Multi-temperature-zone collaborative cooling method for deep-drawing aluminum plate

Abstract

The invention provides a multi-temperature-zone collaborative cooling method for a deep-drawing aluminum plate, and belongs to the technical field of deep-drawing aluminum plate machining.The multi-temperature-zone collaborative cooling method comprises the steps that a thermal stress prediction model is adopted for calculating the target cooling rate and the spraying pressure reference value of each temperature zone, and a pulse film breaking spraying mode or a continuous spraying mode is selected according to the surface temperature; a temperature zone coordination intelligent agent network is adopted to realize parameter collaborative optimization and advanced adjustment of adjacent temperature zones, a gradual transition zone is arranged at the junction of the temperature zones, and smooth change of cooling intensity is realized through a variable-density nozzle array. The plate width direction is divided into independent control units, spraying parameters are dynamically adjusted according to the temperature deviation, edge electromagnetic induction heating compensation is started, temperature field and stress field data are continuously collected for online correction, and parameter adjustment is triggered when the transverse temperature difference or stress exceeds the standard. The problem that the deformation defect of the aluminum plate is caused by heat stress concentration due to non-uniform temperature field in the multi-temperature-zone cooling process is solved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of deep drawing aluminum plate processing, and in particular, relates to a multi-temperature-zone coordinated cooling method for deep drawing aluminum plates. BACKGROUND

[0002] The heat treatment process of deep drawing aluminum plates needs to realize accurate regulation of temperature through a multi-temperature-zone cooling system to ensure the forming performance and uniformity of the material. The traditional multi-temperature-zone cooling method adopts a segmented spraying system, and the cooling parameters are independently set in each temperature zone to control the cooling rate by adjusting the spraying pressure and flow. However, in the prior art, due to the lack of coordinated control mechanism between the temperature zones, there is a sudden change in the cooling parameters at the junction of the temperature zones, which causes the temperature step phenomenon on the surface of the aluminum plate, and the heat dissipation conditions of the edge and the center in the width direction of the plate are significantly different, resulting in uneven transverse temperature distribution. This unevenness of the temperature field produces a significant thermal stress gradient in the cooling process, which can cause local plastic deformation or micro-cracks and other defects when the stress exceeds the yield limit of the material, seriously affecting the deep drawing forming quality. That is, there is a technical problem in the prior art that the unevenness of the temperature field in the multi-temperature-zone cooling process causes thermal stress concentration and induces deformation defects of the aluminum plate. SUMMARY

[0003] Therefore, the application provides a multi-temperature-zone coordinated cooling method for deep drawing aluminum plates, which can solve the technical problem of unevenness of the temperature field in the multi-temperature-zone cooling process causing thermal stress concentration and inducing deformation defects of the aluminum plate in the prior art.

[0004] The application is implemented as follows: The application provides a multi-temperature-zone coordinated cooling method for deep drawing aluminum plates, which includes collecting aluminum plate surface temperature distribution data to establish an initial temperature field matrix, inputting a thermal stress prediction model to calculate and output a target cooling rate sequence of each temperature zone and a spraying pressure reference value, judging the surface temperature to start a pulse membrane-breaking spraying mode or a continuous spraying mode, using a pulse high-pressure spraying system in the first temperature zone for atomized gas-liquid mixed medium cooling, transmitting temperature data between temperature zones through a temperature zone coordination intelligent agent network and adjusting the spraying parameters in advance, setting a gradual transition zone at the junction position to realize continuous transition of the spraying flow, dividing the plate width direction into independent control units and performing flow enhancement and angle adjustment on the edge independent control unit, using a pure liquid intensified cooling method in the second temperature zone and realizing organization uniformization through differential spraying intensity control in the thickness direction of the plate, continuously collecting temperature field data and stress field data to input the thermal stress prediction model for online correction and output corrected stress field prediction values and transverse temperature difference calculation values, and sending parameter correction instructions to each temperature zone control unit when the transverse temperature difference calculation values or the stress field prediction values exceed the threshold value.

[0005] The data on the surface temperature distribution of the aluminum plate was collected using an infrared thermal imaging array with a spatial resolution of 5 mm and a sampling frequency of 50 Hz. The temperature distribution image of the entire surface of the aluminum plate was constructed by linear array scanning, and the temperature measurement accuracy was ±2℃.

[0006] The initial temperature field matrix is ​​a two-dimensional temperature data array divided along the length and width of the aluminum plate, and the matrix elements record the surface temperature value at the corresponding position.

[0007] Among them, the pulsed film-breaking injection mode disrupts the stable vapor film formed by the Leiden-Frost effect through periodic pressure pulsation, and the pulsed high-pressure injection ruptures the vapor film through instantaneous pressure impact, restoring direct contact between the liquid and the solid surface.

[0008] Among them, the atomized gas-liquid mixing medium is a two-phase flow cooling medium formed by atomizing coolant into micron-sized droplets through high-pressure airflow and mixing it with gas.

[0009] Among them, surfactants are added to the cooling medium to reduce the surface tension of the coolant and promote the wetting and spreading behavior of the liquid on the high-temperature surface of the aluminum plate.

[0010] Among them, the temperature zone coordinated agent network is a multi-agent system based on a distributed cooperative control strategy. Each temperature zone control unit acts as an independent agent and exchanges state information through a communication network. Each agent maintains the performance index of its own temperature zone as a local objective function, while maintaining the coordination consistency with adjacent temperature zones as a global objective function.

[0011] The gradual transition zone is a region where cooling parameters change continuously at the boundary between adjacent temperature zones. By changing the nozzle arrangement density, the spatial gradient distribution of spray flow is achieved, eliminating the temperature step phenomenon at the boundary of the temperature zone.

[0012] The independent control unit is a temperature control zone divided according to the width of the aluminum plate, and each independent control unit is equipped with an independent spray valve group and flow regulation device.

[0013] Among them, the edge independent control unit is an independent control unit located on both sides of the aluminum plate in the width direction. When the temperature deviation value of the edge independent control unit exceeds the threshold, the nozzle flow of the edge independent control unit is increased and the spray angle is adjusted to the tilt angle.

[0014] Among them, the edge electromagnetic induction heating compensation device generates Joule heat by inducing eddy currents in the edge of the aluminum plate through a high-frequency alternating magnetic field, which compensates for the temperature loss caused by excessive heat dissipation in the edge. The heating power is dynamically adjusted according to the temperature deviation value.

[0015] Among them, the differential spray intensity control in the thickness direction of the plate is achieved by adjusting the nozzle angle and flow distribution to obtain different cooling intensities for the surface and the interior of the plate. The surface is cooled more strongly, while the interior is cooled more gently through heat conduction.

[0016] Among them, the gas fine control mode is a method of cooling the aluminum plate using a gaseous cooling medium.

[0017] The online correction involves feeding back the measured temperature and stress field data of the cooling process to the thermal stress prediction model, and updating the model's internal parameters based on the deviation between the model output and the measured values.

[0018] Among them, the thermal stress prediction model is a hybrid calculation model that integrates the boundary element method and the analytical solution method. The boundary element method only needs to discretize the surface and edge of the plate without dividing the internal region into a mesh, transforming the three-dimensional volume integral problem into a two-dimensional surface integral problem. The analytical solution layer uses Green's function integration to handle the volume force effect caused by the temperature gradient.

[0019] Among them, the time evolution layer of the thermal stress prediction model uses the fine integration algorithm to calculate the transient response. The fine integration algorithm transforms the differential equation in the time domain into the problem of accurate calculation of the exponential matrix, and achieves high-precision time stepping through the addition theorem and multiplication theorem of the matrix exponent.

[0020] This invention achieves parameter coordinating optimization between adjacent temperature zones by establishing a temperature zone coordination intelligent agent network. A gradual transition zone is set at the temperature zone boundary to eliminate abrupt changes in cooling intensity. Simultaneously, combined with differentiated spray adjustment by independent control units along the plate width and edge electromagnetic induction heating compensation, a uniform temperature field distribution across the entire surface of the aluminum plate is ensured. The temperature zone coordination intelligent agent network, through a distributed collaborative control strategy, enables the exchange of state information between temperature zone control units, achieving smooth transition of cooling parameters between temperature zones. The variable density nozzle array in the gradual transition zone creates a spatial gradient distribution of spray flow, avoiding stress concentration caused by temperature abrupt changes. The independent control units along the plate width dynamically adjust spray parameters based on real-time temperature deviations. The edge heating compensation device offsets temperature losses caused by excessively rapid heat dissipation at the edges. The thermal stress prediction model monitors the stress field state in real time through online correction and triggers parameter adjustments, ensuring that the stress level remains within a safe threshold range. In summary, this invention solves the technical problem mentioned in the background art where uneven temperature fields during multi-temperature zone cooling lead to thermal stress concentration and deformation defects in aluminum plates. Attached Figure Description

[0021] Figure 1 This is a flowchart of the method of the present invention.

