretical aging temperature used as a reference (usually taken as...) ); S64. Convergence Criterion: Condition 1: The relative deviation between the measured core hardness at the maximum wall thickness of the profile and the target hardness is less than 3%; Condition 2: The hardness difference between the maximum and minimum wall thickness of the profile is less than 5 HV; Condition 3: The difference in the heat preservation time calculated in two consecutive iterations is less than 5% (indicating that the parameters have stabilized).

[0018] Example 2 This embodiment focuses on a 6063 aluminum alloy building curtain wall profile, developing customized process parameters for its aging treatment. The profile has a maximum wall thickness of 12 mm and a minimum wall thickness of 2 mm. The customized performance requirements are: tensile strength ≥ 210 MPa, yield strength ≥ 180 MPa, and elongation after fracture ≥ 8%, with a weighting coefficient of 0.6 for the strength index and 0.4 for the elongation index.

[0019] S1. Multidimensional Feature Acquisition: Pre-extrusion quenching state parameters of the profile were obtained from the extrusion production line. The quenching method was online air cooling + water mist combined quenching. The measured average cooling rate of the profile exiting the quenching section was 200 ℃ / s, corresponding to a high level of solution supersaturation (the conductivity of the quenched state was measured to be 32% IACS, confirming the high supersaturation). The maximum wall thickness was recorded as 12 mm, and the minimum wall thickness as 2 mm. User-specific performance requirements are as described above. S2. Thermodynamic Boundary Delineation: Approximately 15 mg of sample was cut from the quenched sample of the profile. Differential scanning calorimetry (DSC) was used to scan from room temperature to 450 °C at a heating rate of 10 °C / min. A distinct exothermic peak appeared on the DSC curve, with a peak temperature of 210 °C. This is the characteristic peak temperature Tp of the precipitated phase (β″ phase). Considering the high solid solution supersaturation (which is conducive to nucleation at lower temperatures), the dynamic lower limit of the aging temperature was set as Tp - 30 °C = 180 °C, and the dynamic upper limit was set as Tp + 20 °C = 230 °C. The initial holding time range was set to 1 h to 8 h. S3. Response Surface Modeling: Using aging temperature (180–230 °C) and holding time (1–8 h) as decision variables, a central composite design (CCD) was adopted, with three levels for each variable: temperature codes -1, 0, and +1 corresponding to 180, 205, and 230 °C; time codes corresponding to 1, 4.5, and 8 h. The star point α = √2, and the center point was repeated three times, for a total of 13 experimental groups. Standard tensile specimens were cut from the same batch of quenched profiles (processed into non-standard miniature specimens according to GB / T 228.1, with gauge length dimensions of 25 mm × 6 mm × 2 mm). Aging was performed in a small laboratory air-circulating aging furnace under 13 different conditions, with 3 parallel specimens per group. After heat treatment, the specimens were air-cooled. Tensile strength Rm, yield strength Rp0.2, and elongation at break A were measured, and the average values ​​were taken. A second-order polynomial model was fitted to each performance index using Design-Expert software. Taking tensile strength as an example, the model in the encoding space was obtained as follows: Rm = 215.3 + 8.2 × temperature + 5.6 × time – 4.1 × temperature² – 2.3 × time² + 1.9 × temperature × time. Analysis of variance showed that the model p-value < 0.001, the lack-of-fit term p = 0.21, R² = 0.94, and the adjusted R² = 0.91. The center point verification showed that the relative error between the predicted and measured values ​​was within ±3%, indicating the model was effective. S4. Multi-objective optimization solution: The NSGA-II algorithm is used for multi-objective optimization: population size 100, number of generations 200, crossover probability 0.85, mutation probability 0.5. The optimization objectives are to maximize tensile strength, maximize yield strength, and maximize elongation (since the elongation requirement is ≥8%, it is actually desired to be as high as possible). Constraints: temperature 180~230 ℃, time 1~8h. After iteration, a Pareto front solution set is obtained, with a total of 23 non-dominated solutions. According to the performance weights (strength 0.6, elongation 0.4, yield strength and tensile strength are highly correlated, so they are combined into a comprehensive strength score), a comprehensive score is calculated for each solution. The solution with the highest weighted sum after normalization is: aging temperature 195 ℃, holding time 5.2 h, predicted tensile strength 221 MPa, yield strength 188 MPa, elongation 9.2%. The theoretically optimal parameter combination (195℃×5.2h) is output. S5. Industrial-grade temperature field coupling test: The theoretically optimal parameters were applied to pilot production: The material frame with the profile was pushed into a 15-ton industrial air circulation aging furnace. Three K-type thermocouples were arranged in the central and edge areas of the material frame. Thermocouples were also drilled and embedded at the maximum wall thickness (12 mm) and minimum wall thickness (2 mm) of the profile cross-section. The furnace gas temperature and the actual temperature of the profile were recorded in real time at a sampling frequency of 1 Hz. The target furnace gas temperature was set at 195℃ and the holding time was 5.2 h (starting from when the furnace gas reached 190℃). The actual measurements showed that: during the heating stage, the temperature at the maximum wall thickness lagged behind the furnace gas by about 12 min, and the temperature at the minimum wall thickness lagged behind by about 3 min. During the holding stage, the stable temperature at the maximum wall thickness was 188-191℃, and the temperature at the minimum wall thickness was 194-197℃. The spatial temperature difference (from the furnace edge to the center) was about 4℃, and the maximum wall thickness temperature difference was 7℃ (occurring in the early stage of holding). S6. Kinetic-based thickness compensation correction: Performance testing was conducted on samples taken after pilot testing: at the minimum wall thickness (thin-walled region), the tensile strength was 218 MPa, the yield strength was 185 MPa, and the elongation was 9.0%, all meeting the standards; at the maximum wall thickness (thick-walled region), the tensile strength was 196 MPa, the yield strength was 162 MPa, and the elongation was 11.5%, failing to meet the strength standards and exhibiting under-aging. The real-time maximum wall thickness temperature difference of 7℃ exceeded the preset threshold of 5℃, meeting the correction conditions; the precipitation activation energy of 6063 alloy was obtained from literature review: Q=145 kJ / mol. The cooling step size was set to 5℃, the initial temperature was 195℃, and the first round of correction: new temperature T1=190℃. According to the Arrhenius equivalent time formula (gas constant R = 8.314 J / (mol·K), temperature converted to Kelvin), the equivalent holding time is calculated as: t1 = 5.2 × exp[(145000 / 8.314)×(1 / (190+273.15) - 1 / (195+273.15))] ≈ 6.1 h; a second pilot test was conducted at 190℃ for 6.1 h, and the temperature was measured again: the wall thickness temperature difference decreased to 5℃, and the performance was tested: the tensile strength of the thick-walled region was 208 MPa, the yield strength was 175 MPa (still slightly lower than 180), the strength of the thin-walled region was 215 MPa, and the elongation was 8.5%. The yield strength of the thick-walled region is still under-aged; second round of correction: T2=185℃, recalculate the equivalent time: t2 = 6.1 × exp[(145000 / 8.314)×(1 / (185+273.15) - 1 / (190+273.15))]≈ 7.3 h; third pilot test (185℃×7.3h) was conducted, and the temperature measurement showed a wall thickness temperature difference of 3℃, which is below the threshold. Performance testing: tensile strength of the thick-walled region is 215 MPa, yield strength is 182 MPa, and elongation is 9.0%; tensile strength of the thin-walled region is 220 MPa, yield strength is 186 MPa, and elongation is 8.2%, all indicators meet the requirements; stop iteration, output the final personalized aging process parameters: aging temperature 185℃, holding time 7.3 hours, this parameter is included in the process standard of this profile, and subsequent mass production will be carried out according to this.

