Intelligent Optimization Methods and Systems for Process Parameters of Galvanized Steel Sheets

Abstract

This application provides an intelligent optimization method and system for galvanized steel sheet process parameters. The method includes: measuring the thickness of different regions of the already formed coating on the surface of the galvanized steel sheet to obtain the final coating thickness distribution data; generating heat transfer efficiency correction parameters for each heating zone of the zinc bath based on the temperature distribution data and the final coating thickness distribution data; and optimizing the galvanizing time and temperature distribution data using a pre-trained first process parameter intelligent optimization model based on the heat transfer efficiency correction parameters for all heating zones of the zinc bath, and optimizing the electrode pressure parameters and resistance spot welding parameters during the welding stage using a pre-trained second process parameter intelligent optimization model, so that the estimated quality value of the galvanized steel sheet is greater than a preset quality threshold. The technical solution provided in this application intelligently optimizes the galvanizing process and welding parameters through dual models, dynamically adjusting the production line based on real-time data to ensure stable and compliant galvanized steel sheet quality.

Description

Technical Field

[0001] This application relates to the field of galvanized steel sheet production technology, and in particular to an intelligent optimization method and system for galvanized steel sheet process parameters. Background Technology

[0002] In the continuous industrial production of galvanized steel sheets, the heat transfer efficiency and process parameters of the zinc bath directly determine the uniformity of the coating and the quality of the weld. Current production lines require precise control of galvanizing time, temperature distribution in the heating zone, and welding parameters to achieve dynamic stability of the coating thickness and consistency of the mechanical properties of the galvanized steel sheets. Because coating formation involves multi-physics coupling and the welding process window is complex, a multi-parameter collaborative optimization method is urgently needed to overcome overall quality fluctuations caused by adjustments to local parameters, while simultaneously reducing energy consumption and trial-and-error costs.

[0003] Existing technology establishes a dynamic model of heat conduction in a zinc bath to simulate the spatial influence of heating zone temperature on coating growth rate, and uses a genetic algorithm to optimize the zinc plating time and target temperature range. The model uses historical production data to calibrate thermal boundary conditions, solves the temperature field distribution in real time using the finite difference method, and finally recommends a local temperature control parameter adjustment scheme to improve coating uniformity.

[0004] Existing technologies have low computational efficiency, making it difficult to meet the real-time decision-making needs of continuous production lines. Their heat conduction models do not deeply couple the actual zinc liquid flow state and element diffusion effects in production, resulting in temperature field prediction deviations accumulating over time. Single-objective optimization strategies only focus on coating consistency indicators and do not form a closed-loop synergy with welding process parameters, which may lead to the optimized temperature control scheme deteriorating the bonding strength of the weld interface or introducing brittle intergranular phases, ultimately creating a bottleneck in overall quality improvement. Summary of the Invention

[0005] This application provides an intelligent optimization method and system for galvanized steel sheet process parameters to solve the problems of poor quality stability of galvanized steel sheets caused by low computational efficiency, lack of multi-physics coupling, and insufficient coordination of multiple process parameters in the prior art.

[0006] In a first aspect, this application provides an intelligent optimization method for process parameters of galvanized steel sheet, including: in the galvanized steel sheet production process, acquiring the first galvanizing time and temperature distribution data of different heating zones in the zinc bath;

[0007] Thickness measurements were performed on different areas of the coating already formed on the surface of the galvanized steel sheet to obtain the final coating thickness distribution data.

[0008] Based on the temperature distribution data and the final coating thickness distribution data, heat conduction efficiency correction parameters for each heating zone of the zinc bath are generated.

[0009] Based on the heat conduction efficiency correction parameters of all heating zones in the zinc bath, the galvanizing time and temperature distribution data are optimized using a pre-trained first process parameter intelligent optimization model, and the electrode pressure parameters and resistance spot welding parameters in the welding stage are optimized using a pre-trained second process parameter intelligent optimization model, so that the estimated quality value of the galvanized steel sheet is greater than the preset quality threshold.

[0010] Optionally, the step of optimizing the galvanizing time and temperature distribution data using a pre-trained intelligent optimization model of the first process parameters based on the heat transfer efficiency correction parameters of all heating zones in the zinc bath includes:

[0011] Based on the heat conduction efficiency correction parameters of all heating zones in the zinc bath, the temperature gradient parameters, heat conduction response time, and dynamic thermal balance index are calculated respectively to generate a first set of parameters for characterizing the thermal field stability of the zinc bath.

[0012] The first parameter set is input into the first process parameter intelligent optimization model, and the galvanizing time correction and temperature distribution compensation vector are generated through the predefined heat conduction dynamic response function in the first process parameter intelligent optimization model.

[0013] The galvanizing time is optimized based on the galvanizing time correction amount to obtain a second galvanizing time. The power distribution ratio of each heating zone in the zinc bath is dynamically adjusted based on the temperature distribution compensation vector to optimize the temperature distribution data.

[0014] Optionally, the step of using a pre-trained intelligent optimization model for second process parameters to optimize the electrode pressure parameters and resistance spot welding parameters during the welding stage, so as to make the estimated quality value of the galvanized steel sheet greater than a preset quality threshold, includes:

[0015] Determine the welding area of ​​the galvanized steel sheet, and select the heat transfer efficiency correction parameter of the target heating area from the heat transfer efficiency correction parameters of all heating areas of the zinc bath. The target heating area is the area in the zinc bath corresponding to the welding area in the galvanized steel sheet production process.

[0016] Based on the heat conduction efficiency correction parameters of the corresponding heating zone, calculate the temperature fluctuation amplitude, electrode pressure deviation and resistance spot welding fluctuation coefficient to generate a second set of parameters for characterizing the thermal effect of the welding zone.

[0017] The second parameter set is input into the second process parameter intelligent optimization model, and the electrode pressure correction amount and resistance spot welding parameter compensation vector are generated by the predefined welding dynamic compensation function in the second process parameter intelligent optimization model.

[0018] The electrode pressure parameters during the welding stage are optimized based on the electrode pressure correction amount, and the welding pulse duration and interval are dynamically adjusted according to the resistance spot welding parameter compensation vector to optimize the resistance spot welding parameters so that the estimated quality value of the galvanized steel sheet is greater than the preset quality threshold.

[0019] Optionally, the step of inputting the first parameter set into the first process parameter intelligent optimization model, and generating a galvanizing time correction and temperature distribution compensation vector through a predefined heat conduction dynamic response function in the first process parameter intelligent optimization model, includes:

[0020] The first parameter set is input into the first process parameter intelligent optimization model. The first process parameter intelligent optimization model takes the ratio between the axial component of the temperature gradient parameter in the first parameter set and the maximum allowable axial temperature gradient of the zinc bath as the target ratio. The target ratio is combined with the heat conduction response time to obtain the axial gradient weight factor. Based on the dynamic thermal balance index and temperature fluctuation coefficient in the first parameter set, the radial offset attenuation coefficient is calculated using the first exponential attenuation model.

[0021] Construct a dynamic thermal field feature vector composed of the axial gradient weighting factor and the radial offset attenuation coefficient;

[0022] Based on the dynamic thermal field characteristic vector and the galvanizing time fluctuation coefficient, the galvanizing time correction and temperature distribution compensation vector are calculated using the heat conduction dynamic response function.

[0023] Optionally, the step of inputting the second parameter set into the second process parameter intelligent optimization model, and generating an electrode pressure correction and resistance spot welding parameter compensation vector through a predefined welding dynamic compensation function in the second process parameter intelligent optimization model, includes:

[0024] The second parameter set is input into the second process parameter intelligent optimization model. The second process parameter intelligent optimization model obtains the axial gradient dynamic weight factor by using the weight factor calculation formula based on the temperature fluctuation amplitude and the axial component of the temperature gradient parameter in the second parameter set. Based on the electrode pressure deviation and resistance spot welding fluctuation coefficient in the second parameter set, the impedance change attenuation coefficient is calculated using the second exponential attenuation model.

