Stamping lubrication self-adaptive control method and system for hardware stamping part

Abstract

The invention relates to the technical field of information, in particular to a stamping lubrication self-adaptive control method and system for a hardware stamping part in the field of hardware stamping part manufacturing, and the method comprises the steps: comparing oil film thickness measured data after execution of a lubrication system with a predicted value, if an error exceeds a preset range, adjusting a weight parameter of a regression model, and if the error exceeds the preset range, adjusting the weight parameter of the regression model; the prediction precision is optimized; in the continuous die production process, according to material characteristic changes of different stations, lubricating parameter configuration is automatically switched, and it is ensured that the lubricating effect of each station meets the requirement; through data linkage of a lubricating system and a mold temperature sensor, a dynamic adjustment mechanism of lubricating parameters and mold temperature is established, and real-time optimization of the lubricating effect is achieved; based on lubricant consumption data in the production process, in combination with the variation trend of the oil film thickness and the mold temperature, the use efficiency of the lubricant is analyzed, and the lubricant supply strategy is optimized.

Description

Technical Field

[0001] The present invention relates to the field of information technology, specifically to the field of manufacturing hardware stamping parts, and particularly to a stamping lubrication adaptive control method and system for hardware stamping parts. Background Art

[0002] During the production process of multi-station progressive dies for hardware stamping parts, the performance of the lubrication system directly affects the die life and workpiece quality. The traditional manual spraying lubrication method has the problem of uneven oil film thickness, and the oil film thickness ranges from 10 to 50 microns. This non-uniformity leads to increased local wear on the die surface, and the die life is usually less than 500,000 times.

[0003] In addition, it is difficult to precisely control the oil film thickness by manual spraying, which easily causes oil residue on the workpiece surface and affects subsequent surface treatment processes. Although the existing automatic lubrication systems have solved the deficiencies of manual spraying to a certain extent, there is still a key problem: they cannot adjust the lubrication parameters in real time according to the friction coefficient of the stamping material. For example, the friction coefficients of stainless steel and copper alloy are quite different, but the existing systems cannot precisely control the oil film for different material characteristics, resulting in unsatisfactory lubrication effects. This non-adaptability is particularly prominent in the production process of multi-station progressive dies because different workstations may need to process workpieces of different materials, and the system cannot quickly respond to this change. There is also a close relationship between the oil film thickness and the die temperature, but the existing systems lack real-time monitoring and feedback mechanisms for the die temperature and cannot dynamically adjust the lubrication parameters according to the temperature change.

[0004] The above problems together lead to consequences such as shortened die life, unstable workpiece quality, and excessive lubricating oil consumption, restricting the improvement of the production efficiency and quality of hardware stamping parts. Summary of the Invention

[0005] The present invention provides a stamping lubrication adaptive control method for hardware stamping parts, including the following steps:

[0006] Step S101: According to the material information of the stamping material, obtain the friction coefficient of this material from the pre-established database, and combine the real-time monitoring data of the die temperature sensor to determine the optimal oil film thickness range under the current working conditions;

[0007] Step S102: Adopt a regression model in machine learning algorithms, train an oil film thickness prediction model based on the historical data of the friction coefficient and the die temperature, and predict the target oil film thickness value by inputting the current friction coefficient and the die temperature;

[0008] Step S103: For the predicted oil film thickness value, combine the control parameter range of the lubrication system, calculate the injection amount and injection frequency of the lubricant, generate a lubrication parameter adjustment instruction and send it to the execution module of the lubrication system;

[0009] Step S104: During the execution of the lubrication system, by monitoring the changing trend of the die temperature in real time, determine whether the temperature exceeds the preset threshold. If it exceeds the threshold, recalculate the lubrication parameters and update the execution instruction;

[0010] Step S105: Compare the measured data of the oil film thickness after the execution of the lubrication system with the predicted value. If the error exceeds the preset range, adjust the weight parameters of the regression model to optimize the prediction accuracy;