[0022] Figure 2 This is a schematic diagram of the overall layout of a multi-temperature zone coordinated cooling system.

[0023] Figure 3 The output of the thermal stress prediction model shows the distribution curves of the target cooling rate in each temperature zone.

[0024] Figure 4 This is a temperature field evolution diagram for the temperature compensation process of the independent control unit at the edge.

[0025] Figure 5 A three-dimensional simulation diagram of the coupled temperature and stress field changes during the entire cooling process of an aluminum plate. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below.

[0027] like Figure 1 The diagram shows a flowchart of a multi-temperature zone synergistic cooling method for deep-drawn aluminum plates provided by the present invention. This method includes the following steps:

[0028] S01. The aluminum plate to be cooled is transported to the inlet of the multi-temperature zone collaborative cooling system. The surface temperature distribution data of the aluminum plate is collected in real time by the infrared thermal imaging array. An initial temperature field matrix is ​​established and the width and thickness parameters of the aluminum plate are recorded. The initial temperature field matrix, the width and thickness parameters of the aluminum plate are output to the thermal stress prediction model.

[0029] S02. Input the initial temperature field matrix, the width dimension of the aluminum plate and the thickness parameter into the thermal stress prediction model for calculation, output the target cooling rate sequence of each temperature zone and the spray pressure benchmark value, and determine whether the surface temperature of the aluminum plate exceeds 290°C based on the initial temperature field matrix. If it exceeds 290°C, start the pulse film breaking spray mode; if it does not exceed 290°C, start the continuous spray mode.

[0030] S03. In the first temperature zone, a pulsed high-pressure jet system is used to cool the surface of the aluminum plate with an atomized gas-liquid mixture medium. The pulse period is set to 0.2 seconds, the peak pressure is set to 2.0 MPa according to the spray pressure reference value, a surfactant with a mass fraction of 0.3% is added to the cooling medium, the cooling rate is controlled at 15℃ / second, the outlet temperature data of the first temperature zone is collected and output to the temperature zone coordination intelligent agent network.

[0031] S04. Input the outlet temperature data of the first temperature zone into the temperature zone coordination intelligent agent network for processing. When the front end of the aluminum plate reaches the junction of the first temperature zone and the second temperature zone, the temperature zone coordination intelligent agent network transmits the outlet temperature data of the first temperature zone to the second temperature zone control unit. The second temperature zone control unit adjusts the spraying parameters 1.2 seconds in advance and sets a gradual transition zone with a length of 350mm at the junction position. The spraying flow rate is continuously transitioned from the standard flow rate value of the first temperature zone to the standard flow rate value of the second temperature zone through a variable density nozzle array.

[0032] S05. Divide the width of the aluminum plate into 9 independent control units according to the width dimension of the plate. Continuously collect the temperature data of each independent control unit through the infrared thermal imaging array and calculate the temperature deviation value of each independent control unit. When the temperature deviation value of the edge independent control unit exceeds 8°C, increase the nozzle flow rate of the edge independent control unit by 40% and adjust the spray angle to a 53-degree tilt angle. At the same time, start the edge electromagnetic induction heating compensation device and set the heating power density to 15kW per square meter.

[0033] S06. In the second temperature zone, the aluminum plate is cooled by pure liquid enhanced cooling. The spray pressure is set to 1.8MPa according to the parameters output by the temperature zone coordination intelligent network, and the cooling rate is controlled at 35℃ / second. The uniformity of the structure is achieved by controlling the differential spray intensity in the thickness direction of the plate. When the aluminum plate enters the third temperature zone, the cooling medium is switched to the gas fine control mode, and the cooling rate is reduced to 8℃ / second.

[0034] S07. During the cooling process, temperature field data and stress field data of each temperature zone are continuously collected. The temperature field data and stress field data are input into the thermal stress prediction model for online correction. The thermal stress prediction model outputs the corrected stress field prediction value and the calculated value of the transverse temperature difference. When the calculated value of the transverse temperature difference exceeds 5°C or the predicted value of the stress field exceeds the safety threshold, the temperature zone coordination intelligent agent network calculates the parameter adjustment amount and sends the parameter correction command to each temperature zone control unit.

[0035] The infrared thermal imaging array has a spatial resolution of 5 mm and a sampling frequency of 50 Hz. It constructs a temperature distribution image of the entire surface of the aluminum plate through linear array scanning, and the temperature measurement accuracy is ±2℃.

[0036] The initial temperature field matrix is ​​a two-dimensional temperature data array divided along the length and width of the aluminum plate. The matrix elements record the surface temperature values ​​at corresponding positions, which are used for subsequent thermal stress calculations and cooling strategy formulation.

[0037] The pulsed film-breaking injection mode disrupts the stable vapor film formed by the Leidenfrost effect through periodic pressure pulsation. When the surface temperature of the plate is higher than the Leidenfrost effect of the coolant, a heat-insulating vapor layer will form between the liquid and the high-temperature surface, resulting in a significant decrease in the heat transfer coefficient. The pulsed high-pressure injection breaks the vapor film through instantaneous pressure impact, restoring direct contact between the liquid and the solid surface and improving heat transfer efficiency.

[0038] The continuous spraying mode is a stable continuous spraying method used when the surface temperature of the aluminum plate is lower than the Leidenfrost characteristic, and the spraying pressure remains constant without the need for pulse pressure changes.

[0039] The atomized gas-liquid mixture medium is a two-phase flow cooling medium formed by atomizing coolant into micron-sized droplets with high-pressure airflow and mixing them with gas. The atomized droplets have a larger specific surface area and milder cooling characteristics, avoiding uneven grain structure caused by excessively fast cooling rate.

[0040] The surfactant is used to reduce the surface tension of the coolant, promote the wetting and spreading behavior of the liquid on the high-temperature surface of the aluminum plate, enhance the liquid's ability to penetrate the vapor film, and increase the contact area between the cooling medium and the plate surface.

[0041] The temperature zone coordination agent network is a multi-agent system based on a distributed cooperative control strategy. Each temperature zone control unit acts as an independent agent, exchanging state information through a communication network. Each agent maintains its own temperature zone's performance index as a local objective function, while simultaneously maintaining coordination consistency with neighboring temperature zones as a global objective function. A consensus algorithm is used to achieve collaborative parameter optimization between temperature zones. The communication topology of the temperature zone coordination agent network adopts a directed graph structure. The agent control law includes local feedback terms and neighbor coordination terms, with weights determined by the graph's connectivity and algebraic connectivity. An event-triggered mechanism is introduced so that communication only occurs when the state deviation exceeds a set threshold, reducing communication load. The temperature zone coordination agent network is plug-and-play, requiring no reconfiguration of the controller structure when the number of temperature zones changes, and is robust to communication latency and data loss.

[0042] The gradual transition zone is a region where cooling parameters change continuously at the boundary between adjacent temperature zones. By changing the nozzle arrangement density, the spatial gradient distribution of the spray flow rate is achieved, eliminating the temperature step phenomenon at the temperature zone boundary and avoiding microscopic defects caused by stress concentration.

[0043] The variable density nozzle array is a spray arrangement in which the nozzle spacing gradually changes along the aluminum plate conveying direction. In the gradual transition zone, the nozzle density continuously transitions from the standard density of the upstream temperature zone to the standard density of the downstream temperature zone, thereby achieving a smooth change in cooling intensity.

[0044] The standard flow rate of the first temperature zone is the set value of the flow rate per unit time of the first temperature zone spray system under standard operating conditions, and the standard flow rate of the second temperature zone is the set value of the flow rate per unit time of the second temperature zone spray system under standard operating conditions.

[0045] The independent control unit is a temperature control zone divided along the width of the aluminum plate. Each independent control unit is equipped with an independent spray valve group and flow regulating device. The spray parameters are independently adjusted according to the real-time temperature data of the independent control unit to achieve temperature uniformity control along the width of the plate.

