Temperature zoning regulation and control method and system for automobile double-door-ring hot-pressing mold

Abstract

The invention belongs to the technical field of temperature control, and relates to a temperature zoning regulation and control method and system for an automobile double-door-ring hot-pressing mold. The method comprises the following steps: synchronously obtaining multi-mode working condition data of a hot pressing mold in a pressure maintaining stage, and carrying out smooth denoising and numerical mapping processing to obtain a real-time working condition state vector; calculating a phase change activeness characteristic and measuring a dynamic thermal coupling characteristic under the common influence of physical interface thermal resistance and phase change latent heat by combining a thermodynamic change rule of the material; introducing the dynamic thermal coupling characteristics into a time correlation kernel function of a nonlinear regression prediction model, and predicting a nonlinear temperature evolution trajectory of the target area in a future control period; and adding thermal stress penalty constraint aiming at space temperature difference in a cost evaluation process of controlling optimization so as to output a cooling water path control instruction with the minimum global cost. The method improves the prediction precision of the temperature track in the machining process of the double-door-ring part, and effectively prevents the part from cracking.

Description

Technical Field

[0001] This invention belongs to the field of temperature control technology, specifically relating to a method and system for temperature zone control of a hot pressing mold for automotive double door rings. Background Technology

[0002] In modern automotive manufacturing, double door ring parts are typically produced using laser welding technology, which joins sheets of varying thicknesses and strengths into a single unit. After being heated to their austenitizing temperature, these welded sheets are rapidly transferred to a hot press mold for stamping and subsequent pressure cooling. During this pressure cooling phase, an austenitic-to-martensite phase transformation occurs within the material, resulting in ultra-high strength physical properties. In this complex process, precise control of the internal temperature field of the mold is crucial for ensuring the yield strength and dimensional accuracy of the final structural component.

[0003] Currently, the mainstream solutions for temperature control of hot pressing molds in the industry typically employ independent proportional-integral-derivative (PID) control algorithms or basic model predictive control (MMC) algorithms. These conventional solutions mainly rely on thermocouples arranged inside the hot pressing mold. Based on the static deviation between the real-time temperature collected by sensors and the target temperature set by the process, they directly provide feedback and calculate and output control signals to adjust the valve openings of the cooling water circuits in each zone, aiming to achieve independent cooling and temperature reduction targets for each area.

[0004] However, the unique structure of automotive double-door ring parts, characterized by adjacent material thickness differences and the release of intense latent heat during cooling, makes the actual engineering scenario extremely complex. During the mold-closing and pressure-holding cooling stage, the rapid cooling of the thicker plate area triggers a martensitic phase transformation, releasing a significant amount of latent heat within the material in a short period. Simultaneously, the volume expansion caused by metal lattice reconstruction leads to a sudden increase in local contact pressure between the plate and the mold. This micromechanical change causes a highly nonlinear abrupt change in the physical interface thermal resistance between the mold and the plate. The rapid decay of the dynamic contact thermal resistance and the concentrated release of the massive latent heat result in a physical superposition, severely distorting existing prediction models that treat interface thermal resistance as a static constant. This leads to a significant mismatch between the water valve opening calculated by the control algorithm and the actual heat dissipation requirements.

[0005] The aforementioned control mismatch often leads to insufficient cooling rates in the thick material area, or excessive cooling by keeping the water valve fully open for extended periods in an attempt to forcibly lower the temperature of the thick material area. This excessive cooling energy undergoes lateral heat conduction through the mold's metal matrix, severely interfering with adjacent thin material areas and causing unexpected and rapid temperature drops in those areas. This cross-regional heat conduction interference creates an extremely steep spatial temperature gradient at the interface between thick and thin materials. When the material is in a phase transition phase and exhibits brittle sensitivity, the significant difference in spatial thermal stress directly causes the thin material area to shrink too quickly, disrupting the overall deformation balance of the part. This ultimately induces severe local shrinkage and warping, and may even directly cause cracking of the weld at the interface between thick and thin materials, thus limiting the molding yield of double-ring parts. Summary of the Invention

[0006] The purpose of this invention is to propose a temperature zone control method and system for automotive double door ring hot pressing molds, in order to solve the technical problem that the superposition of dynamic thermal resistance abrupt change and latent heat release of phase change in complex engineering scenarios of double door rings leads to the distortion of temperature evolution prediction, which in turn causes shrinkage and warping at the thickness-thin interface and weld cracking.