Claims

1. A method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles, characterized in that, Includes the following steps: S1. Multidimensional feature acquisition: Acquire the pre-extrusion quenching state parameters of the target aluminum alloy profile, the maximum and minimum wall thickness parameters of the profile section, and the personalized performance requirements set for the profile, including at least two mechanical and physical performance indicators and their weighting coefficients. S2. Thermodynamic boundary definition: Differential scanning calorimetry is used to scan the quenched sample of the profile to determine its characteristic peak temperature of the precipitate phase. Based on the characteristic peak temperature of the precipitate phase and combined with the solid solution supersaturation in the previous extrusion quenching state parameters, the dynamic upper and lower limits of the aging temperature are set. S3. Response surface modeling: Within the set dynamic upper and lower limits of the effective temperature, a response surface experiment is designed and a small-scale experiment is performed with the aging temperature and holding time as decision variables. Based on the experimental results, a second-order polynomial prediction model is fitted and constructed with the decision variables as independent variables and the performance index as dependent variables. S4. Multi-objective optimization solution of the model: Construct a fitness function based on the weight coefficients in the personalized performance requirements, use a multi-objective heuristic optimization algorithm to iteratively solve the second-order polynomial prediction model, generate a Pareto front solution set, and select the theoretically optimal combination of time-sensitive parameters from the Pareto front solution set; S5. Industrial-grade temperature field coupling test: The theoretically optimal combination of aging parameters is applied to the industrial aging furnace for pilot production. Thermocouples are installed in the central and edge areas of the material frame, as well as at the maximum and minimum wall thicknesses, to collect furnace gas temperature and material body temperature in real time, and to obtain spatial temperature difference distribution data and wall thickness temperature difference distribution data during the heating and heat preservation stages. S6. Kinetic-based thickness compensation correction: Extract the real-time temperature difference between the maximum wall thickness and the thinnest wall from the wall thickness temperature difference distribution data. When the real-time temperature difference exceeds a preset threshold and the pilot performance test results show that the thinnest wall area meets the standard while the maximum wall thickness area is under-aged, introduce the precipitation activation energy parameter of the target profile, reduce the aging temperature in the theoretically optimal aging parameter combination, and perform parameter iterative correction based on the hysteresis compensation strategy of equivalently extending the heat preservation time according to the Arrhenius equation until the corrected parameters meet the performance standards of both the thick and thin areas, and output the final personalized aging process parameters.