[0025] Construct a welding thermal effect feature vector composed of the axial gradient dynamic weighting factor and the impedance change attenuation coefficient;

[0026] Based on the welding heat-affected characteristic vector and the resistance spot welding fluctuation amplitude, the electrode pressure correction and resistance spot welding parameter compensation vector are calculated using the welding dynamic compensation function.

[0027] Optionally, the step of measuring the thickness of different areas of the coating already formed on the surface of the galvanized steel sheet to obtain the final coating thickness distribution data includes:

[0028] The surface of the galvanized steel sheet is divided into several inspection areas according to a preset process grid, and the galvanized steel sheet surface is the coating.

[0029] The spectral acquisition module of each detection area is triggered by a multispectral sensor array to obtain the raw data of the spectral reflectance intensity of each detection area.

[0030] Based on the original spectral reflectance intensity data of each detection area, the coating thickness of each detection area is dynamically calibrated to generate initial coating thickness distribution data.

[0031] By using the gradient difference in reflection intensity between adjacent detection areas, interpolation correction is performed on the thickness jump anomalies in the initial coating thickness distribution data to generate corrected coating thickness distribution data.

[0032] The corrected coating thickness distribution data is subjected to regional consistency verification to obtain the verified coating thickness data.

[0033] The verified coating thickness data is spatially superimposed with the original spectral reflectance intensity data of all detection areas to generate the final coating thickness distribution data.

[0034] Optionally, generating heat transfer efficiency correction parameters for each heating zone of the zinc bath based on the temperature distribution data and the final coating thickness distribution data includes:

[0035] The temperature distribution data is divided into a temperature gradient feature set according to the axial and radial distribution of the zinc bath, and the final coating thickness distribution data is divided into a thickness difference feature set according to the preset process grid.

[0036] Based on the mean axial temperature gradient corresponding to the temperature gradient feature set and the standard deviation of regional thickness corresponding to the thickness difference feature set, a dynamic heat conduction feature vector composed of axial gradient weighting factor and thickness standard deviation compensation coefficient is constructed.

[0037] The dynamic heat conduction feature vector is input into a predefined heat conduction efficiency correction function. The heat conduction efficiency correction parameter is generated by the exponential decay term of the radial temperature offset amplitude and the piecewise linear correction term of the thermal field fluctuation coefficient in the heat conduction efficiency correction function.

[0038] Secondly, this application provides an intelligent optimization system for galvanized steel sheet process parameters, including:

[0039] The acquisition module is used to acquire the first galvanizing time and the temperature distribution data of different heating zones in the zinc bath during the galvanized steel sheet production process.

[0040] The detection module is used to detect the thickness of different areas of the coating already formed on the surface of the galvanized steel sheet, and obtain the final coating thickness distribution data.

[0041] The calculation module is used to generate heat conduction efficiency correction parameters for each heating zone of the zinc bath based on the temperature distribution data and the final coating thickness distribution data.

[0042] The optimization module is used to correct parameters based on the heat conduction efficiency of all heating zones in the zinc bath, optimize the galvanizing time and temperature distribution data using a pre-trained first process parameter intelligent optimization model, and optimize the electrode pressure parameters and resistance spot welding parameters in the welding stage using a pre-trained second process parameter intelligent optimization model, so that the estimated quality value of the galvanized steel sheet is greater than the preset quality threshold.

[0043] Thirdly, this application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are to be invoked and executed by the processing component to implement an intelligent optimization method for galvanized steel sheet process parameters as described in any of the first aspects.

[0044] Fourthly, this application provides a computer storage medium storing a computer program, which, when executed by a computer, implements an intelligent optimization method for galvanized steel sheet process parameters as described in any of the first aspects.

[0045] This application provides an intelligent optimization method for process parameters of galvanized steel sheets. The method includes: acquiring first galvanizing time and temperature distribution data of different heating zones within the zinc bath during the galvanized steel sheet production process; performing thickness detection on different areas of the already formed coating on the surface of the galvanized steel sheet to obtain final coating thickness distribution data; generating heat transfer efficiency correction parameters for each heating zone of the zinc bath based on the temperature distribution data and the final coating thickness distribution data; and optimizing the galvanizing time and temperature distribution data using a pre-trained first process parameter intelligent optimization model based on the heat transfer efficiency correction parameters for all heating zones of the zinc bath, and optimizing the electrode pressure parameters and resistance spot welding parameters during the welding stage using a pre-trained second process parameter intelligent optimization model, so that the estimated quality value of the galvanized steel sheet is greater than a preset quality threshold.

[0046] This application constructs a digital mapping of the entire production process by real-time acquisition of data on galvanizing time, temperature distribution in multiple heating zones of the zinc bath, and coating thickness distribution. This enables dynamic monitoring and anomaly localization of the process status, overcoming the lag and localization defects of traditional manual sampling inspection, and improving the spatiotemporal resolution and process transparency of data acquisition. Based on the coupling relationship between temperature and coating thickness, the actual heat transfer efficiency of each heating zone is inferred, generating dynamic correction parameters and overcoming the limitations of traditional static heat transfer models. By integrating multi-physical field effects such as heat transfer, flow, and phase change, the accuracy of temperature field prediction is improved, reducing the mismatch between coating growth rate and temperature distribution. A dual intelligent optimization model is adopted to achieve parameter coordination across process sections, resolving the coupling conflict between coating uniformity and weld interface strength. Through a multi-objective optimization strategy for quality estimates, coating adhesion, corrosion resistance, and weld joint strength are simultaneously improved, ensuring that the overall quality indicators consistently exceed preset thresholds while reducing zinc and energy consumption. Based on data-driven heat conduction correction parameters and model optimization results, a closed-loop control of "perception-analysis-decision-execution" is formed, which dynamically adapts to disturbances such as fluctuations in zinc liquid composition and equipment aging, reduces the frequency of manual intervention, and improves the stability and process robustness of continuous production line operation.

[0047] Furthermore, based on the heat transfer efficiency correction parameters of each heating zone in the zinc bath, the temperature gradient, heat transfer response time, and dynamic thermal balance index are calculated using the first model to generate a set of thermal field stability parameters. The galvanizing time and temperature distribution are then optimized using the heat transfer dynamic response function. Simultaneously, for the target heating zone associated with the welding zone, the second model analyzes temperature fluctuations, electrode pressure deviations, and resistance spot welding fluctuation coefficients to generate a set of thermal effect parameters. This is combined with a welding dynamic compensation function to optimize electrode pressure and welding pulse timing, achieving joint control of coating quality and welding quality. Through the synergistic optimization of thermal field stability and welding thermal effects using dual models, galvanizing process parameters and welding parameters are dynamically corrected, improving the uniformity of zinc bath heat transfer and the accuracy of thermal balance in the welding zone. This ensures consistent galvanized layer thickness and reduces welding defects, ensuring that the quality estimate consistently exceeds the preset threshold.

[0048] These or other aspects of this application will become more apparent from the description of the following embodiments. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1A flowchart illustrating an intelligent optimization method for galvanized steel sheet process parameters provided in this application embodiment;

[0051] Figure 2 A schematic diagram of the structure of an intelligent optimization system for galvanized steel sheet process parameters provided in this application embodiment;

[0052] Figure 3 This is a schematic diagram of the structure of a computing device provided in an embodiment of this application. Detailed Implementation

[0053] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0054] In some of the processes described in the specification, claims, and accompanying drawings of this application, multiple operations appearing in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order they appear herein, or may be executed in parallel. The operation numbers, such as 11, 12, etc., are merely used to distinguish different operations and do not themselves represent any execution order. Furthermore, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions such as "first," "second," etc., in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to different types.

[0055] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0056] To address the issues of poor quality stability in galvanized steel sheets caused by low computational efficiency, lack of multi-physics coupling, and insufficient coordination of multiple process parameters in existing technologies, this application provides an intelligent optimization method for galvanized steel sheet process parameters. The overall concept of this method is as follows: By real-time acquisition of temperature and coating thickness distribution data in multiple heating zones of the zinc bath during the galvanizing stage, a mapping relationship between thermodynamic parameters and coating quality results is established; based on a data-driven method, the influence of temperature distribution on coating thickness uniformity is analyzed, and dynamic correction parameters for heat conduction efficiency in each heating zone are generated to compensate for process fluctuations; subsequently, the corrected thermal parameters and galvanizing time are input into a first process optimization model to iteratively solve for the optimal combination of process parameters; simultaneously, a second process optimization model is used to collaboratively optimize parameters such as electrode pressure and current waveform during the welding stage, ultimately achieving dual assurance of coating quality and welding strength, ensuring that quality assessment indicators exceed preset thresholds.