[0007] To solve the above problems, the technical solution of the temperature zone control method for automotive double-door ring hot pressing mold proposed in this invention is as follows: A method for temperature zone control of hot pressing molds for automotive double door rings includes: Simultaneously acquire multimodal working condition data of hot pressing mold in the pressure holding stage and perform smoothing, noise reduction and numerical mapping processing to obtain the real-time working condition state vector at the current moment. Based on the temperature and pressure states reflected by the real-time operating condition state vector, and combined with the thermodynamic change law of the material, the phase transformation activity characteristics used to characterize the degree of martensitic phase transformation of the material and the dynamic thermal coupling characteristics used to measure the combined influence of physical interface thermal resistance and latent heat of phase transformation are calculated. The dynamic thermal coupling feature is introduced into the time correlation kernel function of the nonlinear regression prediction model. By combining the historical working condition state vector and the current real-time working condition state vector, the nonlinear temperature evolution trajectory of the target thick material area and the adjacent thin material area in the future control cycle is predicted. Based on the nonlinear temperature evolution trajectory and the phase transition activity characteristics, a thermal stress penalty constraint for spatial temperature differences is added during the cost evaluation process of control optimization to find and output the cooling water circuit control command that minimizes the global cost.

[0008] Beneficial effects: This invention obtains an accurate real-time operating condition vector by synchronously acquiring multimodal operating condition data and performing smoothing, denoising, and numerical mapping processing; by calculating phase change activity characteristics and measuring the dynamic thermal coupling characteristics that measure the combined influence of physical interface thermal resistance and latent heat of phase change, it achieves accurate extraction of the intensity of phase change and dynamic nonlinear thermal resistance changes within the material; by introducing dynamic thermal coupling characteristics into the time correlation kernel function of the nonlinear regression prediction model, it improves the accuracy of predicting the nonlinear temperature evolution trajectory of the target thick material region and adjacent thin material regions in future control cycles, eliminating prediction distortion; by adding thermal stress penalty constraints for spatial temperature differences in the cost evaluation process of control optimization, it finds and outputs the cooling water circuit control command that minimizes the global cost, effectively suppressing the phenomenon of excessive temperature gradient at the thick-thin interface, fundamentally avoiding shrinkage, warping, and cracking of parts, and achieving precise temperature zoning control of automotive double-door ring hot pressing molds.

[0009] Furthermore, the formula for calculating the phase transition activity characteristic is as follows:

[0010] In the formula, Indicates at a time node The phase transition activity characteristics, Indicates at a time node Real-time temperature of the secondary surface of the mold. This indicates the temperature at which the material's inherent martensitic phase transformation begins. This represents the pre-set phase transition temperature window constant.

[0011] Beneficial effects: This formula reflects the intensity of phase transformation and the latent heat release law of metallic materials when they are close to the martensitic phase transformation initiation temperature through an exponential mathematical relationship. It transforms the complex thermodynamic phase transformation process of materials into a continuous state variable that can be directly processed by the control algorithm, and realizes the accurate characterization of the martensitic phase transformation behavior of materials.

[0012] Furthermore, the calculation formula for the dynamic thermal coupling characteristic is as follows:

[0013] In the formula, Indicates at a time node The dynamic thermal coupling characteristics, Indicates at a time node The real-time contact pressure obtained This represents the standard rated holding pressure constant. and These represent the first weighting coefficient and the second weighting coefficient, respectively.