2. The method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles as described in claim 1, characterized in that, The personalized performance requirements set for this profile in step S1, which include at least two mechanical and physical performance indicators and their weighting coefficients, are as follows: Mechanical and physical properties include at least two of the following: tensile strength, yield strength, elongation after fracture, microhardness, electrical conductivity, and corrosion resistance. Personalized performance requirements include at least two of the following: target tensile strength, target yield strength, target elongation, and target electrical conductivity, as well as weighting coefficients for each performance indicator.

3. The method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles as described in claim 1, characterized in that, In step S2, the dynamic upper and lower limits of the aging temperature are set. The specific setting method is as follows: Upper limit of aging temperature = peak characteristic temperature of precipitated phase + Lower limit of aging temperature = characteristic peak temperature of precipitated phase - ,in, and This is a correction value determined based on the aforementioned solid solution supersaturation, and , The values ​​range from 10℃ to 30℃.

4. The method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles as described in claim 1, characterized in that, In step S3, response surface experimental design is carried out with aging temperature and heat preservation time as decision variables, and a small-sample experiment is performed. Based on the experimental results, a second-order polynomial prediction model with decision variables as independent variables and performance indicators as dependent variables is fitted and constructed. The specific steps are as follows. S31. Select the response surface experimental design type: Use a central composite design with two factors, including axial points and central point replications. The specific structure includes: Factorial part: 2² = 4 corner points ((-1, -1), (+1, -1), (-1, +1), (+1, +1)). Axial portion: 2×2=4 star points ((±α, 0), (0, ±α)), where α = Rotational design; Center point: 3 repetitions; Total number of experiments ,recommend ,Right now Second-rate; S32. Preparation of small-scale experimental samples: Cut small samples from the target aluminum alloy profile. The sample size is designed according to the corresponding mechanical property testing standards. Non-standard micro-samples can be used. Ensure that all samples come from the same batch and the same quenching state. S33. Perform small-scale aging experiments: Use a small laboratory air-circulating aging furnace and perform aging treatments according to the conditions designed in S31. Record the actual furnace temperature curve to ensure that the heating rate is similar to that of an industrial furnace. After the heat preservation is completed, immediately remove the furnace and air-cool it to room temperature. S34. Performance Index Testing: Based on the performance indexes set in step S1, each aged sample is tested, and the arithmetic mean of the test results of parallel samples under each experimental condition is taken to obtain the response value. j=1,…,M, where M is the number of performance indicators; S35. Establish a second-order polynomial prediction model: for each performance index Fit the following second-order polynomial models respectively: ,in: Encoded values ​​for aging temperature and holding time: ranging from -1 to +1, not the original values, regression coefficients. Estimated using the least squares method, based on experimental data. , i = 1 ~ N; S36. Model quality evaluation and verification: Under the condition of the center point, conduct 2 to 3 additional independent experiments, compare the measured values ​​with the model prediction values, and if the relative error is within ±5%, the model is considered to be effective. If the model quality does not meet the requirements, consider adding higher-order terms, changing the response variables, narrowing the variable range and redesigning the model. S37. Output usable prediction models: Finally, obtain a set of second-order response surface models of aging parameters and performance indices applicable to the target aluminum alloy profile within its dynamic temperature boundary.