[0046] The temperature deviation value is the absolute value of the difference between the measured temperature of the independent control unit and the target temperature.

[0047] The independent edge control unit is an independent control unit located on both sides of the aluminum plate in the width direction. Since the heat dissipation conditions at the edge are different from those in the central area, compensation measures need to be applied.

[0048] The edge electromagnetic induction heating compensation device generates Joule heating by inducing eddy currents at the edge of the aluminum plate using a high-frequency alternating magnetic field, compensating for temperature loss caused by excessive heat dissipation at the edge. The heating area is 100mm wide, and the heating power is dynamically adjusted according to the temperature deviation value.

[0049] The pure liquid enhanced cooling method is a direct spray cooling method using an un-atomized liquid cooling medium, which has a stronger cooling intensity compared to an atomized gas-liquid mixture medium.

[0050] The differential spraying intensity control in the thickness direction of the sheet metal is designed to address the differences in cooling rates between different layers in the thickness direction of the aluminum sheet. By adjusting the nozzle angle and flow distribution, different cooling intensities are achieved between the surface and interior of the sheet metal. The surface layer is cooled more strongly, while the interior is cooled more gently through heat conduction. This reduces the grain size difference in the thickness direction and ensures the performance of deep drawing.

[0051] The gas fine control mode is a method of cooling the aluminum plate using a gaseous cooling medium. The cooling intensity is lower than that of liquid cooling, and it is used for fine temperature control in the later stages of cooling.

[0052] The temperature field data is the surface temperature distribution data of the aluminum plate continuously collected by the infrared thermal imaging array during the cooling process.

[0053] The stress field data is the stress distribution data of the aluminum plate collected by a stress sensor array during the cooling process.

[0054] The online correction involves feeding back the measured temperature field data and stress field data from the cooling process to the thermal stress prediction model. The model's internal parameters are updated based on the deviation between the model output and the measured values, thereby achieving adaptive adjustment of the model and improving prediction accuracy and robustness.

[0055] The safety threshold is the maximum allowable stress field determined based on the mechanical properties of the aluminum plate material. If the safety threshold is exceeded, the aluminum plate is at risk of developing cracks or deformation defects.

[0056] The parameter adjustment amount is the correction value of the spray pressure, flow rate and cooling medium composition of each temperature zone calculated by the temperature zone coordination intelligent agent network based on the calculated value of the transverse temperature difference and the predicted value of the stress field.

[0057] The parameter correction instruction is a control signal sent by the temperature zone coordination agent network to each temperature zone control unit, which includes the specific value of the parameter adjustment amount and the execution time information.

[0058] The thermal stress prediction model is a hybrid computational model integrating the boundary element method and analytical solution method, used to quickly predict the transient temperature and stress field distributions during the cooling process of aluminum plates. The structure of the thermal stress prediction model is as follows: the input layer receives the initial temperature field matrix, the width dimension of the aluminum plate, the thickness parameter, and the cooling process parameters; the feature extraction layer extracts the spatial features of the temperature distribution using a convolutional neural network; the boundary processing layer uses a boundary element discretization algorithm to discretize the plate surface and edges into boundary units; the analytical solution layer uses Green's function integration to handle the volume force effect caused by the temperature gradient; the time evolution layer uses a refined integration algorithm to calculate the transient response; the dimensionality reduction layer uses Fourier series expansion to transform the three-dimensional problem into a two-dimensional problem for the periodic spray arrangement; the fast calculation layer accelerates the boundary integration calculation through a fast multi-pole algorithm; and the output layer generates the predicted stress field value, the calculated lateral temperature difference value, the target cooling rate sequence for each temperature zone, and the spray pressure reference value.

[0059] The steps for establishing the training dataset for the thermal stress prediction model are as follows: Measured temperature and stress field data of aluminum plates under different cooling processes are collected; temperature-stress coupling field data under multiple operating conditions are generated using finite element simulation software; the measured data and simulation data are fused to form a hybrid dataset; the hybrid dataset is standardized and divided into training, validation, and test sets, with a training-to-validation ratio of 8:2. The training steps for the thermal stress prediction model are as follows: Model parameters are initialized, mean squared error is used as the loss function, adaptive moment estimation optimization algorithm is used to update model weights, the learning rate is set to 0.001 and an exponential decay strategy is adopted; during training, model performance is evaluated on the validation set every 10 iterations; when the validation set loss no longer decreases after 5 consecutive iterations, an early stopping mechanism is triggered; and the model parameters with the best performance on the validation set are saved as the final model.

[0060] The boundary element method (BEM) only requires discretization of the plate surface and edges without meshing the internal regions, transforming the three-dimensional volume integral problem into a two-dimensional surface integral problem. This significantly reduces the number of computational units and is naturally adaptable to complex geometric boundaries. The influence between boundary units is expressed through the Green's function of the fundamental solution, forming a system of boundary integral equations. The analytical solution layer addresses the thermal stress volume force term generated by the internal temperature gradient of the plate by performing integration using the analytical expression of the Green's function of elasticity, avoiding volume discretization and ensuring computational accuracy while reducing computational complexity. The refined integration algorithm transforms the time-domain differential equations into a problem of precise calculation of the exponential matrix. High-precision time stepping is achieved through the addition and multiplication theorems of matrix exponents, maintaining numerical stability for long-term transient problems and avoiding the cumulative errors of traditional time integration methods. The Fourier series expansion utilizes the periodicity of the spray arrangement to expand the three-dimensional temperature and stress fields into Fourier series along the periodic direction. Each series term corresponds to a two-dimensional problem. Dimensionality reduction is achieved by truncating a finite number of series terms, reducing the computational increase caused by spatial dimension. The fast multipole algorithm approximates the interaction between distant boundary elements through multipole expansion, reducing the computational complexity from quadratic to linear, and significantly improving the computational speed for large-scale boundary element problems.

[0061] The hybrid solution approach of the thermal stress prediction model combines the dimensionality reduction characteristics of the boundary element method with the accuracy of analytical methods. This significantly improves computational efficiency while ensuring the accuracy of stress field calculations, enabling the solution speed of transient thermoelastic coupling problems to meet the real-time requirements of online control systems. The model can continuously predict the stress state of the sheet metal during cooling, providing a reliable basis for the dynamic adjustment of parameters in each temperature zone and avoiding sheet metal defects caused by stress concentration. By transforming the three-dimensional transient problem into a boundary two-dimensional problem and a Fourier series superposition, it overcomes the speed bottleneck of traditional finite element methods in real-time calculations, providing an efficient stress monitoring method for multi-temperature zone collaborative control, ensuring the safety of the cooling process and the stability of product quality. The hybrid solution approach enables the model to perform full-field stress field calculations without volumetric meshing when dealing with complex geometric boundaries, simply through surface discretization. The introduction of the analytical Green's function eliminates numerical integration errors in the volumetric force term, while the refined integration algorithm ensures computational stability and accuracy consistency over long-term evolution. Fourier dimension reduction utilizes the regularity of the spray layout in industrial cooling systems, decomposing high-dimensional spatial problems into a superposition of multiple low-dimensional problems. The fast multi-pole algorithm, through hierarchical spatial decomposition and far-field approximation, reduces the solution time for large-scale boundary element systems from hours to seconds. This hybrid solution approach achieves a balance between computational accuracy and efficiency, providing a theoretical foundation and technical support for real-time stress monitoring and early warning.

[0062] The specific implementation methods of the above steps are described in detail below.

[0063] The specific implementation of step S01 involves feeding the aluminum plate to be cooled into the inlet of the multi-temperature zone collaborative cooling system via a conveyor roller. An infrared thermal imaging array is erected above the inlet. The infrared thermal imaging array contains multiple infrared temperature sensors, evenly arranged along the width of the aluminum plate with a sensor spacing of 5mm and a sampling frequency of 50Hz. Temperature data of the aluminum plate surface is collected line by line using a linear array scanning method. The linear array scanning achieves full surface coverage by utilizing the continuous movement of the aluminum plate. One line of temperature data is acquired in each scanning cycle. Multiple lines of data are arranged in chronological order to form a two-dimensional temperature distribution image. The two-dimensional temperature distribution image is stored as an initial temperature field matrix. The number of rows in the initial temperature field matrix corresponds to the number of sampling points along the length of the aluminum plate, and the number of columns corresponds to the number of sampling points along the width of the aluminum plate. The value of each element in the matrix is ​​the temperature measurement value at the corresponding location. At the same time, the width dimension of the aluminum plate is measured by a laser rangefinder, and the thickness parameter of the aluminum plate is measured by an ultrasonic thickness gauge. The width dimension and the thickness parameter are packaged together with the initial temperature field matrix into a dataset, which is transmitted to the input end of the thermal stress prediction model through a data bus. The purpose of this step is to obtain complete thermal state information and geometric parameter information of the aluminum plate before it enters the cooling system, so as to provide basic data for subsequent stress prediction and cooling strategy formulation.