[0014] Beneficial effects: This formula accurately reflects the dynamic heat conduction resistance effect under the combined action of the interface thermal resistance at the mold and the sheet metal during the pressure holding process by superimposing the interface thermal resistance exponential decay term dominated by the contact pressure and the apparent thermal resistance logarithmic growth term dominated by the latent heat of phase change. It eliminates the prediction benchmark bias caused by the traditional model treating the interface thermal resistance as a static constant. At the same time, by setting the two weighting coefficients, it can flexibly adapt to the thermal coupling characteristics under different material properties and pressure holding conditions.

[0015] Furthermore, the formula for calculating the time correlation kernel function is as follows:

[0016] In the formula, Indicates at a time node and time nodes The covariance kernel function values ​​between them Represents the signal variance parameter. Indicates the time scale parameter. and Represented at the time node and the time nodes The calculated dynamic thermal coupling characteristics.

[0017] Beneficial effects: By embedding dynamic thermal coupling feature difference constraint terms into the time correlation kernel function of the nonlinear regression prediction model, the model can adaptively adjust the reference weights of historical observation data at different times according to the changes in the thermal conduction boundary conditions inside the mold. When the thermodynamic state difference between two comparison times is large, the correlation of the corresponding data will be automatically attenuated, avoiding the interference of the drastically different thermal conduction characteristics before and after the phase change on the prediction results. This enhances the reliability of the model in predicting the latent heat disturbance conditions of the phase change, solves the problem of temperature trajectory prediction distortion in the phase change stage of traditional prediction models from the algorithm level, and improves the prediction accuracy of temperature evolution trajectory within the future control cycle.

[0018] Furthermore, the addition of thermal stress penalty constraints for space temperature differences during the cost evaluation process of control optimization includes: Calculate the absolute value of the predicted temperature difference between the target thick material region and the adjacent thin material region at each prediction step size; The absolute value is compared to the product of the physical Euclidean distance between the target thick material region and the adjacent thin material region and the reference safe temperature gradient; After calculating the square of the ratio, multiply it by the sum of the phase transition activity characteristics of the target thick material region and the adjacent thin material region, as well as the penalty weight constant; The calculation results of all predicted step sizes are summed to obtain the cross-regional thermal stress penalty term, which is used as the thermal stress penalty constraint for the spatial temperature difference.

[0019] Beneficial effects: It enables the control optimization process to accurately perceive the thermal stress risk at the interface of different material thicknesses, especially the cracking risk when the material is in the brittle stage of phase transformation. While ensuring that the cooling rate of each region meets the requirements of martensitic phase transformation, it can autonomously balance the cooling intensity of different zones, avoiding the problem of excessive temperature gradient between adjacent regions caused by strong cooling of a single zone.

[0020] Furthermore, the multimodal operating condition data includes the real-time temperature of the mold subsurface, local contact pressure, coolant inlet temperature, coolant outlet temperature, and the real-time flow rate corresponding to the current valve opening.

[0021] Furthermore, in the process of smoothing, denoising, and numerical mapping of multimodal operating condition data, electromagnetic interference noise is eliminated by a moving average filtering algorithm, and the flow rate data and pressure data are linearly normalized to construct the real-time operating condition state vector at the current moment.

[0022] Furthermore, the cooling water circuit control command includes a valve opening control sequence generated by a quadratic programming algorithm under given valve maximum opening constraints and flow upper limit constraints.

[0023] Beneficial effects: It ensures that the generated cooling water circuit control commands not only achieve the optimal goal of minimizing global cost, but are also strictly constrained by the objective boundary constraints of the cooling water circuit physical actuators, thus guaranteeing the safe and executable nature of the control commands.

[0024] Furthermore, the first control command of the valve opening control sequence is extracted, converted into an electrical pulse signal, and output to the proportional-integral regulating water valve of the corresponding zone to perform dynamic allocation of the cooling water circuit.

[0025] The technical solution of the temperature zone control system for automotive double-door ring hot pressing mold proposed in this invention is as follows: A temperature zone control system for a double-door ring hot press mold for automobiles includes a processor and a memory. The memory stores computer program instructions. When the computer program instructions are executed by the processor, the temperature zone control method for the double-door ring hot press mold for automobiles described in any of the above technical solutions is implemented.