5. The method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles as described in claim 1, characterized in that, In step S4, a multi-objective heuristic optimization algorithm is used to iteratively solve the second-order polynomial prediction model to generate a Pareto front solution set. The specific steps are as follows: S41. Define a multi-objective optimization problem: the decision variables are aging temperature and holding time, the objective function is the prediction model of each performance index obtained in step S3, and the optimization direction is set as follows: maximizing strength and ensuring that the elongation is not lower than the lower limit. The constraints are that the temperature is between the dynamic upper and lower limits and the time is between the set upper and lower limits. S42. The non-dominated sorting genetic algorithm NSGA-II with elitist strategy is used to solve the problem: and the first layer of non-dominated solutions is extracted to form the Pareto front solution set. S43. Based on the weight coefficients of each performance index set by the user in step S1, select the theoretically optimal solution from the Pareto front solution set, find the maximum and minimum values ​​of each performance index on the Pareto front, normalize each index of each solution, and calculate the weighted sum. The solution with the highest score is the theoretically optimal combination of time-efficiency parameters. If no weight is specified, the weights are equal by default. S44. Output the theoretically optimal aging temperature and holding time, as well as the corresponding predicted performance values, and record the Pareto front solution set for future reference.

6. The method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles as described in claim 1, characterized in that, In step S5, thermocouples are installed in the central and edge areas of the molded material frame, as well as at the locations of maximum and minimum wall thickness, according to the following rules: Temperature field distribution rules in the furnace space: monitor the temperature uniformity inside the aging furnace and identify cold and hot spots in the furnace; Layout dimensions: A three-dimensional layout method is adopted, covering the length, width, and height of the furnace; Central area placement: 1-2 K-type and N-type thermocouples are placed at the geometric center of the material loading area as reference points for the furnace reference temperature; Points of contact in the edge areas: Four-corner placement: One thermocouple is placed at each of the four corners of the loading area of ​​the material frame: upper left, lower left, upper right, and lower right. Tubular outlets and furnace doors: One tube is installed near the hot air inlet and the furnace door seal, respectively, to monitor the extreme temperatures in the strong convection zone and the heat dissipation zone; Temperature field layout rules for profiles: Monitor the cross-sectional temperature difference of the profile during the heating and heat preservation stages, and evaluate the thermal hysteresis effect at thick-walled sections; Points of placement at the maximum wall thickness: Heart area distribution The sensing end of the thermocouple is pre-embedded or drilled into the geometric center of the profile's maximum wall thickness section: at 1 / 2 wall thickness. Surface dot distribution A thermocouple is installed close to the surface of the profile at the point of maximum wall thickness to calculate the radial temperature difference at that cross-section. ; Points to be placed at the minimum wall thickness: Surface dot distribution One thermocouple is placed on the surface of the profile with the smallest wall thickness. Since the temperature rises very quickly at the thin wall, it is very easy to over-aging. The data at this point is used to evaluate the synchronicity of the thick and thin areas.

7. The method for optimizing personalized process parameters for aging treatment of aluminum alloy profiles as described in claim 1, characterized in that, In step S6, the aging temperature in the theoretically optimal combination of aging parameters is reduced, and the parameters are iteratively corrected based on the hysteresis compensation strategy of equivalently extending the heat preservation time according to the Arrhenius equation, as follows; S61. Determine the triggering conditions and objectives for compensation correction: Input data: Theoretical optimal time-efficiency parameters: The global optimal solution obtained by the S4 multi-objective optimization algorithm, denoted as... ; Measured temperature difference data: The maximum temperature difference between the core at the thickest wall section and the furnace gas / thin wall section, as monitored by the S5 thermocouple. ; Intrinsic parameters of materials: precipitation activation energy of the target profile Unit: J / mol, determined by differential scanning calorimetry; When the measured temperature difference If the temperature exceeds the preset process tolerance threshold by 10°C, the hysteresis compensation strategy will be activated. S62. Reduce aging temperature To ensure that the thin-walled sections do not undergo aging, the overall aging temperature needs to be reduced. The reduction should be proportional to the maximum temperature difference measured in the actual measurements. Related; Calculation formula: ,in, The corrected aging temperature. The theoretically optimal aging temperature, For correction factor, This means that the target temperature will be directly lowered to a level close to the actual temperature that the thick-walled core can achieve. S63. Equivalent extension of holding time based on the Arrhenius equation: According to the Arrhenius equation, the reaction rate constant... With temperature It exhibits an exponential relationship, especially at lower temperatures. To obtain high temperature For the same degree of reaction, extending the heat preservation time, the time correction formula is as follows: ,in This is the corrected heat preservation time. For the theoretically optimal heat preservation time: take , Let be the ideal gas constant. The theoretical aging temperature used as a benchmark: (take) ; S64. Convergence Criterion: Condition 1: The relative deviation between the measured core hardness at the maximum wall thickness of the profile and the target hardness is less than 3%; Condition 2: The hardness difference between the maximum and minimum wall thickness of the profile is less than 5 HV; Condition 3: The difference in insulation time calculated in two consecutive iterations is less than 5%.

Images