[0064] The specific implementation of step S02 is as follows: After receiving the initial temperature field matrix, the width dimension of the aluminum plate, and the thickness parameters, the thermal stress prediction model first extracts spatial features from the initial temperature field matrix through the feature extraction layer of a convolutional neural network to identify the gradient pattern and abnormal regions of the temperature distribution. Then, the boundary processing layer discretizes the surface and edges of the aluminum plate into several boundary elements according to the requirements of the boundary element method. The boundary element method only requires meshing of the boundaries without discretizing the internal regions. The analytical solution layer calculates the thermal stress volume force caused by the temperature gradient based on the temperature field matrix and obtains the analytical solution of the volume force by performing integration using the Green's function of elasticity. The time evolution layer uses a refined integration algorithm to simulate the transient response of the cooling process. The refined integration algorithm uses matrix exponential operation... To ensure numerical stability over long periods, the dimensionality reduction layer utilizes the periodic arrangement of the spray system to perform Fourier series expansion of the three-dimensional temperature and stress fields along the periodic direction, transforming the three-dimensional problem into a superposition of multiple two-dimensional problems. The fast calculation layer accelerates the solution of the boundary integral equations through a fast multi-pole algorithm. The output layer generates the target cooling rate sequence and spray pressure benchmark value for each temperature zone based on the calculation results. Simultaneously, the model determines whether the highest temperature value in the initial temperature field matrix exceeds 290℃. If it does, a command to start the pulse film-breaking spray mode is sent to the control system; if it does not exceed 290℃, a command to start the continuous spray mode is sent. The purpose of these steps is to quickly predict the stress response of the aluminum plate under different cooling strategies using a hybrid numerical method, providing optimized process parameters for each temperature zone.

[0065] The specific implementation of step S03 is as follows: After receiving the pulse film-breaking injection mode command, the first temperature zone starts the pulse-type high-pressure injection system. The pulse-type high-pressure injection system includes a high-pressure plunger pump, a pulse control valve, and an atomizing nozzle array. The high-pressure plunger pump pressurizes the coolant to 2.0 MPa. The pulse control valve switches between opening and closing at a cycle of 0.2 seconds, with an opening time of 60%, forming a pulse injection with a peak pressure of 2.0 MPa. The pulse pressure change disrupts the stable vapor film on the surface of the high-temperature aluminum plate. A nonionic surfactant with a mass fraction of 0.3% is pre-added to the coolant. The surfactant reduces the surface tension of the liquid and enhances the liquid's resistance to high temperatures. The surface wetting ability is achieved by the atomizing nozzle mixing coolant and high-pressure gas to form an atomized gas-liquid mixture medium. The droplet diameter of the atomized gas-liquid mixture medium ranges from 10 to 50 μm. The specific surface area of ​​the atomized droplets is increased, resulting in a gentler and more uniform cooling process. The cooling rate of the first temperature zone is controlled at 15°C / second by adjusting the spray flow rate. An infrared temperature sensor array is installed at the outlet of the first temperature zone to collect the temperature distribution of the aluminum plate. The temperature distribution data is output as the outlet temperature data of the first temperature zone to the temperature zone coordination intelligent agent network. The purpose of these steps is to achieve rapid cooling in the high-temperature stage through pulse film breaking and atomized cooling, avoiding the decrease in heat transfer efficiency caused by the Leidenfrost effect.

[0066] The specific implementation of step S04 is as follows: After receiving the outlet temperature data of the first temperature zone, the temperature zone coordination agent network processes it through a distributed cooperative control algorithm. Each temperature zone control unit in the network acts as an agent node. Agents exchange state information through a directed communication graph, where the connection matrix describes the topological relationship of information transmission. When the position sensor detects that the front end of the aluminum plate has reached the boundary between the first and second temperature zones, the network calculates the parameter adjustment amount for the second temperature zone. This parameter adjustment amount is calculated based on a consensus algorithm, which makes the state variables of each agent tend to be consistent while satisfying their respective local constraints. The control unit receives an adjustment command 1.2 seconds before the aluminum plate reaches the boundary position and executes parameter modification. The advance adjustment time is determined based on the moving speed of the aluminum plate and the heat conduction time constant. A gradual transition zone with a length of 350mm is set at the boundary position. A variable density nozzle array is installed in the gradual transition zone. The nozzle spacing gradually transitions from the standard spacing of 50mm in the first temperature zone to the standard spacing of 30mm in the second temperature zone. The rate of change of the nozzle spacing is a linear gradient distribution, so that the spray flow rate changes continuously in the gradual transition zone, eliminating temperature jumps and stress abrupt changes. The purpose of this step is to avoid temperature field coupling mismatch and stress concentration at the boundary of adjacent temperature zones through coordinated control between temperature zones and the design of the gradual transition zone.

[0067] The specific implementation of step S05 involves calculating the division scheme of independent control units based on the width of the aluminum plate, dividing the plate width into 9 independent control units. The width of each independent control unit is equal to the width of the aluminum plate divided by 9. The infrared thermal imaging array continuously collects temperature data from each independent control unit, calculates the average temperature value of each independent control unit, and obtains the temperature deviation value by subtracting the average temperature value from the target temperature value. The target temperature value is dynamically given by the thermal stress prediction model based on the cooling process. When the temperature deviation value of the independent control units located on both sides of the aluminum plate exceeds 8°C, the edge compensation control process is triggered. The edge compensation control process first sends an enhancement command to the flow regulating valve corresponding to the edge independent control unit, increasing the nozzle flow rate by 40% based on the current value. Simultaneously, the pitch angle of the nozzle is adjusted to a 53-degree tilt angle via a servo motor. This tilt angle design increases the relative velocity between the jet flow and the plate surface, enhancing the convective heat transfer coefficient, and activates the edge electromagnetic induction heating compensation device. The heating power density of the edge electromagnetic induction heating compensation device is set to 15kW / The heating area covers a 100mm width range around the edge of the plate. The heating power is dynamically adjusted according to the temperature deviation value using a proportional-integral-derivative control algorithm. The purpose of this step is to solve the problem of uneven temperature distribution between the edge and the center during the cooling process of the wide plate. Lateral temperature uniformity is achieved through zone control and edge compensation.

[0068] The specific implementation of step S06 is that the second temperature zone adopts a pure liquid enhanced cooling method. This pure liquid enhanced cooling method stops the gas atomization process and directly sprays liquid cooling medium. The spray pressure is set to 1.8 MPa according to the parameters output by the temperature zone coordination intelligent agent network. The spray pressure is precisely controlled by a variable frequency speed-driven plunger pump. The cooling rate of the second temperature zone is controlled at 35°C / second. This cooling rate is achieved by adjusting the spray flow rate and nozzle coverage density. To compensate for the cooling unevenness in the aluminum plate thickness direction, differentiated spray intensity control in the plate thickness direction is adopted. This differentiated spray intensity control is achieved by adjusting the flow rate ratio of the upper and lower nozzles. The surface spray intensity is set to 1.3 times that of the lower surface, enabling rapid cooling of the surface layer while achieving gentle cooling of the interior through heat conduction. This reduces the temperature gradient and grain size differences in the thickness direction. When the aluminum plate moves to the inlet of the third temperature zone, the cooling medium is switched from liquid to gas via a three-way switching valve. The gas is compressed air or inert gas, and the spray pressure is reduced to 0.5 MPa. The cooling rate of the third temperature zone is reduced to 8°C / second. The gas fine control mode is used for temperature fine-tuning and surface drying in the later stage of cooling. The purpose of these steps is to achieve enhanced cooling in the medium temperature stage and ensure uniformity of the microstructure through differentiated spray control, and to achieve fine temperature control in the low temperature stage.