[0026] The beneficial effects of this invention are as follows: This invention constructs a complete closed-loop control logic from multimodal working condition data acquisition and preprocessing to core feature extraction, temperature trajectory prediction, and then to constrained control optimization, breaking the limitation of the traditional control framework that treats heat conduction parameters as static constants; for complex working conditions with different material thicknesses of double gate rings accompanied by severe latent heat of phase change, this invention can predict in advance the temperature field changes caused by the disturbance of latent heat of phase change of materials and the sudden change of interface thermal resistance, improve the prediction accuracy of mold temperature trajectory, and enable the adjustment of cooling water channels to match the actual heat dissipation requirements of different areas, effectively avoiding the problems of insufficient cooling rate in the thick material area and warping and weld cracking at the junction caused by the interference of cold amount in the adjacent thin material area, thus ensuring the molding quality and mechanical properties of double gate ring parts from the root. Attached Figure Description

[0027] Figure 1 This is a flowchart of the temperature zone control method for the hot pressing mold of automotive double door rings according to an embodiment of the present invention; Figure 2 This is a comparison chart of the data processing results of the cooling curve of the thick material region in an embodiment of the present invention; Figure 3 This is a comparison diagram of the temperature difference evolution at the interface between adjacent thick and thin materials in an embodiment of the present invention. Detailed Implementation

[0028] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0029] Specific embodiments of the temperature zone control method for automotive double-door ring hot pressing molds proposed in this invention: like Figure 1 As shown, the temperature zone control method for the automotive double-door ring hot pressing mold in this embodiment includes the following steps: S1. Simultaneously acquire multimodal working condition data of the hot pressing mold in the pressure holding stage, and perform smoothing, noise reduction and numerical mapping processing to obtain the real-time working condition state vector at the current moment.

[0030] In this step, high-temperature thermocouples and miniature pressure sensors are pre-embedded 2mm to 5mm below the cavity area corresponding to different thicknesses of the hot pressing mold, and high-frequency flow meters and water temperature sensors are installed in the inlet and outlet pipes of the cooling water circuits in each zone.

[0031] During the die-closing and pressure-holding process of the stamping equipment, initial multimodal operating condition data are synchronously acquired at a sampling frequency of 10Hz. The acquired underlying physical data includes the real-time temperature of the die subsurface, local contact pressure, coolant inlet temperature, coolant outlet temperature, and the real-time flow rate corresponding to the current valve opening.

[0032] After acquiring the raw data, the smoothing, denoising, and numerical mapping processes are initiated. The industrial control computer uses a moving average filtering algorithm on the acquired high-frequency time series data to eliminate high-frequency electromagnetic interference noise caused by the operation of heavy-duty stamping equipment on site, and performs linear normalization on flow and pressure data due to significant differences in dimensions. For example, if five discrete temperature values ​​of thermocouples are continuously acquired within a half-second sampling period, the industrial control computer sums the five readings and divides them by 5 for mean smoothing. This operation effectively eliminates abnormal spike signals caused by abrupt jumps due to electromagnetic pulses. Through the above processing steps, the real-time operating condition state vector of each mold partition at the current moment is constructed and temporarily stored in the high-speed memory of the industrial control computer for subsequent prediction model calls.

[0033] In this way, through multi-dimensional synchronous acquisition and noise reduction preprocessing of multimodal working condition data, the system can obtain a set of real-time working condition features of the mold with high fidelity and high signal-to-noise ratio, providing reliable data support for subsequent precise temperature zone control.

[0034] S2. Based on the temperature and pressure states reflected by the real-time operating condition state vector, and combined with the thermodynamic change law of the material, calculate the phase transformation activity characteristics used to characterize the degree of martensitic phase transformation of the material, as well as the dynamic thermal coupling characteristics used to measure the combined influence of physical interface thermal resistance and latent heat of phase transformation.