[0057] Figure 1 A flowchart of a method for intelligent optimization of process parameters for galvanized steel sheets provided in this application embodiment is shown below. Figure 1 As shown, the method includes:

[0058] S11. In the galvanized steel sheet production process, obtain the temperature distribution data of the first galvanizing time and different heating zones in the zinc bath.

[0059] In this context, galvanized steel sheet can be simply referred to as steel sheet. The first galvanizing time is the initially set duration for which the steel sheet is immersed in the zinc bath, used to form the basic coating. The zinc bath temperature distribution data is the real-time temperature value of different heating zones within the zinc bath collected by sensors, used to characterize the uniformity of the thermal field.

[0060] In this embodiment, the timing module of the production control system automatically records the time parameter, which starts timing from the trigger signal when the steel plate is immersed in the zinc bath and ends when the exit signal when the steel plate leaves the zinc bath. The value is directly derived from the process timing database of the programmable logic controller or the distributed control system. The temperature distribution data of different heating zones in the zinc bath are collected in real time by thermocouple arrays or infrared thermal imagers embedded in each partition of the tank.

[0061] S12. Thickness detection is performed on different areas of the coating already formed on the surface of the galvanized steel sheet to obtain the final coating thickness distribution data.

[0062] The final coating thickness distribution data refers to the coating thickness values ​​of each region on the surface of the galvanized steel sheet. In this embodiment, the thickness detection process is as described in steps 121 to 126 below, and will not be repeated here.

[0063] S13. Based on the temperature distribution data and the final coating thickness distribution data, generate the heat transfer efficiency correction parameters for each heating zone of the zinc bath.

[0064] Among them, the heat conduction efficiency correction parameter is a parameter calculated based on temperature distribution data and final coating thickness distribution data. This parameter can be a temperature gradient compensation coefficient, thermal field balance factor, etc., and is used to adjust the heat output efficiency of the heating zone.

[0065] In this embodiment of the application, based on the temperature distribution data and the final coating thickness distribution data, the heat conduction efficiency correction parameters for each heating zone of the zinc bath are generated based on the mean axial temperature gradient and the standard deviation of the regional thickness. The specific description is as described in steps 131 to 133, and will not be repeated here.

[0066] S14. Based on the heat conduction efficiency correction parameters of all heating zones in the zinc bath, the galvanizing time and temperature distribution data are optimized using a pre-trained intelligent optimization model of the first process parameter, and the electrode pressure parameters and resistance spot welding parameters of the welding stage are optimized using a pre-trained intelligent optimization model of the second process parameter, so that the quality estimate of the galvanized steel sheet is greater than the preset quality threshold.

[0067] The first intelligent optimization model for process parameters can be a zinc melt flow-heat transfer coupling model, specifically a machine learning model based on the dynamic response function of heat conduction, outputting galvanizing time and temperature compensation values. The second intelligent optimization model for process parameters can be an electrode-coating contact mechanics model, specifically a deep learning model based on the welding heat-affected zone effect, optimizing electrode pressure parameters and spot welding pulse parameters. Specifically, the first intelligent optimization model is a time-series prediction model based on neural network training, with the input layer being temperature gradient parameters, heat conduction response time, and dynamic thermal balance index, and the output layer being the galvanizing time correction and temperature distribution compensation vector. The second intelligent optimization model is a decision-making model based on reinforcement learning, employing the Q-learning algorithm, with the reward function being the difference between the quality estimate and a preset threshold.

[0068] In the embodiments of this application, the process of optimizing the galvanizing process by the intelligent optimization model of the first process parameter is as described in steps a1 to a3, and the process of optimizing the welding parameters by the intelligent optimization model of the second process parameter is as described in steps b1 to b4, which will not be repeated here.

[0069] Here is a specific example: A car steel sheet factory producing galvanized door panels needs to address issues such as excessively thick edge coatings and incomplete welds. Initially, the galvanizing time is set to 22 seconds. Six temperature sensors deployed in the zinc bath collect real-time temperature distribution data for each heating zone. Testing reveals that the temperature of the target heating zone (e.g., heating zone D) at 635℃ is lower than the preset value of 650℃. X-ray thickness measurement shows an edge coating thickness of 25μm (target 20μm) and a center thickness of 18μm (target 20μm), indicating insufficient heat conduction in zone D, leading to poor zinc flow. Based on the temperature distribution data and the final coating thickness distribution data, the heat conduction efficiency of zone D needs to be improved by 12%. A heat conduction efficiency correction parameter is generated to achieve a power increase coefficient of 1.12. The first process parameter intelligent optimization model optimizes the galvanizing process. Specifically, after inputting the heat conduction efficiency correction parameter for heating zone D, the first process parameter intelligent optimization model calculates a galvanizing time correction of -3 seconds, shortening the galvanizing time from 22 seconds to 19 seconds, and dynamically adjusting the power distribution ratio for heating zone D. After execution, the coating thickness in the edge area was reduced, and the standard deviation was lowered. The second process parameter intelligent optimization model optimized the welding parameters. Specifically, the target heating zone corresponding to the welding area was determined as heating zone D. Based on the corrected temperature distribution data, the electrode pressure was adjusted from 400N to 430N, and the welding pulse was extended from 12ms to 15ms, which improved the welding strength and reduced the rate of false welds.

[0070] By executing S11 to S14, this embodiment of the application dynamically generates heat conduction correction parameters by collecting temperature and coating data in real time, and uses a dual intelligent model to collaboratively optimize the galvanizing process (time, temperature, power) and welding parameters, thereby achieving a simultaneous improvement in coating thickness uniformity and welding strength, ensuring that the quality of galvanized steel sheet consistently exceeds the preset threshold, and reducing rework rate and production costs.

[0071] In one possible embodiment, S14, based on the heat transfer efficiency correction parameters of all heating zones in the zinc bath, the galvanizing time and temperature distribution data are optimized using a pre-trained intelligent optimization model of the first process parameters, including:

[0072] Step a1: Based on the heat conduction efficiency correction parameters of all heating zones in the zinc bath, calculate the temperature gradient parameters, heat conduction response time, and dynamic thermal balance index respectively to generate the first parameter set used to characterize the thermal field stability of the zinc bath.

[0073] The temperature gradient parameter can be the average temperature difference between adjacent heating zones, used to characterize the uniformity of lateral heat conduction in the zinc bath. The heat conduction response time is the average time it takes for the temperature in the heating zone to recover from a fluctuating state to a set value, reflecting the dynamic adjustment capability of the heating system. The dynamic thermal balance index is a normalized index calculated based on the temperature fluctuation amplitude and recovery speed, characterizing the thermal field's resistance to disturbances; a higher value indicates a stronger resistance to interference.

[0074] In this embodiment, based on the heat conduction efficiency correction parameters of each heating zone in the zinc bath, the temperature gradient parameters are calculated using finite element thermal field simulation or three-dimensional spatial interpolation algorithms to determine the rate of temperature change between adjacent heating zones. The formula is as follows: Where ΔT is the temperature difference between adjacent heating zones, and Δx is the distance between the heating zones. The parameters for adjusting heat conduction efficiency were adjusted; the heat conduction response time was determined by transfer function analysis or fitting transient thermal response experimental data, and the unsteady-state heat conduction equation was solved. Extract the characteristic time constant of the system's step response, where ρ is the density of the zinc liquid, and c p is the specific heat capacity, k is the thermal conductivity, and Q is the heat source term; the dynamic thermal balance index is based on the law of conservation of energy to construct a state-space model, and uses Kalman filter or recursive least squares method to calculate the dynamic ratio of heating power to heat dissipation loss in real time. Finally, the first parameter set of the index is generated through multi-parameter fusion technology to quantify the thermal field stability.