[0069] The specific implementation of step S07 is as follows: throughout the cooling process, the infrared thermal imaging array and stress sensor array continuously collect temperature field data and stress field data for each temperature zone. The stress sensor array includes piezoelectric stress sensors arranged at the support positions of the aluminum plate conveyor rollers. By measuring the changes in support force, the stress state of the aluminum plate is inferred. The temperature field data and stress field data are transmitted in real time to the thermal stress prediction model for online correction. The online correction process uses a Kalman filter algorithm to fuse the model's predicted values ​​with the measured values, updating the model's internal state variables and weight parameters. The Kalman filter algorithm achieves optimal estimation by minimizing the variance of the prediction error. The thermal stress prediction model recalculates the predicted stress field value and the calculated transverse temperature difference value based on the corrected parameters. The calculated transverse temperature difference value is the maximum temperature difference in the width direction of the aluminum plate. When the calculated transverse temperature difference exceeds 5°C or the predicted stress field exceeds the safety threshold, a dynamic parameter adjustment process is triggered. The safety threshold is set based on the yield strength of the aluminum plate material, typically taking 70% of the yield strength as a reference value. After receiving the over-limit signal, the temperature zone coordination intelligent agent network calculates the parameter adjustment amount for each temperature zone using a distributed optimization algorithm. The distributed optimization algorithm solves for the parameter combination that minimizes the global cost function. The global cost function includes a weighted sum of temperature uniformity terms, stress minimization terms, and energy consumption constraint terms. The calculated parameter adjustment amount is sent to each temperature zone control unit in the form of a parameter correction command. Upon receiving the command, each temperature zone control unit immediately executes the parameter modification to achieve closed-loop control. The purpose of these steps is to dynamically optimize cooling parameters through a combination of real-time monitoring and model prediction, ensuring the safety of the cooling process and the stability of product quality.

[0070] It should be noted that one of the key technical ideas of this invention is the synergistic cooling mechanism of pulsed film-breaking injection and atomized gas-liquid mixed medium. By periodically pulsating pressure to break the stable vapor film on the high-temperature surface, the problem of a sharp drop in heat transfer efficiency caused by the Leiden-Frost effect is solved. Compared with the traditional continuous spraying method, pulsed injection establishes direct liquid-solid contact at the moment the vapor film breaks, and the heat transfer coefficient is restored. The micron-sized droplets of the atomized gas-liquid mixed medium have a larger specific surface area and a gentler cooling characteristic, avoiding the non-uniformity of grain structure caused by pure liquid cooling. The addition of surfactants further enhances the wetting and spreading ability of the liquid. This synergistic mechanism ensures rapid cooling in the high-temperature stage while controlling the spatiotemporal distribution of the cooling rate, creating good initial conditions for the cooling process in the subsequent temperature range.

[0071] The second key technical approach is temperature zone collaborative control and gradual transition zone design based on intelligent agent networks. Traditional multi-temperature zone cooling systems control each temperature zone independently, resulting in temperature jumps and stress concentrations at the boundaries. This invention models each temperature zone controller as an autonomous intelligent agent, achieving parameter linkage between temperature zones through a distributed collaborative control algorithm. The consensus algorithm of the intelligent agent network keeps the cooling states of adjacent temperature zones coordinated, avoiding global performance degradation caused by local optimization. The advance adjustment mechanism uses the outlet temperature data of the upstream temperature zone to predict the inlet state of the downstream temperature zone, enabling the downstream temperature zone to respond in advance. The gradual transition zone achieves a continuous transition of cooling intensity through a variable density nozzle array, eliminating temperature abruptness interfaces. This collaborative control strategy is robust to changes in the number of temperature zones and communication failures, and can adapt to cooling systems of different sizes and complex production conditions.

[0072] The third key technical approach is a hybrid solution method for the thermal stress prediction model. Traditional finite element methods require three-dimensional meshing of the entire plate, resulting in large computational loads and difficulty in meeting real-time control requirements. This invention combines the boundary element method with analytical solution methods, discretizing only the plate surface and edges, transforming the three-dimensional volume integral problem into a two-dimensional surface integral problem, significantly reducing the number of computational units. The analytical Green's function handles the volume force term, avoiding numerical integration errors, while the refined integration algorithm ensures the long-term stability of transient evolution. Fourier series expansion utilizes the periodic characteristics of the spray system to achieve dimensionality reduction, and the fast multi-pole algorithm reduces the computational complexity of boundary integrals from quadratic to linear. This hybrid method ensures the accuracy of stress field calculation while meeting the real-time requirements of online control, providing a reliable prediction basis for dynamic parameter adjustment.

[0073] The synergistic effect of the above-mentioned technical approaches is reflected in the fact that pulsed film-breaking injection solves the heat transfer bottleneck in the high-temperature stage, creating a suitable temperature starting point for subsequent temperature zones; the collaborative control of the intelligent agent network ensures parameter matching and smooth transition between temperature zones, avoiding stress concentration at the interface; and the real-time calculation capability of the thermal stress prediction model enables the system to dynamically adjust the parameters of each temperature zone according to the actual stress state of the plate, forming a complete closed-loop control chain from high-temperature film breaking to temperature zone coordination and stress monitoring. The three support each other and constitute the core technical system of multi-temperature zone coordinated cooling. Compared with traditional methods that only focus on the local optimization of a single temperature zone, this invention achieves overall optimization and dynamic response of the entire process, effectively controlling temperature uniformity and stress level while ensuring cooling efficiency, and improving the forming performance and product quality stability of deep-drawing aluminum plates.

[0074] It should be noted that this invention also solves the following technical problem: the significant decrease in heat transfer efficiency caused by the stable vapor film formed by the Leidenfrost effect on the surface of high-temperature aluminum plates. When the surface temperature of the aluminum plate exceeds 290°C and is higher than the Leidenfrost characteristic of the coolant, a heat-insulating vapor layer forms between the liquid and the high-temperature surface, hindering direct contact between the liquid and the solid surface, significantly reducing the heat transfer coefficient and severely attenuating the cooling efficiency. This invention, by activating a pulse-driven film-breaking injection mode, employs a pulsed high-pressure injection system with a pulse period of 0.2 seconds and a peak pressure of 2.0 MPa. The instantaneous pressure impact generated by the periodic pressure pulsation disrupts the stability of the vapor film, causing it to rupture and restoring direct contact between the liquid and the solid surface. Simultaneously, a surfactant with a mass fraction of 0.3% is added to the cooling medium to reduce surface tension, promoting the wetting and spreading of the liquid on the high-temperature surface and enhancing its penetration ability through the vapor film, significantly improving the heat transfer efficiency at high temperatures and solving the technical problem of decreased cooling efficiency caused by the Leidenfrost effect.

[0075] Furthermore, this invention addresses the technical problem that the real-time stress field calculation speed cannot meet the requirements of online control. Traditional finite element methods (FEMs) require meshing the entire volume of the plate when dealing with three-dimensional transient thermoelastic coupling problems. This results in a large number of computational units and cumulative errors during time integration, making it difficult to meet the response requirements of real-time control systems. The thermal stress prediction model constructed in this invention integrates the boundary element method and analytical solution methods. The boundary element method discretizes only the plate surface and edges, transforming the three-dimensional volume integral into a two-dimensional surface integral. An analytical Green's function is used to handle the volumetric force term caused by temperature gradients, avoiding volume discretization. A refined integration algorithm ensures numerical stability over long periods. Fourier series expansion utilizes the periodicity of the spray arrangement to transform the three-dimensional problem into a two-dimensional one. A fast multi-stage algorithm is introduced to reduce the computational complexity from quadratic to linear, reducing the stress field solution time from hours to seconds, thus meeting the timeliness requirements for real-time monitoring of stress state and dynamic parameter adjustment during cooling.

[0076] Specifically, the principle of this invention is as follows: The reason this invention can solve the technical problem of thermal stress concentration caused by uneven temperature fields during multi-temperature zone cooling is that it establishes an information interaction mechanism between the control units of each temperature zone through a temperature zone coordination intelligent agent network. Each agent optimizes the cooling effect of its own temperature zone while maintaining coordination and consistency with adjacent temperature zones, so that the cooling parameter transformation between temperature zones is no longer isolated but forms a coordinated response. This distributed coordinated control strategy essentially eliminates the control blind spot at the temperature zone boundary. The setting of the gradual transition zone achieves continuous change in cooling intensity through a variable density nozzle array, transforming the original step parameter switching into gradual parameter evolution, eliminating the cause of temperature abrupt changes at the physical level. The division of independent control units in the plate width direction refines the lateral temperature regulation from overall control to precise regional control. Addressing the difference between the heat dissipation conditions at the edges and the central area, closed-loop regulation is formed by enhancing spray flow and electromagnetic induction heating compensation, offsetting the destructive effect of geometric boundary effects on temperature uniformity. The thermal stress prediction model integrates the boundary element method and analytical solution method to realize real-time calculation of the stress field, providing a reliable prediction basis for dynamic parameter adjustment and ensuring that the cooling process always remains within the stress safety range.