[0035] In this step, the phase transformation activity characteristics are first calculated based on the physical approximation between the real-time temperature of the mold's subsurface and the material's inherent phase transformation initiation temperature. When the double-ring thick plate is cooled to the martensitic phase transformation initiation temperature, the internal phase transformation of the material is most intense, and the latent heat energy released reaches its peak. Therefore, this step utilizes the smooth transition characteristics of the Gaussian distribution function to fit the peak state of latent heat release in the metallic material within a specific temperature range. The formula for calculating the phase transformation activity characteristics is:

[0036] In the formula, Indicates at a time node The phase transition activity characteristics, Indicates at a time node Real-time temperature of the mold subsurface obtained by thermocouple; This indicates the inherent martensitic phase transformation initiation temperature of the material. This value is obtained from the material's manufacturer's property data sheet and is preset in the system. This represents the preset phase transition temperature window constant, used to control the width of the phase transition influence range, and its value is greater than 0.

[0037] The principle behind this formula is that when the real-time temperature of the mold subsurface is closer to the martensitic phase transformation initiation temperature, the value of the molecule approaches 0, making the result of the entire exponential function approach 1, accurately representing that the current moment is in the peak period of latent heat release during phase transformation; conversely, when the real-time temperature of the mold deviates from the set phase transformation temperature range, the value of the phase transformation activity characteristic will rapidly and smoothly decay to close to 0.

[0038] Furthermore, this step combines local contact pressure calculation to measure the dynamic thermal coupling characteristics influenced by the combined effects of physical interface thermal resistance and latent heat of phase change. During high-pressure holding and cooling, the greater the local contact pressure, the tighter the fit between the sheet metal and the mold surface, and the smaller the interface contact thermal resistance. Simultaneously, when the phase change activity is high, the latent heat released violently within the material resists external cooling capacity, manifesting as a logarithmic increase in apparent heat conduction resistance. This step superimposes the exponential decay law of contact thermal resistance caused by contact pressure with the logarithmic buffering law of latent heat release from a thermodynamic perspective to obtain a dynamic thermal coupling characteristic that accurately characterizes the degree of nonlinear heat conduction resistance within the mold at the current moment. The calculation formula is as follows:

[0039] In the formula, Indicates at a time node Dynamic thermal coupling characteristics; This indicates the real-time contact pressure acquired by the miniature pressure sensor. This represents the standard rated holding pressure constant initially set by the hydraulic press. This constant value is greater than 0 to avoid logical errors where the denominator is 0. and Let represent the first weighting coefficient and the second weighting coefficient, respectively. Both are dimensionless empirical weighting coefficients and their sum is 1.

[0040] This formula shows that an increase in real-time contact pressure leads to a decrease in the first term, indicating a reduction in physical contact thermal resistance. Conversely, an increase in the phase transition activity characteristic leads to an increase in the second term, indicating that latent heat offsets some of the external cooling capacity, effectively increasing the overall system thermal resistance. The second term uses the natural logarithm function with a constant of 1 added to the logarithm, ensuring that even at the most intense phase transition (i.e., when the phase transition activity characteristic equals 1), the increase in this characteristic factor exhibits a smooth convergence, preventing the extreme case of unbounded numerical amplification.

[0041] S3. Introduce the dynamic thermal coupling characteristics into the time correlation kernel function of the nonlinear regression prediction model, and combine the historical operating condition state vector and the current real-time operating condition state vector to predict the nonlinear temperature evolution trajectory of the target thick material region and the adjacent thin material region in the future control cycle.

[0042] This step aims to address the technical challenge of traditional Gaussian process regression and other algorithms, which, when predicting temperature distribution in future time windows, rely solely on absolute scalar distances in their covariance kernel functions. This leads to model divergence and distortion in the early stages of phase transitions. This step embeds the obtained dynamic thermal coupling characteristics as strong physical prior constraints into the kernel function of the time series, constructing a modified time-dependent kernel function.

[0043] The formula for calculating the time correlation kernel function after incorporating the dynamic thermal coupling feature correction is as follows:

[0044] In the formula, Indicates at a time node and time nodes The covariance kernel function values ​​between them; This represents the signal variance parameter; This represents the time scale parameter, and its value is greater than 0. and These represent two different historical or current sampling comparison moments. and These represent the time nodes. and time nodes The calculated dynamic thermal coupling characteristics.