[0011] Step S106: During the production process of the progressive die, automatically switch the lubrication parameter configuration according to the change of material characteristics at different stations to ensure that the lubrication effect at each station meets the requirements;

[0012] Step S107: Through the data linkage between the lubrication system and the die temperature sensor, establish a dynamic adjustment mechanism for lubrication parameters and die temperature to achieve real-time optimization of the lubrication effect;

[0013] Step S108: Based on the lubricant consumption data during the production process, combined with the changing trends of the oil film thickness and die temperature, analyze the usage efficiency of the lubricant and optimize the lubricant supply strategy.

[0014] The present invention provides a stamping lubrication adaptive control system for hardware stamping parts, mainly including:

[0015] A friction coefficient acquisition module, which is used to obtain the friction coefficient of the material from the pre-established database according to the material information of the stamping material, and combine the real-time monitoring data of the die temperature sensor to determine the optimal oil film thickness range under the current working conditions;

[0016] An oil film thickness prediction module, which is used to train an oil film thickness prediction model based on the historical data of the friction coefficient and die temperature by using a regression model in machine learning algorithms, and predict the target oil film thickness value by inputting the current friction coefficient and die temperature;

[0017] A lubrication parameter calculation module, which is used to calculate the injection amount and injection frequency of the lubricant for the predicted oil film thickness value, combine the control parameter range of the lubrication system, generate a lubrication parameter adjustment instruction and send it to the lubrication system execution module;

[0018] A temperature monitoring module, which is used to monitor the changing trend of the die temperature in real time during the execution of the lubrication system, determine whether the temperature exceeds the preset threshold. If it exceeds the threshold, recalculate the lubrication parameters and update the execution instruction;

[0019] A model optimization module, which is used to compare the measured data of the oil film thickness after the execution of the lubrication system with the predicted value. If the error exceeds the preset range, adjust the weight parameters of the regression model to optimize the prediction accuracy;

[0020] A lubrication parameter switching module, which is used to automatically switch the lubrication parameter configuration according to the material property changes of different stations during the production process of progressive dies, so as to ensure that the lubrication effect of each station meets the requirements;

[0021] A dynamic adjustment module, which is used to establish a dynamic adjustment mechanism between lubrication parameters and die temperature through data linkage between the lubrication system and die temperature sensors, so as to realize real-time optimization of the lubrication effect;

[0022] A lubricant optimization module, which is used to analyze the usage efficiency of lubricants based on the lubricant consumption data during the production process, and combine the change trends of oil film thickness and die temperature to optimize the lubricant supply strategy.

[0023] The technical solution provided by the embodiment of the present invention may include the following beneficial effects:

[0024] The present invention discloses a stamping lubrication adaptive control method for metal stamping parts. This method predicts the optimal oil film thickness by establishing a material friction coefficient database and combining the real-time monitoring data of die temperature sensors. A regression model is used to train the oil film thickness prediction model, and the lubrication parameters are calculated and executed according to the current working conditions. During the production process, the present invention monitors the change of die temperature in real time, dynamically adjusts the lubrication parameters, and optimizes the prediction model according to the measured data. For the material property changes of different stations, the lubrication configuration is automatically switched to ensure the lubrication effect of each station. By establishing a dynamic adjustment mechanism between lubrication parameters and die temperature and combining the analysis of lubricant consumption data, the continuous optimization of the lubrication effect and the improvement of the lubricant usage efficiency are realized, effectively improving the stability and efficiency of stamping production. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a flowchart of a stamping lubrication adaptive control method for metal stamping parts of the present invention.

[0026] Figure 2 It is a schematic diagram of a stamping lubrication adaptive control method and system for metal stamping parts of the present invention.

[0027] Figure 3 It is another schematic diagram of a stamping lubrication adaptive control method and system for metal stamping parts of the present invention.