[0077] The following provides a specific embodiment 1 of the present invention, and the specific implementation of each step in this embodiment 1 is described in detail below.

[0078] The specific implementation of step S01 involves conveying the aluminum plate to be cooled to the inlet of the multi-temperature zone collaborative cooling system. Real-time temperature acquisition of the aluminum plate surface is performed using an infrared thermal imaging array. This array has a spatial resolution of 5mm, a sampling frequency of 50Hz, and a temperature measurement accuracy of ±2℃. A linear array scanning method is used to construct a temperature distribution image of the entire aluminum plate surface. Initial temperature field matrix. The two-dimensional temperature data array is constructed by dividing the aluminum plate along its length and width, as described below:

[0079] ;

[0080] In the formula, These are the elements of the normalized initial temperature field matrix, dimensionless. This is the length direction index, with values ​​ranging from 1 to... , The total number of measuring points along the length is determined by dividing the length of the aluminum plate by the spatial resolution of 5mm. This is the width direction index, with values ​​ranging from 1 to... , The total number of measuring points in the width direction is determined by dividing the width of the aluminum plate by the spatial resolution of 5mm. For the corresponding position The measured surface temperature value is in °C. For reference temperature, a value of 100℃ is typically used. Also record the width of the aluminum plate. and thickness parameters All units are in mm. The normalized initial temperature field matrix... Aluminum plate width dimensions and thickness parameters The output is fed into the thermal stress prediction model for further calculations.

[0081] The specific implementation of step S02 is to use the normalized initial temperature field matrix obtained in step S01. Aluminum plate width dimensions and thickness parameters The thermal stress prediction model is input for calculation, and the model outputs the target cooling rate sequence for each temperature zone. and spray pressure benchmark value ,in Temperature zone numbering, These correspond to the first, second, and third temperature zones, respectively. The unit is ℃ / second. The unit is MPa. The determination of whether the aluminum plate surface temperature exceeds 290℃ based on the initial temperature field matrix is ​​as follows:

[0082] ;

[0083] In the formula, The maximum temperature value at all measuring points in the initial temperature field matrix is ​​expressed in °C. The pulsed film-breaking injection mode is activated when the above conditions are met; the continuous injection mode is activated when the conditions are not met. This determination is based on the critical temperature point of the Leiden-Frost effect.

[0084] The specific implementation of step S03 involves using a pulsed high-pressure jet system in the first temperature zone to cool the aluminum plate surface with an atomized gas-liquid mixture medium, with a pulse cycle of... Set to 0.2 seconds, peak pressure Based on the spray pressure benchmark value The pressure was set to 2.0 MPa, and a surfactant with a mass fraction of 0.3% was added to the cooling medium to reduce surface tension. The cooling rate... The speed is controlled at 15℃ / second. Collect the outlet temperature data for the first temperature zone and construct the outlet temperature distribution vector. The vector is represented as follows:

[0085] ;

[0086] In the formula, The width direction of the first temperature zone outlet position The temperature values ​​at each measuring point are given in °C. This outlet temperature distribution vector is then calculated. The output is sent to the temperature zone coordination agent network for processing.

[0087] The specific implementation of step S04 is to convert the outlet temperature distribution vector of the first temperature zone. The input temperature zone coordination intelligent agent network performs collaborative processing. When the front end of the aluminum plate reaches the boundary between the first and second temperature zones, the network transmits the outlet temperature data to the second temperature zone control unit. The second temperature zone control unit adjusts the spray parameters 1.2 seconds in advance based on the received data. A 350mm long gradual transition zone is set at the boundary, and a variable density nozzle array is used to achieve a spray flow rate that transitions from the standard flow rate value of the first temperature zone. Standard flow rate to the second temperature zone The continuous transition, the flow distribution function within the transition zone is expressed as follows:

[0088] ;

[0089] In the formula, The distance from the starting point of the boundary within the transition zone The spray flow rate at the location is expressed in units of... , This is the standard flow rate value for the first temperature zone, in units of... , This is the standard flow rate value for the second temperature zone, in units of... , The length of the gradient transition zone is set to 350mm. This is the distance from the starting point of the boundary along the conveying direction, in mm, with a value range of 0 to 350 mm.

[0090] The specific implementation of step S05 is based on the width dimension of the aluminum plate. The board width is divided into 9 independent control units. Temperature data of each independent control unit is continuously acquired using an infrared thermal imaging array, and the temperature deviation value is calculated. Temperature deviation of each independent control unit The statement is as follows:

[0091] ;

[0092] In the formula, For the first Temperature deviation values ​​for each independent control unit, in °C. For the first The measured average temperature of each independent control unit at the current moment, in °C, is obtained by analyzing the temperature of the first independent control unit. The average temperature of all measuring points within the area of ​​each independent control unit is obtained. For the first The target temperature for each independent control unit, expressed in °C, is calculated by a thermal stress prediction model based on a cooling rate sequence. The value range is 1 to 9. The edge independent control unit refers to... or The temperature deviation value of the control unit is denoted as Calculation method and Same. When the temperature deviation value of the independent control unit at the edge... When the temperature exceeds 8°C, the nozzle flow rate of the independent control unit at the edge is increased by 40%, and the spray angle is adjusted to a 53-degree tilt angle. Simultaneously, the edge electromagnetic induction heating compensation device is activated, increasing the heating power density. Set to 15 The heating zone width is 100mm, and the heating power is dynamically adjusted according to the temperature deviation. The adjustment relationship is described as follows:

[0093] ;

[0094] In the formula, The adjusted heating power density, in units of , For the standardized temperature parameter, the empirical value is 10℃.

[0095] The specific implementation of step S06 involves cooling the aluminum plate using a pure liquid enhanced cooling method in the second temperature zone. The spray pressure is set according to the parameters output by the temperature zone coordination intelligent agent network. The statement is as follows:

[0096] ;

[0097] In the formula, This represents the actual spray pressure in the second temperature zone, in MPa. The reference value for the spray pressure in the second temperature zone is typically 1.8 MPa. The average outlet temperature of the first temperature zone is expressed in °C. This is determined by the temperature distribution vector at the outlet of the first temperature zone. The average temperature of all measuring points is obtained by averaging the values. The calculation formula is: , This is the standard temperature parameter, with an empirical value of 200℃. Cooling rate. The temperature is controlled at 35℃ / second. Uniform microstructure is achieved through differentiated spraying intensity control along the thickness of the sheet material. When the aluminum sheet enters the third temperature zone, the cooling medium is switched to a finely controlled gas mode. The cooling rate in the third temperature zone... Reduced to 8℃ / second.

[0098] The specific implementation of step S07 involves continuously collecting temperature field data and stress field data for each temperature zone during the cooling process, and inputting the collected data into the thermal stress prediction model for online correction. (Calculated value of transverse temperature difference) The statement is as follows:

[0099] ;

[0100] In the formula, This is a calculated value for the transverse temperature difference, in °C. This represents the maximum measured temperature of each independent control unit. These are the minimum measured temperatures of each independent control unit, all in °C. The value ranges from 1 to 9. (Stress field prediction value) Output from the thermal stress prediction model, in MPa. Safety threshold. The value is determined based on the yield strength of the aluminum plate material, expressed in MPa, and is empirically considered to be 0.8 times the material's yield strength. This is used when calculating the transverse temperature difference. Temperatures exceeding 5°C or predicted stress field values Exceeding the safety threshold At that time, the temperature zone coordination agent network calculates the parameter adjustment amount and sends parameter correction instructions to each temperature zone control unit.