[0045] The first half of the formula retains the fundamental characteristic of traditional algorithms that a smaller time difference indicates a stronger correlation, while the second half incorporates thermodynamic physical state constraints. If the absolute value of the difference in dynamic thermal coupling characteristics between two comparison moments is large, it indicates that the mold system is under drastically different heat conduction boundary modes, such as before a phase change and during a violent phase change releasing latent heat. In this case, the constraint term in the denominator of the calculation formula will increase significantly, forcing a substantial physical attenuation of the correlation between the covariance data at these two moments. Since the absolute value of the difference in dynamic thermal coupling characteristics must be greater than or equal to 0, adding a constant 1 to the denominator ensures that the overall denominator value is strictly greater than or equal to 1, avoiding the risk of logical operation collapse caused by a denominator of zero.

[0046] In practical implementation, edge computing nodes utilize the improved time-dependent kernel function to train the nonlinear regression prediction model online. The system continuously inputs real-time operating condition vectors from the current and historical time windows, thereby stably outputting the nonlinear temperature evolution trajectory of the target thick material region and adjacent thin material regions over multiple future control cycles. By deeply internalizing the dynamic thermal coupling physical characteristics into the lowest-level kernel function of the data-driven prediction model, this temperature evolution trajectory incorporates prior physical information, enabling the algorithm to adaptively and accurately predict the impending latent heat hysteresis effect of phase change, eliminating the prediction blind spots and distortions of traditional pure data prediction models under extreme phase change disturbance conditions.

[0047] S4. Based on the nonlinear temperature evolution trajectory and the phase transition activity characteristics, a thermal stress penalty constraint for spatial temperature differences is added during the cost evaluation process of control optimization, and the cooling water circuit control command that minimizes the global cost is found and output.

[0048] Excessive spatial temperature gradients are the root cause of deformation and cracking in molded parts, especially during the cooling stage when the material undergoes a martensitic transformation and becomes extremely brittle, at which point the material's tolerance to spatial temperature gradients drops to its lowest point. Therefore, this step adds a cross-regional thermal stress penalty constraint term to the conventional temperature tracking error, constructing a penalty index that accurately reflects the actual physical cracking risk.

[0049] In this step, the calculation formula for the cross-regional thermal stress penalty constraint term is as follows:

[0050] In the formula, This represents the cross-regional thermal stress penalty term obtained through calculation and accumulation; This represents the total prediction step size of the model predictive control algorithm, and its value is a positive integer. This is the penalty weight constant. and The target thick material area and the adjacent thin material area are respectively the future... The predicted temperature for each prediction step is obtained from the nonlinear temperature evolution trajectory output by the previous step. The fixed physical Euclidean distance between two measuring points in the target thick material area and the adjacent thin material area is obtained by measuring the three-dimensional digital model of the mold and its value is greater than 0. It is a reference safe temperature gradient constant based on material properties and allowed in engineering, a fixed value determined by previous process testing and greater than 0. and The phase transition activity characteristics of the target thick material region and the adjacent thin material region at the corresponding prediction time.

[0051] The calculation process of this formula is as follows: Calculate the absolute value of the predicted temperature difference between the target thick material region and the adjacent thin material region at each prediction step; compare this absolute value with the product of the physical Euclidean distance between the target thick material region and the adjacent thin material region and the reference safe temperature gradient; square this ratio and multiply it by the sum of the phase transformation activity characteristics of the target thick material region and the adjacent thin material region, plus the penalty weight constant; sum the calculation results for all prediction steps to obtain the cross-regional thermal stress penalty term, which is used as the thermal stress penalty constraint for spatial temperature differences. This formula combines the spatial temperature gradient term with the regional phase transformation activity characteristics through multiplicative logic. If it is predicted that any region is in a period of intense phase transformation (i.e., when the activity value is large), and the material is in a highly brittle state, the penalty term will be further amplified by the activity parameter of the corresponding region, thereby extremely sensitively intercepting unreasonable temperature differences that induce microcracks.