[0075] Step a2: Input the first parameter set into the intelligent optimization model of the first process parameter, and generate the galvanizing time correction amount and temperature distribution compensation vector through the predefined heat conduction dynamic response function in the intelligent optimization model of the first process parameter.

[0076] The heat conduction dynamic response function is based on a time-series prediction model trained by a neural network. Its input is the first parameter set, and its output is the galvanizing time correction and temperature compensation vector. The temperature distribution compensation vector includes the adjustment amount of the temperature setpoint of each heating zone and the power distribution weight.

[0077] Furthermore, the formula for the dynamic response function of heat conduction is:

[0078] f LSTM (v t δ T )=[ΔT,Δτ]=Decoder(LSTM(x t ), where v t δ is the eigenvector of the dynamic thermal field. T ΔT is the galvanizing time fluctuation coefficient, ΔT is the galvanizing time correction amount, Δτ is the temperature distribution compensation vector, Decoder is the fully connected layer, and LSTM(x) is the galvanizing time correction factor. t () is in a hidden state.

[0079] The implementation process of step a2 is as described in steps a21 to a23 below, and will not be repeated here.

[0080] Step a3: Optimize the galvanizing time based on the galvanizing time correction amount to obtain the second galvanizing time. Dynamically adjust the power distribution ratio of each heating zone in the zinc bath according to the temperature distribution compensation vector to optimize the temperature distribution data.

[0081] The second galvanizing time is the actual process execution time of 16.8 seconds after adjustment, compared to 18 seconds before adjustment. The power distribution ratio is the percentage of the total system power allocated to each heating zone of the zinc bath, dynamically distributed according to the compensation vector.

[0082] In this embodiment, for optimizing the galvanizing time, a reinforcement learning algorithm or model predictive control based on time constraints is employed. By establishing a coating thickness-time response model, using the heat transfer efficiency correction parameter as input, and with coating uniformity as the optimization objective, the shortest second galvanizing time satisfying the quality threshold is iteratively solved. Its mathematical expression is as follows: Where Δt is the time correction amount, f is the optimization function, and t new For the optimized second galvanizing time, t initial For the initial first galvanizing time, For multi-parameter coupling function, η is the dynamic thermal balance index; for dynamic adjustment of temperature distribution, the distributed collaborative control algorithm driven by temperature distribution compensation vector is used, combined with adaptive sliding mode control technology, to calculate the power distribution weight of each heating zone in real time, and send it to the heater actuator through the industrial Internet of Things platform to realize closed-loop balanced control of the temperature field of the zinc bath.

[0083] Continuing the example above, in the production scenario of galvanized car door panels in an automotive steel sheet factory, based on the correction parameter that the heat conduction efficiency of heating zone D needs to be improved by 12%, and combined with the heat conduction data of the other five heating zones, the temperature gradient parameters are calculated through finite element thermal field simulation. The response time of heating zone D is fitted using the unsteady-state heat conduction equation, and the dynamic thermal balance index is calculated in real time through Kalman filtering to generate the first parameter set characterizing the stability of the thermal field. The parameters are input into the first process optimization model. The model's built-in dynamic response function of heat conduction, combined with the coating thickness deviation, outputs a galvanizing time correction of -3 seconds and a temperature distribution compensation vector. The galvanizing time is shortened from 22 seconds to 19 seconds. At the same time, the power of each heating zone is dynamically adjusted through a distributed collaborative control algorithm. After optimization, the standard deviation of the temperature distribution is reduced from 9.2℃ to 4.5℃, and the edge coating thickness is reduced to 21μm. Simultaneously, the second model is triggered to adjust the welding parameters, ultimately achieving a synergistic improvement in coating uniformity and welding strength.

[0084] By executing steps a1 to a3, the embodiments of this application achieve synergistic optimization of the thermal field stability of the zinc bath and the coating quality: the accurate calculation of temperature gradient parameters and dynamic thermal balance index supports the scientific decision-making of the zinc plating time correction amount, the dynamic adjustment of power distribution ratio reduces the excessive thickness of the edge coating, the shortening of zinc plating time improves production efficiency, and ultimately ensures that the coating uniformity, adhesion and weld quality of the galvanized steel sheet reach the preset threshold.

[0085] In one possible embodiment, S14, optimizing the electrode pressure parameters and resistance spot welding parameters during the welding stage using a pre-trained intelligent optimization model for second process parameters, so that the estimated quality value of the galvanized steel sheet is greater than a preset quality threshold, includes:

[0086] Step b1: Determine the welding area of ​​the galvanized steel sheet. Select the heat transfer efficiency correction parameter of the target heating area from the heat transfer efficiency correction parameters of all heating areas in the zinc bath. The target heating area is the area in the zinc bath corresponding to the welding area in the galvanized steel sheet production process.

[0087] The welding zone is the designated area on the galvanized steel sheet where resistance spot welding is required, such as the weld points on the edge of a car door panel. The target heating zone is the heating zone number corresponding to the welding zone when it is immersed in the zinc bath during the galvanizing stage.

[0088] In this embodiment, the welding area is determined by the physical location mapping relationship of the galvanized steel plate. Specifically, the coordinates of the welding area on the steel plate are matched with the spatial distribution of the heating area of ​​the zinc bath to select the target heating area. Then, the parameters of the target heating area are extracted from the heat conduction efficiency correction parameters of all heating areas of the zinc bath. This process relies on the area positioning module and data filtering algorithm of the programmable logic controller system or the distributed control system.

[0089] Step b2: Based on the heat conduction efficiency correction parameters of the corresponding heating zone, calculate the temperature fluctuation amplitude, electrode pressure deviation, and resistance spot welding fluctuation coefficient to generate a second parameter set for characterizing the thermal effect of the welding zone.

[0090] Among them, the temperature fluctuation amplitude is the standard deviation of the temperature in the target heating zone during the galvanizing stage, reflecting the stability of heat conduction. The electrode pressure deviation is the percentage deviation between the actual electrode pressure and the theoretical value during welding. The resistance spot welding fluctuation coefficient is the fluctuation rate of the welding current / voltage caused by uneven heat conduction. The second parameter set includes the temperature fluctuation amplitude of the welding zone, the electrode pressure deviation, and the resistance spot welding fluctuation coefficient.

[0091] In this embodiment, based on the heat conduction efficiency correction parameter of the target heating zone, the temperature fluctuation amplitude is calculated by the sliding window algorithm, the electrode pressure deviation is calculated by combining historical electrode pressure data, and the resistance spot welding fluctuation coefficient is calculated by using real-time resistance spot welding data. Finally, a second parameter set is generated by multi-source data fusion. The temperature fluctuation amplitude and resistance spot welding fluctuation coefficient are calculated by time series analysis, and the electrode pressure deviation is calculated by a multiple regression model.

[0092] Step b3: Input the second parameter set into the intelligent optimization model of the second process parameters, and generate the electrode pressure correction amount and resistance spot welding parameter compensation vector through the predefined welding dynamic compensation function in the intelligent optimization model of the second process parameters.

[0093] The welding dynamic compensation function is a decision model trained using reinforcement learning. Its input is the second parameter set, and its output is the electrode pressure correction and the spot welding pulse timing compensation vector. Optionally, the formula for the welding dynamic compensation function is:

[0094] s=[σ T ΔF dev η weld ] T , where σ T Let ΔF be the amplitude of temperature fluctuation. dev η is the electrode pressure deviation. weld This refers to the resistance spot welding fluctuation coefficient. Furthermore, the specific implementation of step b3 can be found in steps b31 to b33, which will not be repeated here.

[0095] Step b4: Optimize the electrode pressure parameters during the welding stage based on the electrode pressure correction amount, and dynamically adjust the welding pulse duration and interval based on the resistance spot welding parameter compensation vector to optimize the resistance spot welding parameters so that the estimated quality value of the galvanized steel sheet is greater than the preset quality threshold.

[0096] Among them, the resistance spot welding parameter compensation vector is the adjustment amount of the welding pulse duration and interval time.