[0101] The specific implementation of the temperature zone coordinated agent network is a multi-agent system based on a distributed cooperative control strategy, where each temperature zone control unit acts as an independent agent, exchanging status information through a communication network. The control law for each agent is expressed as follows:

[0102] ;

[0103] In the formula, For the first Each temperature zone at time The normalized control input is dimensionless and represents the ratio of the control output to the standard control quantity. This is the local feedback gain coefficient, dimensionless, with a default value of 0.5. The normalized local error signal is dimensionless and is calculated by dividing the difference between the actual cooling rate and the target cooling rate by the target cooling rate. ,in For the first Each temperature zone at time The actual cooling rate, expressed in °C / second. For the first The neighbor set of an agent, For intelligent agents With intelligent agents The communication weight between agents is dimensionless and determined by the elements of the graph's adjacency matrix. It takes a value of 1 when two agents are adjacent, and a value of 0 otherwise. For the first The normalized state variables of an agent, dimensionless, are defined as the ratio of the actual temperature to the reference temperature, i.e. ,in For the first Each temperature zone at time The average temperature, in °C. For the first Normalized state variables of an agent, dimensionless, defined in the same way as... same.

[0104] The specific implementation of the thermal stress prediction model is a hybrid computational model that integrates the boundary element method and analytical solution methods. The boundary element method discretizes the surface of the plate into... One boundary unit, Given the total number of boundary elements, the boundary integral equation is expressed as follows:

[0105] ;;

[0106] In the formula, is the coefficient of the free term, dimensionless, and takes a value of 0.5 for smooth boundaries. Boundary unit The normalized displacement components on the reference length are dimensionless and are obtained by removing the actual displacement. The normalized displacement kernel function of the fundamental solution is dimensionless. The normalized surface force kernel function of the fundamental solution is dimensionless. The normalized actual displacement on the boundary is dimensionless. The normalized actual surface force on the boundary is dimensionless. For the boundary surface, the equation eliminates the dimensional inconsistency problem through normalization. The analytical solution layer, addressing the thermal stress volume force term generated by the temperature gradient, is expressed using Green's function integral as follows:

[0107] ;

[0108] In the formula, The normalized thermal stress components are dimensionless. The coefficient of thermal expansion of the material is given by 1000 ppm. , This refers to the elastic modulus, expressed in gigabytes of pressure (GPa). This is a reference stress, expressed in GPa, with an empirical value of 100 GPa, used for normalization. Green's function, unit: , indicating the source point Opposing point The impact, Temperature gradient, unit: , This refers to the volume of the sheet material, in units of... The precise integration algorithm transforms the time-domain differential equation into an exponential matrix calculation. The time-stepping format is as follows:

[0109] ;

[0110] In the formula, For a moment The normalized state vector is dimensionless and contains the normalized temperature and stress components for each temperature region. The time step is in seconds. This is the system matrix, with units of . , This is the normalized external force vector, in units of . , It is an identity matrix, dimensionless. The matrix exponent is dimensionless. This is the inverse of the system matrix, expressed in seconds.

[0111] To better understand and implement this invention, Example 2 of a specific application scenario is provided below: A batch of 5182 aluminum alloy deep-drawing sheets with a thickness of 2.5mm, a width of 1850mm, and a length of 6000mm are required to achieve the uniformity of the microstructure of the cooled sheet to meet the deep-drawing forming Grade II standard, and the yield strength deviation needs to be controlled within ±8MPa. Traditional single-temperature zone cooling methods suffer from overcooling at the edges and uneven cooling in the central area when processing aluminum sheets of this size, leading to localized cracking and orange peel defects during deep-drawing, resulting in a finished product qualification rate of only 72%. The technical team decided to use the multi-temperature zone synergistic cooling method of this invention to cool the aluminum sheets, such as... Figure 2 As shown. The initial temperature of the aluminum sheet exiting the hot rolling process is 485℃. A 300×370 dimensional initial temperature field matrix was constructed by using an infrared thermal imaging array to perform a full-area scan of the sheet surface at a frequency of 50Hz, with a spatial resolution of 5mm. The temperature measurement accuracy reached ±2℃. The temperature field matrix shows that the temperature in the center of the aluminum sheet is 487℃, the edge temperature is 478℃, and a high-temperature point of 492℃ exists in a local area on the left edge. After inputting the initial temperature field matrix, width 1850mm, and thickness 2.5mm into the thermal stress prediction model, the model completed the calculation within 0.8 seconds, outputting the target cooling rate of 15℃ / second for the first temperature zone, 35℃ / second for the second temperature zone, and 8℃ / second for the third temperature zone. The spray pressure baseline value was set to 2.0MPa. Figure 3As shown. Because the surface temperature of the aluminum plate exceeds 290℃, the system automatically starts the pulse film breaking spray mode, with the pulse period set to 0.2 seconds, the peak pressure to 2.0MPa, and the valley pressure to 0.6MPa.

[0112] The first temperature zone uses an atomized gas-liquid mixture for cooling. 0.3% (w / w) of a nonionic surfactant, polyoxyethylene ether, is added to the coolant to reduce the surface tension from 72 mN / m to 28 mN / m, enhancing the wettability of the liquid on the high-temperature aluminum plate surface. The pulsed high-pressure jet system disrupts the vapor film formed on the aluminum plate surface through periodic pressure fluctuations, achieving an instantaneous impact force of 350 N / m at the pressure peak. This reduces the vapor film thickness from a stable 1.8 mm to below 0.3 mm, restoring direct contact between the liquid and solid surface. The atomizing nozzle breaks the coolant into droplets with an average particle size of 45 μm, increasing the specific surface area to [missing information]. / The cooling rate is precisely controlled at 15℃ / second. After the aluminum plate stays in the first temperature zone for 12 seconds, the outlet temperature drops to 305℃, and the temperature field data is transmitted to the temperature zone coordination intelligent network in real time. When the front end of the aluminum plate reaches the boundary between the first and second temperature zones, the second temperature zone control unit starts adjusting the spray parameters 1.2 seconds in advance, setting a 350mm long gradual transition zone at the boundary. Through a variable density nozzle array, the spray flow rate is linearly transitioned from the standard flow rate of 120L / min in the first temperature zone to the standard flow rate of 280L / min in the second temperature zone, eliminating the temperature step phenomenon that may occur at the temperature zone boundary.

[0113] The system divides the 1850mm wide aluminum plate into 9 independent control units, each unit being approximately 205mm wide, as shown in Table 1.

[0114] Table 1 Temperature Distribution Data for Independent Control Units

[0115]

[0116] Table 1 shows that the temperature deviations of Unit 1 and Unit 9 are 13℃ and 17℃ respectively, both exceeding the 8℃ threshold. The system immediately increases the nozzle flow rate of the two independent edge control units by 40% and adjusts the spray angle from vertical to a 53-degree tilt, causing the coolant to cover the edge area in an elliptical trajectory, increasing the coolant's residence time at the edge. Simultaneously, the edge electromagnetic induction heating compensation device is activated, with a heating area width of 100mm and a heating power density set to 15kW / m². Eddy currents are induced at the edges of the aluminum plate by a high-frequency alternating magnetic field, compensating for temperature loss caused by rapid heat dissipation at the edges. Figure 4 As shown. After 6 seconds of compensatory heating and differentiated spray adjustment, the edge temperature deviation was reduced to within 5°C.

[0117] After the aluminum plate enters the second temperature zone, the system switches to a pure liquid enhanced cooling mode. The spray pressure is set to 1.8 MPa according to the parameters output by the temperature zone coordination intelligent agent network, and the cooling rate is increased to 35℃ / second. Considering the 2.5mm thickness of the plate, the system employs a differentiated spray intensity control strategy: the surface nozzle flow rate is 320 L / min, and the inner nozzle flow rate is 240 L / min. By adjusting the nozzle angle, the surface layer achieves stronger cooling intensity, while the interior achieves a relatively gentle cooling process through heat conduction, reducing grain size differences in the thickness direction. The residence time in the second temperature zone is 8 seconds, and the aluminum plate outlet temperature drops to 125℃, significantly improving temperature field uniformity. When the aluminum plate enters the third temperature zone, the system switches the cooling medium to a gas fine-control mode, using compressed air as the cooling medium. The cooling rate is reduced to 8℃ / second to avoid increased residual stress caused by excessively rapid cooling in the low-temperature stage. Figure 5 As shown.