[0052] In the optimization and execution phase of the cooling water circuit control commands, the control system adds the traditional tracking error cost function to the aforementioned cross-zone thermal stress penalty term to form a global cost function. Using a quadratic programming algorithm, under given constraints on maximum valve opening and flow rate limit, the system finds and generates the valve opening control sequence that minimizes the global cost function; the cooling water circuit control commands include this control sequence. The system then extracts the first control command of the valve opening control sequence and converts it into an electrical pulse signal, outputting it to the proportional-integral regulating water valve of the corresponding mold zone to execute precise dynamic allocation of the cooling water circuit. Subsequently, the control system enters the next prediction and control cycle, achieving long-term stable closed-loop physical regulation.

[0053] The following combination Figure 2 and Figure 3 The effects of the present invention will be further described.

[0054] like Figure 2 As shown in the figure, the holding time is plotted on the horizontal axis, and the subsurface temperature of the mold is plotted on the vertical axis, reproducing the temperature decay trajectory of the core region of the thick material during the entire cycle of mold closure, holding, and cooling. The two core curves in the figure are naturally and smoothly connected at both ends, and both are realistically superimposed with micro-fluctuation data that conform to the characteristics of electromechanical composite interference in industrial settings. The cooling curves of existing technologies exhibit a significant lag in temperature recovery and a plateau phenomenon when the temperature drops through the pre-calibrated sensitive range of martensitic phase transformation, due to the concentrated release of a large amount of latent heat, resulting in the cooling water not carrying away heat in time. The stable and rapid cooling curve of this invention, while effectively overcoming the interference of micro-fluctuations in the environment, relies on the predictive model's forward-looking identification of nonlinear thermal resistance and the advance injection of water valve flow rate, still maintaining a stable and steeply sloping rapid cooling trend without stagnation, and spanning the entire phase transformation range.

[0055] like Figure 3As shown, this graph also uses holding time as the horizontal axis and the temperature difference at the interface between measuring points at a specific Euclidean distance as the vertical axis, reflecting the deterioration of thermodynamic cross-conduction at the interface during rapid cooling. A horizontal safety threshold reference line representing the risk of cracking when the residual tensile stress inside the material exceeds the yield strength is marked in the middle of the graph. Existing technology curves reflect that under existing independent proportional-integral-differential zone control technology, the latent heat of phase change in the middle section causes the thick material area to passively execute a fully open cooling valve, resulting in rapid penetration of cold energy into the thin material area. This leads to severe nonlinear oscillations in the temperature difference at the interface, with its peak significantly exceeding the cracking risk threshold. The curves of this invention are generally below the cracking risk threshold, confirming that the collaborative optimization control algorithm of this invention successfully suppresses the aforementioned surge in temperature difference at the interface caused by thermal coupling.

[0056] A specific embodiment of the temperature zone control system for automotive double-door ring hot pressing molds provided by the present invention: The temperature zone control system for automotive double-door ring hot pressing mold includes a processor and a memory. The memory stores computer program instructions. When the computer program instructions are executed by the processor, the temperature zone control method for automotive double-door ring hot pressing mold in the above embodiments is implemented.

[0057] The temperature zone control system of the automotive double door ring hot pressing mold also includes other components well known to those skilled in the art, such as communication buses and communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

[0058] While various embodiments of the invention have been shown and described in this specification, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention.