[0097] In this embodiment, the electrode pressure parameters during the welding stage are adjusted according to the electrode pressure correction amount, and the duration and interval of the welding pulse are dynamically adjusted through the timing control module according to the resistance spot welding parameter compensation vector. This process ensures that the welding energy input matches the coating characteristics after the heat conduction efficiency correction through real-time feedback control, and ultimately makes the quality estimate of the galvanized steel sheet exceed the preset threshold.

[0098] Continuing the example above, in the production scenario of galvanized car door panels in an automotive steel sheet factory, based on the mapping relationship of the welding area position, the target heating area is determined to be heating area D of the zinc bath. Based on the heat conduction efficiency correction parameter of area D, the temperature fluctuation amplitude is calculated to be (655℃-640℃)÷647℃≈2.3%, the electrode pressure deviation is ÷430N≈-7%, and the resistance spot welding fluctuation coefficient is welding current standard deviation 15A÷mean 120A=12.5%, generating a second parameter set. The second parameter set is input into the intelligent optimization model of the second process parameters, and the electrode pressure correction amount +30N and the resistance spot welding parameter compensation vector are output through the welding dynamic compensation function. After optimization, the electrode pressure is adjusted from 400N to 430N, the welding pulse is extended from 12ms to 15ms, the welding strength is increased from 350MPa to 390MPa, the false weld rate is reduced from 5% to 1.5%, and the coating uniformity is improved, further reducing the volatilization of zinc layer in the welding heat-affected zone, achieving full compliance with the quality standards of the galvanized car door panels.

[0099] By executing steps b1 to b4, the embodiments of this application achieve synergistic optimization of welding parameters and galvanizing process: the heat conduction efficiency correction parameter of the target heating zone drives dynamic compensation of welding, the precise adjustment of electrode pressure and welding pulse can reduce the rate of false welding, and at the same time improve the uniformity of coating, so that the comprehensive quality index of galvanized steel sheet stably exceeds the preset threshold.

[0100] In one possible embodiment, step a2, inputting the first parameter set into the first process parameter intelligent optimization model, and generating the galvanizing time correction and temperature distribution compensation vector through the predefined heat conduction dynamic response function in the first process parameter intelligent optimization model, includes:

[0101] Step a21: Input the first parameter set into the intelligent optimization model of the first process parameter. The intelligent optimization model of the first process parameter takes the ratio between the axial component of the temperature gradient parameter in the first parameter set and the maximum allowable axial temperature gradient of the zinc bath as the target ratio. Combine the target ratio with the heat conduction response time to obtain the axial gradient weight factor. Based on the dynamic thermal balance index and temperature fluctuation coefficient in the first parameter set, the radial offset attenuation coefficient is calculated using the first exponential attenuation model.

[0102] The axial component of the temperature gradient parameter is the temperature gradient value of adjacent heating zones along the length (axial direction) of the zinc bath. The maximum allowable axial temperature gradient of the zinc bath is the axial gradient threshold defined by the process specification. The target ratio is the ratio of the axial component to the maximum allowable gradient. The axial gradient weighting factor is a normalized parameter combining the target ratio and the heat conduction response time, used to quantify the axial heat conduction efficiency. The dynamic thermal balance index is an indicator reflecting the thermal field's resistance to disturbances, ranging from 0 to 1; a higher value indicates stronger stability. The temperature fluctuation coefficient is the standard deviation of the temperature data, characterizing the amplitude of thermal field fluctuations. The radial offset attenuation coefficient is a parameter calculated using an exponential attenuation model, ranging from 0 to 1, used to suppress radial (width direction) heat conduction deviations.

[0103] Optionally, the first exponential decay model is α=I×e-σT / τ, where I is the dynamic thermal balance exponent, σT is the temperature fluctuation coefficient, and τ is the preset decay time constant.

[0104] In this embodiment, the first parameter set is input into the intelligent optimization model of the first process parameters. First, the axial component of the temperature gradient parameter in the first parameter set is extracted, and the ratio of it to the maximum allowable axial temperature gradient of the zinc bath is taken as the target ratio. Then, the target ratio is multiplied by the reciprocal of the heat conduction response time to obtain the axial gradient weight factor. At the same time, based on the dynamic thermal balance index and temperature fluctuation coefficient in the first parameter set, the radial offset attenuation coefficient is calculated using the first exponential decay model, where the attenuation constant is calibrated by historical data.

[0105] Step a22: Construct a dynamic thermal field feature vector consisting of an axial gradient weighting factor and a radial offset attenuation coefficient.

[0106] Among them, the dynamic thermal field feature vector is a two-dimensional vector composed of the axial gradient weighting factor and the radial offset attenuation coefficient, which is used to characterize the dynamic properties of the thermal field.

[0107] In this embodiment, the output axial gradient weight factor and radial offset attenuation coefficient are fused in multiple dimensions to construct a dynamic thermal field feature vector. This vector is compressed in dimension through principal component analysis or feature dimensionality reduction algorithm to characterize the coupling relationship between the axial and radial thermal field stability of the zinc bath.

[0108] Step a23: Based on the dynamic thermal field characteristic vector and the galvanizing time fluctuation coefficient, calculate the galvanizing time correction and temperature distribution compensation vector through the heat conduction dynamic response function.

[0109] Among them, the galvanizing time fluctuation coefficient is the deviation rate between the historical galvanizing time and the actual coating thickness.

[0110] In this embodiment, based on the dynamic thermal field feature vector and the galvanizing time fluctuation coefficient, the galvanizing time correction and temperature distribution compensation vector are calculated through the heat conduction dynamic response function. The temperature distribution compensation vector optimizes the power distribution ratio of each heating zone through the gradient backpropagation algorithm.

[0111] Continuing the example above, when a car factory produces galvanized door panels, based on the axial temperature gradient component of 10℃ / m and the maximum allowable gradient of 8℃ / m in the first parameter set, a target ratio of 1.25 is calculated. Combined with the heat conduction response time of 14 seconds, an axial gradient weighting factor of 0.089 is generated. At the same time, using the dynamic thermal balance index of 0.93 and the temperature fluctuation coefficient of 2.3%, a radial offset attenuation coefficient of 0.92 is calculated through an exponential decay model. Step a22 constructs a dynamic thermal field feature vector [0.089, 0.92]. Step a23, based on this vector and the galvanizing time fluctuation coefficient of 1.2 seconds, outputs a galvanizing time correction of -3 seconds and a temperature distribution compensation vector through the heat conduction dynamic response function. After execution, the temperature in area D rises from 635℃ to 650℃, the edge coating thickness decreases to 21μm, and the welding failure rate decreases to 1.5%.

[0112] By executing steps a21 to a23, the embodiments of this application achieve precise matching between multi-dimensional features of the thermal field and process parameters: the axial gradient weighting factor suppresses coating thickness deviation in the length direction of the zinc bath, the radial offset attenuation coefficient improves the uniformity of zinc flow in the edge area, the galvanizing time correction amount driven by the dynamic thermal field feature vector effectively reduces the dispersion of coating thickness distribution, the temperature distribution compensation vector optimizes the power distribution ratio of the heating zone, synergistically improves coating uniformity and reduces the instability of the welding heat-affected zone, and finally optimizes the coating quality and welding strength of the galvanized steel sheet simultaneously, so that the comprehensive quality index stably reaches the preset process standard.

[0113] In one possible embodiment, step b3, inputting the second parameter set into the second process parameter intelligent optimization model, and generating electrode pressure correction and resistance spot welding parameter compensation vectors through the predefined welding dynamic compensation function in the second process parameter intelligent optimization model, includes:

[0114] Step b31: Input the second parameter set into the intelligent optimization model of the second process parameters. The intelligent optimization model of the second process parameters obtains the dynamic weight factor of the axial gradient by using the weight factor calculation formula based on the temperature fluctuation amplitude and the axial component of the temperature gradient parameter in the second parameter set. Based on the electrode pressure deviation and resistance spot welding fluctuation coefficient in the second parameter set, the impedance change attenuation coefficient is calculated by using the second exponential attenuation model.