[0118] Throughout the cooling process, an infrared thermal imaging array continuously acquired temperature field data for each temperature zone, while a stress sensor array simultaneously acquired stress field data for the sheet metal at a frequency of 50 Hz. The thermal stress prediction model performed online corrections on the acquired data. The sheet metal surface was discretized into 2800 boundary elements using the boundary element method. The volume force effect caused by the temperature gradient was handled using Green's function integration. A refined integration algorithm was used to calculate the transient response. Fourier series expansion transformed the three-dimensional problem into a superposition of 15 two-dimensional problems. A fast multipole algorithm reduced the computational complexity to linear levels. The model's output stress field predictions showed that the maximum principal stress in the central region of the aluminum sheet was 85 MPa, and the maximum principal stress in the edge region was 92 MPa, both below the safety threshold of 120 MPa for 5182 aluminum alloy. The calculated lateral temperature difference during the cooling process reached a maximum of 4.2℃, consistently remaining within the 5℃ threshold. At the 18th second, the calculated value of the transverse temperature difference instantly rose to 5.8℃. The temperature zone coordination intelligent agent network immediately calculated the parameter adjustment amount and sent a parameter correction command to the second temperature zone control unit, adjusting the spray pressure from 1.8MPa to 1.65MPa and reducing the cooling rate from 35℃ / second to 32℃ / second. After 2 seconds of parameter adjustment, the transverse temperature difference decreased to 3.5℃, and the predicted stress field value returned to the safe range.

[0119] After cooling, the final temperature of the aluminum sheet was 68℃, with a temperature field uniformity of ±3℃. Yield strength tests showed that the yield strength deviation at different locations was controlled within ±5MPa, with uniform grain size distribution, an average grain diameter of 18μm, and a standard deviation of 1.2μm. Deep drawing tests indicated that the forming limit curve of this batch of aluminum sheets was within the deep drawing forming Class II standard range, with a cupping test value reaching 11.2mm. The finished product qualification rate increased to 96%, and localized cracking and orange peel defects were effectively eliminated.

[0120] The technological advancements of this invention compared to traditional single-temperature-zone cooling methods are reflected in several aspects. First, the pulsed film-breaking spray mode disrupts the stable vapor film formed by the Leidenfrost effect through periodic pressure pulsations, solving the problem of decreased heat transfer coefficient caused by the vapor layer between the liquid and solid surfaces at high temperatures. The instantaneous pressure impact ruptures the vapor film and restores direct contact, significantly improving cooling efficiency at high temperatures. Second, the temperature-zone coordinated intelligent agent network achieves parameter coordinating optimization across temperature zones based on a distributed collaborative control strategy. Through event triggering mechanisms and consensus algorithms, it enables a smooth transition of cooling parameters at the boundaries of adjacent temperature zones, eliminating temperature abrupt changes and stress concentration at temperature zone boundaries in traditional methods. Third, the strategy of dividing the plate width into multiple independent control units and coordinating with an edge electromagnetic induction heating compensation device allows for differentiated control based on the different physical characteristics of the edge heat dissipation conditions compared to the central region. By enhancing the edge spray flow rate and compensating for heating, it effectively solves the problem of temperature unevenness across the plate width. Furthermore, the thermal stress prediction model employs a hybrid computational architecture that integrates the boundary element method (BEM) and analytical solution methods. Leveraging the dimensionality reduction properties of the BEM, the three-dimensional volume integral problem is transformed into a two-dimensional surface integral problem. The analytical expression of the Green's function is used to handle the thermal stress volume force term, and a refined integration algorithm ensures the numerical stability of long-term transient problems. A fast multi-stage algorithm reduces the solution time for large-scale boundary element systems from hours to seconds, enabling the model to continuously predict the stress state of the sheet metal during cooling and provide real-time data for dynamic parameter adjustments. Finally, differentiated spraying intensity control along the sheet thickness reduces grain size differences in the thickness direction by adjusting the cooling intensity difference between the surface and inner layers. This combination of strong surface cooling and gentle internal heat conduction reduces the microstructure uniformity required for deep-drawing performance, ensuring the necessary uniformity of microstructure for deep-drawing forming performance from a physical mechanism perspective. These technological advancements enable the multi-temperature zone synergistic cooling method to achieve precise control of the temperature and stress fields while ensuring cooling efficiency, fundamentally solving the problems of microstructure uniformity and forming performance during the cooling process of large-size aluminum sheets.

[0121] It should be noted that the variables involved in this invention are explained in detail in Tables 2 and 3.

[0122] Table 2. Variable Explanation Table (Part 1)

[0123]

[0124] Table 3. Variable Explanation Table (Part Two)

[0125]

[0126] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-temperature zone synergistic cooling method for deep-drawing aluminum sheets, characterized in that, The method comprises collecting aluminum plate surface temperature distribution data to establish an initial temperature field matrix, inputting a thermal stress prediction model to calculate and output a target cooling rate sequence and a spraying pressure reference value of each temperature zone, judging to start a pulse membrane-breaking spraying mode or a continuous spraying mode according to the surface temperature, adopting a pulse high-pressure spraying system to spray an atomized gas-liquid mixed medium in the first temperature zone, transmitting temperature data between temperature zones through a temperature zone coordination intelligent agent network and adjusting the spraying parameters in advance, setting a gradual transition zone at the junction of adjacent temperature zones to realize continuous transition of the spraying flow, dividing the plate width direction into independent control units and enhancing the flow and adjusting the angle of the edge independent control units, adopting a pure liquid enhanced cooling mode in the second temperature zone and realizing uniformization of the structure through differential spraying intensity control in the thickness direction of the plate, continuously collecting temperature field data and stress field data to input the thermal stress prediction model for online correction and output corrected stress field prediction values and transverse temperature difference calculation values, and sending parameter correction instructions to each temperature zone control unit when the transverse temperature difference calculation values or the stress field prediction values exceed the threshold values.

2. The multi-temperature zone synergistic cooling method of deep-drawing aluminum sheet according to claim 1, characterized in that, The infrared thermal imaging array is used to collect the aluminum plate surface temperature distribution data, the spatial resolution is 5 mm, the sampling frequency is 50 Hz, the temperature distribution image of the whole surface of the aluminum plate is constructed through line array scanning, and the temperature measurement accuracy is ±2℃.

3. The multi-temperature zone synergistic cooling method of deep-drawing aluminum sheet according to claim 2, characterized in that, The initial temperature field matrix is a two-dimensional temperature data array divided according to the length direction and the width direction of the aluminum plate, and the matrix elements record the surface temperature values of the corresponding positions.

4. The multi-temperature zone synergistic cooling method of deep-drawn aluminum sheet according to claim 3, characterized in that, The pulse membrane-breaking spraying mode breaks the stable vapor film formed by the Leidenfrost effect through periodic pressure pulsation, and the pulse high-pressure spraying breaks the vapor film through instantaneous pressure impact to restore the direct contact between the liquid and the solid surface.

5. The multi-temperature zone synergistic cooling method of deep-drawn aluminum sheet according to claim 4, characterized in that, The atomized gas-liquid mixed medium is a two-phase flow cooling medium formed by atomizing the cooling liquid into micron-sized droplets through high-pressure gas flow and mixing with the gas.

6. The multi-temperature zone synergistic cooling method of a deep-drawn aluminum sheet according to claim 5, characterized by, A surfactant is added to the cooling medium to reduce the surface tension of the cooling liquid and promote the wetting and spreading behavior of the liquid on the high-temperature surface of the aluminum plate.

7. The multi-temperature zone synergistic cooling method of deep-drawn aluminum sheet according to claim 6, characterized in that, The temperature zone coordination intelligent agent network is a multi-agent system based on a distributed collaborative control strategy, each temperature zone control unit acts as an independent agent to exchange state information through a communication network, and each agent maintains the performance index of the temperature zone as a local objective function and maintains the coordination consistency with adjacent temperature zones as a global objective function.

8. The multi-temperature zone synergistic cooling method of a deep-drawn aluminum sheet according to claim 7, characterized by, The gradual transition zone is a cooling parameter continuous change region set at the junction of adjacent temperature zones, the spatial gradient distribution of the spraying flow is realized by changing the nozzle arrangement density, and the temperature step phenomenon at the temperature zone boundary is eliminated.

9. The multi-zone synergistic cooling method of deep-drawn aluminum sheet according to claim 8, characterized in that, The independent control unit is a temperature regulation and control region divided according to the width direction of the aluminum plate, and each independent control unit is provided with an independent spraying valve group and a flow adjusting device.

10. The multi-zone synergistic cooling method of a deep-drawn aluminum sheet according to claim 9, characterized in that, The edge independent control unit is an independent control unit located at the edges of the width direction of the aluminum plate, when the temperature deviation value of the edge independent control unit exceeds the threshold value, the nozzle flow of the edge independent control unit is enhanced and the spraying angle is adjusted to an inclination angle.

Images