Claims

1. A method for temperature zone control of a hot pressing mold for automotive double door rings, characterized in that, include: Simultaneously acquire multimodal working condition data of hot pressing mold in the pressure holding stage and perform smoothing, noise reduction and numerical mapping processing to obtain the real-time working condition state vector at the current moment. Based on the temperature and pressure states reflected by the real-time operating condition state vector, and combined with the thermodynamic change law of the material, the phase transformation activity characteristics used to characterize the degree of martensitic phase transformation of the material and the dynamic thermal coupling characteristics used to measure the combined influence of physical interface thermal resistance and latent heat of phase transformation are calculated. The dynamic thermal coupling feature is introduced into the time correlation kernel function of the nonlinear regression prediction model. By combining the historical working condition state vector and the current real-time working condition state vector, the nonlinear temperature evolution trajectory of the target thick material area and the adjacent thin material area in the future control cycle is predicted. Based on the nonlinear temperature evolution trajectory and the phase transition activity characteristics, a thermal stress penalty constraint for spatial temperature differences is added during the cost evaluation process of control optimization to find and output the cooling water circuit control command that minimizes the global cost.

2. The temperature zone control method for automotive double-door ring hot pressing mold according to claim 1, characterized in that, The formula for calculating the phase transition activity characteristic is as follows: In the formula, Indicates at time node The phase transition activity characteristics, Indicates at time node Real-time temperature of the secondary surface of the mold. This indicates the temperature at which the material's inherent martensitic phase transformation begins. This represents the pre-set phase transition temperature window constant.

3. The temperature zone control method for automotive double-door ring hot pressing mold according to claim 2, characterized in that, The formula for calculating the dynamic thermal coupling characteristic is as follows: In the formula, Indicates at time node The dynamic thermal coupling characteristics, Indicates at time node The real-time contact pressure obtained Indicates the standard rated holding pressure constant. and These represent the first weighting coefficient and the second weighting coefficient, respectively.

4. The temperature zone control method for automotive double-door ring hot pressing mold according to claim 3, characterized in that, The formula for calculating the time correlation kernel function is as follows: In the formula, Indicates at time node and time nodes The covariance kernel function values ​​between them Represents the signal variance parameter. Indicates the time scale parameter. and Represented at the time node and the time nodes The calculated dynamic thermal coupling characteristics.

5. The temperature zone control method for automotive double-door ring hot pressing mold according to claim 1, characterized in that, The addition of thermal stress penalty constraints for space temperature differences during the cost evaluation process of control optimization includes: Calculate the absolute value of the predicted temperature difference between the target thick material region and the adjacent thin material region at each prediction step size; The absolute value is compared to the product of the physical Euclidean distance between the target thick material region and the adjacent thin material region and the reference safe temperature gradient; After calculating the square of the ratio, multiply it by the sum of the phase transition activity characteristics of the target thick material region and the adjacent thin material region, as well as the penalty weight constant; The calculation results of all predicted step sizes are summed to obtain the cross-regional thermal stress penalty term, which is used as the thermal stress penalty constraint for the spatial temperature difference.

6. The temperature zone control method for automotive double-door ring hot pressing mold according to claim 1, characterized in that, The multimodal operating condition data includes the real-time temperature of the mold subsurface, local contact pressure, coolant inlet temperature, coolant outlet temperature, and the real-time flow rate corresponding to the current valve opening.

7. The temperature zone control method for automotive double-door ring hot pressing mold according to claim 6, characterized in that, In the process of smoothing, denoising and numerical mapping of multimodal operating condition data, electromagnetic interference noise is eliminated by moving average filtering algorithm, and flow and pressure data are linearly normalized to construct the real-time operating condition state vector at the current moment.

8. The temperature zone control method for automotive double-door ring hot pressing mold according to claim 1, characterized in that, The cooling water circuit control command includes a valve opening control sequence generated by a quadratic programming algorithm under given valve maximum opening constraints and flow upper limit constraints.

9. The temperature zone control method for automotive double-door ring hot pressing mold according to claim 8, characterized in that, The first control command of the valve opening control sequence is extracted, converted into an electrical pulse signal, and output to the proportional-integral regulating water valve of the corresponding zone to perform dynamic allocation of cooling water circuit.

10. A temperature zone control system for a hot pressing mold for automotive double door rings, characterized in that, The device includes a processor and a memory storing computer instructions, wherein the processor, when executing the computer instructions, implements the temperature zoning control method for automotive double-door ring hot pressing molds as described in any one of claims 1 to 9.

Images