[0115] Among them, the axial gradient dynamic weighting factor is a parameter that combines temperature fluctuations and axial gradients, reflecting the intensity of the dynamic influence of heat conduction on the weld zone. The electrode pressure deviation is the percentage deviation between the actual electrode pressure and the theoretical value. The resistance spot welding fluctuation coefficient is the welding current fluctuation rate. The impedance change attenuation coefficient is a parameter calculated based on electrode pressure deviation and current fluctuations, ranging from 0 to 1, used to suppress the influence of uneven resistance in the galvanized layer.

[0116] Step b32: Construct a welding heat effect feature vector consisting of an axial gradient dynamic weighting factor and an impedance change attenuation coefficient.

[0117] Among them, the welding heat-affected zone characteristic vector is a two-dimensional vector used to quantify the combined effects of heat conduction and resistance changes in the weld zone.

[0118] Step b33: Based on the characteristic vector of welding heat influence and the fluctuation amplitude of resistance spot welding, calculate the electrode pressure correction and the resistance spot welding parameter compensation vector through the welding dynamic compensation function.

[0119] The welding dynamic compensation function is based on a reinforcement learning-based decision model. It takes a feature vector and the resistance spot welding fluctuation amplitude as input, and outputs an electrode pressure correction and a spot welding parameter compensation vector. The resistance spot welding fluctuation amplitude is used to quantify the stability of the current during welding, reflecting the impact of changes in the galvanized layer resistance or the accuracy of equipment control on welding quality.

[0120] Continuing the above example, in the production scenario of galvanized car door panels in an automotive steel sheet factory, the second parameter set (input to the intelligent optimization model of the second process parameters, through the weight factor calculation formula, combined with the second exponential decay model, calculates the axial gradient dynamic weight factor of 2.3% and the impedance change decay coefficient of 0.28; step b32 constructs the two into a welding heat-affected feature vector [2.3%, 0.28], which characterizes the dynamic influence of abnormal heat conduction on welding impedance; based on this feature vector and the resistance spot welding fluctuation amplitude, the electrode pressure correction amount +30N and the resistance spot welding parameter compensation vector are output through the welding dynamic compensation function. After execution, the electrode pressure increases from 400N to 430N, the welding pulse is adjusted from 12ms to 15ms, the uniformity of heat-affected zone in the welding area is improved, the false weld rate is significantly reduced, and the coating edge thickness is optimized to the target range.

[0121] By executing steps b31 to b33, this embodiment of the application effectively solves problems such as incomplete welds and insufficient weld strength caused by uneven heat conduction and resistance fluctuations through dynamic optimization control. By quantifying the dynamic weighting factor of the influence of heat conduction and the attenuation coefficient for suppressing uneven resistance, the model accurately identifies the synergistic relationship between the thermal field distribution and resistance characteristics of the welding zone, and optimizes the electrode pressure parameters and spot welding process parameters in a targeted manner. After optimization, the incomplete weld rate is significantly reduced to almost eliminating weld point failures caused by thermal deformation of the galvanized layer; the weld strength is comprehensively improved and consistently exceeds industry standard requirements; the weld nugget diameter is significantly increased and fully meets the process specification threshold. At the same time, the welding current fluctuation rate is greatly reduced, and the impact of resistance changes on weld quality is effectively suppressed; the temperature fluctuation range of the zinc bath associated area is significantly narrowed, and the thermal field stability is significantly enhanced. Welding cycle efficiency is improved, and overall production capacity is optimized; by adjusting the pulse timing strategy, the unit welding energy consumption is significantly reduced; rework costs are significantly reduced due to the improvement in the incomplete weld rate, and production costs are effectively controlled.

[0122] In one possible embodiment, S12, the thickness of the coating already formed on the surface of the galvanized steel sheet is measured in different areas to obtain the final coating thickness distribution data, including:

[0123] Step 121: Divide the surface of the galvanized steel sheet into several inspection areas according to the preset process grid. The surface of the galvanized steel sheet is the coating.

[0124] The preset process grid divides the surface of the galvanized steel sheet into regular rectangular grid units at fixed intervals (e.g., 10cm × 10cm). The detection area is the coating detection area corresponding to each grid unit, used for independently acquiring spectral data.

[0125] Step 122: Trigger the spectral acquisition module of each detection area through the multispectral sensor array to obtain the raw data of the spectral reflectance intensity of each detection area.

[0126] The multispectral sensor array is a matrix composed of multiple spectral sensors, covering the entire detection area. The raw data of spectral reflectance intensity is the reflected light intensity value at different wavelengths (such as visible light and near-infrared).

[0127] For example, the multispectral sensor array consists of 10 near-infrared sensors in different wavelength bands, and the spectral reflectance intensity of each detection region is calibrated by reflectance in the wavelength range of 550nm-850nm.

[0128] Step 123: Based on the original data of spectral reflectance intensity of each detection area, dynamically calibrate the coating thickness of each detection area to generate initial coating thickness distribution data.

[0129] Among them, dynamic calibration is based on a pre-trained relationship model between spectral reflectance intensity and coating thickness, which calculates the thickness of each region in real time.

[0130] Step 124: By using the gradient of reflection intensity difference between adjacent detection areas, interpolate and correct the thickness jump anomalies in the initial coating thickness distribution data to generate corrected coating thickness distribution data.

[0131] The reflection intensity difference gradient is the absolute value of the rate of change of reflection intensity between adjacent regions (e.g., the difference in reflection intensity between regions A1 and A2 / spacing). Thickness jump anomalies are abrupt changes where the thickness difference between adjacent regions exceeds a threshold (e.g., ±5μm).

[0132] Step 125: Perform regional consistency verification on the corrected coating thickness distribution data to obtain the verified coating thickness data.

[0133] Among them, regional consistency verification verifies the continuity of thickness distribution through cluster analysis and removes outlier data. The verified coating thickness data refers to the highly reliable coating thickness distribution data obtained after multiple steps of detection, correction, and verification during the galvanized steel sheet production process.

[0134] Step 126: Spatially overlay the calibrated coating thickness data with the original spectral reflectance intensity data of all detection areas to generate the final coating thickness distribution data.

[0135] Spatial overlay involves mapping the verified thickness data to the original spectral reflectance intensity using coordinates to generate a distribution map with spectral-thickness correlation.

[0136] Continuing the example above, in the production scenario of galvanized car door panels in an automotive steel sheet factory, the surface of the galvanized steel sheet is divided into rectangular process grid units according to a fixed spacing rule, with each grid serving as an independent detection area. A multispectral sensor array deployed along the steel sheet's moving track sequentially triggers the spectral acquisition module of each grid unit to acquire raw data on the reflection intensity of each area at a specific wavelength. Based on the pre-calibrated relationship curve between reflection intensity and coating thickness, the thickness of each grid unit is dynamically calibrated to generate initial coating thickness distribution data. The gradient difference in reflection intensity between adjacent grids is detected, and bilinear interpolation is used to correct abnormal points where thickness jumps exceed the process threshold, generating smooth, corrected thickness distribution data. A regional consistency verification algorithm is used to remove local outliers to ensure data continuity. The verified thickness data and spectral reflection intensity are spatially superimposed and fused to generate a final coating thickness distribution map with fused thermodynamic properties, which is used to guide galvanizing process optimization and welding parameter adjustment.

[0137] By executing steps 121 to 126, this embodiment of the application achieves high-precision full-domain quantitative analysis of coating thickness through systematic detection and data processing: process grid division ensures detection coverage, dynamic calibration of spectral reflectance intensity improves thickness calculation efficiency, interpolation correction and consistency verification eliminate local data anomalies, spatial superposition and fusion enhance the multidimensional correlation of data, and the finally generated coating thickness distribution data provides a reliable basis for process optimization, significantly improves coating uniformity and reduces the defect rate of subsequent welding processes.

[0138] In one possible embodiment, S13, based on temperature distribution data and final coating thickness distribution data, generates heat transfer efficiency correction parameters for each heating zone of the zinc bath, including:

[0139] Step 131: Divide the temperature distribution data into a temperature gradient feature set according to the axial and radial distribution of the zinc bath, and divide the final coating thickness distribution data into a thickness difference feature set according to the preset process grid.

[0140] The temperature gradient feature set consists of temperature distribution data along the axial (length) and radial (width) directions within the zinc bath, including the temperature values ​​of each region and the temperature difference between adjacent regions. The thickness difference feature set consists of coating thickness data divided according to a preset process grid, including the thickness value of each grid region and the thickness difference between adjacent regions.

[0141] Step 132: Based on the mean axial temperature gradient corresponding to the temperature gradient feature set and the standard deviation of regional thickness corresponding to the thickness difference feature set, construct a dynamic heat conduction feature vector composed of axial gradient weighting factor and thickness standard deviation compensation coefficient.

[0142] The thickness standard deviation compensation coefficient is a compensation parameter calculated by normalizing the percentage difference between the thickness standard deviation and the target value, ranging from 0 to 1. The axial gradient weighting factor is used to quantify the deviation between the actual axial gradient and the maximum allowable gradient, typically ranging from 0 to 1. The closer the value is to 1, the closer the heat conduction efficiency is to the critical state, and the more priority should be given to adjusting it.

[0143] Step 133: Input the dynamic heat conduction feature vector into the predefined heat conduction efficiency correction function. Generate the heat conduction efficiency correction parameter by using the exponential decay term of the radial temperature offset amplitude and the piecewise linear correction term of the thermal field fluctuation coefficient in the heat conduction efficiency correction function.

[0144] The radial temperature offset amplitude is the maximum temperature difference between different radial regions at the same axial position (e.g., a left-to-right temperature difference of 10℃). The thermal field fluctuation coefficient within the tank is the standard deviation of the overall temperature of the zinc bath (e.g., ±3℃). The piecewise linear correction term for the thermal field fluctuation coefficient adjusts the power compensation ratio according to the interval of the thermal field fluctuation coefficient. The exponential decay term of the radial temperature offset amplitude is a nonlinear correction term based on the radial offset amplitude, suppressing local thermal deviations. The heat transfer efficiency correction function can achieve piecewise linear correction, for example: K = w1 × e -ΔTr +w2×max(0,σ T -T0), where K is the heat transfer efficiency correction parameter, ΔTr is the radial temperature offset amplitude, T0 is the maximum allowable temperature fluctuation reference value, and σ T Let w1 be the thermal field fluctuation coefficient within the tank, and w2 be the weighting factors for the corresponding terms. Furthermore, this application does not specifically limit the way the heat transfer efficiency correction function is expressed.

[0145] Continuing the example above, when a car factory produces galvanized door panels, it decomposes the zinc bath temperature data into an axial gradient feature set and a thickness difference feature set. The axial gradient weighting factor is calculated as 8 ÷ 10 = 0.8, and the thickness standard deviation compensation coefficient is calculated as 3.5 × 0.3 = 1.05, constructing a dynamic heat conduction feature vector [0.8, 1.05]. Using the heat conduction efficiency correction function, the exponential decay term = 5℃ × e^(-0.2 × 1.05) ≈ 4.1, the piecewise linear correction term = 0.8 × 1.2 = 0.96, generating a power enhancement coefficient of 1.12. This increases the power of the driving heating zone D from 12kW to 13.44kW, and after optimization, the edge coating thickness decreases from 25μm to 21μm, with the welding failure rate simultaneously decreasing to 1.5%.

[0146] By executing steps 131 to 133, the embodiments of this application achieve precise control of the heat transfer efficiency of the zinc bath through multi-dimensional feature fusion and dynamic correction: the axial gradient weight factor suppresses excessive coating thickness in the length direction, the thickness standard deviation compensation coefficient reduces the thickness distribution dispersion, and the power distribution optimization driven by the heat transfer efficiency correction function significantly improves the coating uniformity while reducing the heat-affected zone effect, ultimately ensuring that the coating quality and welding strength of the galvanized steel sheet meet the standards in a coordinated manner.

[0147] Figure 2 A schematic diagram of a smart optimization system for galvanized steel sheet process parameters provided in this application embodiment is shown below. Figure 2 As shown, the system includes:

[0148] The acquisition module 21 is used to acquire data on galvanizing time and temperature distribution in different areas of the zinc bath during the galvanizing steel sheet production process.

[0149] The detection module 22 is used to detect the thickness of different areas of the coating already formed on the surface of the galvanized steel sheet and obtain coating thickness distribution data.

[0150] The calculation module 23 is used to generate heat conduction efficiency correction parameters for each heating zone of the zinc bath based on temperature distribution data and coating thickness distribution data.

[0151] The optimization module 24 is used to correct parameters based on the heat conduction efficiency of all heating zones in the zinc bath, optimize the galvanizing time and temperature distribution data using a pre-trained first process parameter intelligent optimization model, and optimize the electrode pressure parameters and resistance spot welding parameters in the welding stage using a pre-trained second process parameter intelligent optimization model, so as to make the quality estimate of the galvanized steel sheet greater than the preset quality threshold.

[0152] Figure 2 The intelligent optimization system for galvanized steel sheet process parameters can perform... Figure 1 The implementation principle and technical effects of the intelligent optimization method for galvanized steel sheet process parameters described in the illustrated embodiment will not be repeated here. The specific methods of operation of each module and unit in the intelligent optimization system for galvanized steel sheet process parameters described in the above embodiments have been described in detail in the embodiments related to this method, and will not be elaborated upon here.

[0153] In one possible design, Figure 2 The intelligent optimization system for galvanized steel sheet process parameters shown in the embodiment can be implemented as a computing device, such as... Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32.

[0154] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are invoked and executed by the processing component 32.

[0155] The processing component 32 is used to: acquire the first galvanizing time and temperature distribution data of different heating zones in the zinc bath during the galvanized steel sheet production process; perform thickness detection on different areas of the coating already formed on the surface of the galvanized steel sheet to obtain the final coating thickness distribution data; generate heat transfer efficiency correction parameters for each heating zone of the zinc bath based on the temperature distribution data and the final coating thickness distribution data; and optimize the galvanizing time and temperature distribution data using a pre-trained first process parameter intelligent optimization model based on the heat transfer efficiency correction parameters of all heating zones in the zinc bath, and optimize the electrode pressure parameters and resistance spot welding parameters in the welding stage using a pre-trained second process parameter intelligent optimization model, so that the estimated quality value of the galvanized steel sheet is greater than a preset quality threshold.

[0156] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above-described method. Alternatively, the processing component may be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described method.

[0157] Storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented from any type of volatile or non-volatile storage device or a combination thereof, such as Random Access Memory (RAM), Static Random-Access Memory (SRAM), Electrically Erasable Programmable Read Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0158] Of course, computing devices may also include other components, such as input / output interfaces, display components, communication components, etc.

[0159] Input / output interfaces provide interfaces between processing components and peripheral interface modules, which can be output devices, input devices, etc.

[0160] The communication components are configured to facilitate wired or wireless communication between computing devices and other devices.

[0161] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform. In this case, the computing device can refer to a cloud server, and the aforementioned processing components, storage components, etc., can be basic server resources rented or purchased from the cloud computing platform.

[0162] This application also provides a computer storage medium storing a computer program, which, when executed by a computer, can perform the above-described functions. Figure 1 The embodiment shown illustrates an intelligent optimization method for process parameters of galvanized steel sheets.

[0163] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0164] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0165] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0166] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for intelligent optimization of process parameters for galvanized steel sheets, characterized in that, include: In the galvanized steel sheet production process, obtain the temperature distribution data of the first galvanizing time and different heating zones in the zinc bath; Thickness measurements were performed on different areas of the coating already formed on the surface of the galvanized steel sheet to obtain the final coating thickness distribution data. Based on the temperature distribution data and the final coating thickness distribution data, heat conduction efficiency correction parameters for each heating zone of the zinc bath are generated. Based on the heat conduction efficiency correction parameters of all heating zones in the zinc bath, a pre-trained intelligent optimization model for the first process parameter is used to optimize the galvanizing time and temperature distribution data, and a pre-trained intelligent optimization model for the second process parameter is used to optimize the electrode pressure parameters and resistance spot welding parameters during the welding stage, so that the estimated quality value of the galvanized steel sheet is greater than the preset quality threshold. The first intelligent optimization model for the first process parameter is a time-series prediction model based on neural network training, with the input layer being temperature gradient parameters, heat conduction response time, and dynamic thermal balance index, and the output layer being the galvanizing time correction amount and temperature distribution compensation vector. The second intelligent optimization model for the second process parameter is a decision model based on reinforcement learning, using the Q-learning algorithm, with the reward function being the difference between the estimated quality value and the preset quality threshold.

2. The method according to claim 1, characterized in that, The step of correcting the heat transfer efficiency of all heating zones in the zinc bath and optimizing the galvanizing time and temperature distribution data using a pre-trained intelligent optimization model of the first process parameters includes: Based on the heat conduction efficiency correction parameters of all heating zones in the zinc bath, the temperature gradient parameters, heat conduction response time, and dynamic thermal balance index are calculated respectively to generate a first set of parameters for characterizing the thermal field stability of the zinc bath. The first parameter set is input into the first process parameter intelligent optimization model, and the galvanizing time correction and temperature distribution compensation vector are generated through the predefined heat conduction dynamic response function in the first process parameter intelligent optimization model. The galvanizing time is optimized based on the galvanizing time correction amount to obtain a second galvanizing time. The power distribution ratio of each heating zone in the zinc bath is dynamically adjusted based on the temperature distribution compensation vector to optimize the temperature distribution data.

3. The method according to claim 1, characterized in that, The process of optimizing electrode pressure parameters and resistance spot welding parameters during the welding stage using a pre-trained intelligent optimization model for second process parameters, so as to ensure that the estimated quality value of the galvanized steel sheet is greater than a preset quality threshold, includes: Determine the welding area of ​​the galvanized steel sheet, and select the heat transfer efficiency correction parameter of the target heating area from the heat transfer efficiency correction parameters of all heating areas of the zinc bath. The target heating area is the area in the zinc bath corresponding to the welding area in the galvanized steel sheet production process. Based on the heat conduction efficiency correction parameters of the corresponding heating zone, calculate the temperature fluctuation amplitude, electrode pressure deviation and resistance spot welding fluctuation coefficient to generate a second set of parameters for characterizing the thermal effect of the welding zone. The second parameter set is input into the second process parameter intelligent optimization model, and the electrode pressure correction amount and resistance spot welding parameter compensation vector are generated by the predefined welding dynamic compensation function in the second process parameter intelligent optimization model. The electrode pressure parameters during the welding stage are optimized based on the electrode pressure correction amount, and the welding pulse duration and interval are dynamically adjusted according to the resistance spot welding parameter compensation vector to optimize the resistance spot welding parameters so that the estimated quality value of the galvanized steel sheet is greater than the preset quality threshold.

4. The method according to claim 2, characterized in that, The step of inputting the first parameter set into the first process parameter intelligent optimization model, and generating the galvanizing time correction and temperature distribution compensation vector through the predefined heat conduction dynamic response function in the first process parameter intelligent optimization model, includes: The first parameter set is input into the first process parameter intelligent optimization model. The first process parameter intelligent optimization model takes the ratio between the axial component of the temperature gradient parameter in the first parameter set and the maximum allowable axial temperature gradient of the zinc bath as the target ratio. The target ratio is combined with the heat conduction response time to obtain the axial gradient weight factor. Based on the dynamic thermal balance index and temperature fluctuation coefficient in the first parameter set, the radial offset attenuation coefficient is calculated using the first exponential attenuation model. Construct a dynamic thermal field feature vector composed of the axial gradient weighting factor and the radial offset attenuation coefficient; Based on the dynamic thermal field characteristic vector and the galvanizing time fluctuation coefficient, the galvanizing time correction and temperature distribution compensation vector are calculated using the heat conduction dynamic response function.

5. The method according to claim 3, characterized in that, The step of inputting the second parameter set into the second process parameter intelligent optimization model, and generating electrode pressure correction and resistance spot welding parameter compensation vectors through the predefined welding dynamic compensation function in the second process parameter intelligent optimization model, includes: The second parameter set is input into the second process parameter intelligent optimization model. The second process parameter intelligent optimization model obtains the axial gradient dynamic weight factor by using the weight factor calculation formula based on the temperature fluctuation amplitude and the axial component of the temperature gradient parameter in the second parameter set. Based on the electrode pressure deviation and resistance spot welding fluctuation coefficient in the second parameter set, the impedance change attenuation coefficient is calculated using the second exponential attenuation model. Construct a welding thermal effect feature vector composed of the axial gradient dynamic weighting factor and the impedance change attenuation coefficient; Based on the welding heat-affected characteristic vector and the resistance spot welding fluctuation amplitude, the electrode pressure correction and resistance spot welding parameter compensation vector are calculated using the welding dynamic compensation function.

6. The method according to claim 1, characterized in that, The process of measuring the thickness of different areas of the coating already formed on the surface of the galvanized steel sheet to obtain the final coating thickness distribution data includes: The surface of the galvanized steel sheet is divided into several inspection areas according to a preset process grid, and the galvanized steel sheet surface is the coating. The spectral acquisition module of each detection area is triggered by a multispectral sensor array to obtain the raw data of the spectral reflectance intensity of each detection area. Based on the original spectral reflectance intensity data of each detection area, the coating thickness of each detection area is dynamically calibrated to generate initial coating thickness distribution data. By using the gradient difference in reflection intensity between adjacent detection areas, interpolation correction is performed on the thickness jump anomalies in the initial coating thickness distribution data to generate corrected coating thickness distribution data. The corrected coating thickness distribution data is subjected to regional consistency verification to obtain the verified coating thickness data. The verified coating thickness data is spatially superimposed with the original spectral reflectance intensity data of all detection areas to generate the final coating thickness distribution data.

7. The method according to claim 1, characterized in that, The step of generating heat transfer efficiency correction parameters for each heating zone of the zinc bath based on the temperature distribution data and the final coating thickness distribution data includes: The temperature distribution data is divided into a temperature gradient feature set according to the axial and radial distribution of the zinc bath, and the final coating thickness distribution data is divided into a thickness difference feature set according to the preset process grid. Based on the mean axial temperature gradient corresponding to the temperature gradient feature set and the standard deviation of regional thickness corresponding to the thickness difference feature set, a dynamic heat conduction feature vector composed of axial gradient weighting factor and thickness standard deviation compensation coefficient is constructed. The dynamic heat conduction feature vector is input into a predefined heat conduction efficiency correction function. The heat conduction efficiency correction parameter is generated by the exponential decay term of the radial temperature offset amplitude and the piecewise linear correction term of the thermal field fluctuation coefficient in the heat conduction efficiency correction function.

8. An intelligent optimization system for galvanized steel sheet process parameters, characterized in that, include: The acquisition module is used to acquire the first galvanizing time and the temperature distribution data of different heating zones in the zinc bath during the galvanized steel sheet production process. The detection module is used to detect the thickness of different areas of the coating already formed on the surface of the galvanized steel sheet, and obtain the final coating thickness distribution data. The calculation module is used to generate heat conduction efficiency correction parameters for each heating zone of the zinc bath based on the temperature distribution data and the final coating thickness distribution data. The optimization module is used to correct parameters based on the heat conduction efficiency of all heating zones in the zinc bath, optimize the galvanizing time and temperature distribution data using a pre-trained first process parameter intelligent optimization model, and optimize the electrode pressure parameters and resistance spot welding parameters during the welding stage using a pre-trained second process parameter intelligent optimization model, so that the estimated quality value of the galvanized steel sheet is greater than a preset quality threshold. The first process parameter intelligent optimization model is a time-series prediction model based on neural network training, with temperature gradient parameters, heat conduction response time and dynamic thermal balance index as input layers, and galvanizing time correction amount and temperature distribution compensation vector as output layers. The second process parameter intelligent optimization model is a decision model based on reinforcement learning, using the Q-learning algorithm, with the reward function being the difference between the estimated quality value and the preset quality threshold.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement the intelligent optimization method for galvanized steel sheet process parameters as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The device stores a computer program, which, when executed by a computer, implements an intelligent optimization method for galvanized steel sheet process parameters as described in any one of claims 1 to 7.

Images