[0028] Figure 4 It is a framework schematic diagram of a stamping lubrication adaptive control system for metal stamping parts of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0029] Next, the technical solutions in the embodiments of the present invention will be clearly and detailedly described in conjunction with the accompanying drawings in the embodiments of the present invention. The described embodiments are only a part of the embodiments of the present invention.

[0030] As Figures 1-4 , a stamping lubrication adaptive control method for a hardware stamping part in this embodiment specifically includes:

[0031] Step S101, according to the material information of the stamping material, obtain the friction coefficient of the material from a pre-established database, and combine the real-time monitoring data of the mold temperature sensor to determine the optimal oil film thickness range under the current working conditions.

[0032] Obtain the initial friction parameters of the stamping material, where the initial friction parameters are obtained from a pre-established material information database; obtain the current temperature data of the mold, where the current temperature data is obtained by real-time monitoring of the temperature sensor; determine whether there is a preset matching relationship between the initial friction parameters and the current temperature data, and if so, obtain the adjusted friction coefficient from a preset mapping table; according to the adjusted friction coefficient, obtain an oil film thickness range table associated with the current working conditions; according to the oil film thickness range table and the current temperature data, determine the upper and lower limit thresholds of the optimal oil film thickness interval; obtain historical working condition data and friction coefficient data, and use a machine learning algorithm to analyze the historical working condition data and friction coefficient data to obtain the dynamic adjustment value of the oil film thickness range; according to the matching degree between the dynamic adjustment value and the current working conditions, determine the final optimal oil film thickness range.

[0033] Exemplarily, obtain the friction coefficient of the stamping material from a pre-established database through the material information, and the acquisition of the initial friction parameters is a basic link in the stamping process.

[0034] For example, for common stamping materials such as cold-rolled steel sheet SPCC, its initial friction coefficient can be found from the database to be 0.15, and this value is measured under standard conditions.

[0035] Exemplarily, if the material is changed to aluminum alloy AA5052, the database may show that the initial friction coefficient is 0.12. Different materials have different friction coefficients due to differences in surface characteristics. The advantage of this method is to quickly establish a parameter benchmark and avoid the cost of repeated tests. Using a temperature sensor to monitor the mold temperature in real time and obtain the current temperature data is the key to dynamic adjustment.

[0036] Specifically, during the stamping process, the mold temperature may rise from the initial 25°C to 60°C, and the temperature sensor can collect data once per second to ensure real-time performance.

[0037] It can be understood that temperature changes will affect the lubricant performance and thus indirectly affect the friction coefficient.

[0038] Preferably, a high-precision thermocouple sensor is used to control the error within ±1°C and improve data reliability. The technical effect brought by this real-time monitoring is significant and can provide an accurate basis for subsequent parameter adjustments. If the initial friction parameter matches the temperature data, the adjusted friction coefficient is determined through a preset mapping table.

[0039] In one possible implementation, when the mold temperature rises to 60°C, the mapping table shows that the friction coefficient of the SPCC steel plate is adjusted from 0.15 to 0.17, because the increase in temperature reduces the viscosity of the lubricant. This mapping table is usually constructed based on experimental data and can reflect the correlation between temperature and friction.

[0040] It should be noted that this method improves the scientific nature of parameter adjustment and avoids the arbitrariness of empirical estimation. According to the correlation between the adjusted friction coefficient and the current working conditions, the corresponding oil film thickness range table is obtained.

[0041] For example, when the friction coefficient is 0.17, the oil film thickness range table may recommend a thickness between 2-5 microns. This range is based on lubrication theory and ensures that friction can be reduced without affecting stamping accuracy due to excessive oil film thickness.

[0042] In one embodiment, if the working condition is high-speed stamping, the range may be biased towards 2-3 microns to accommodate higher sliding speeds. This correlation analysis helps optimize the lubrication strategy. The upper and lower limits of the optimal oil film thickness range are determined by combining the thickness range table with the real-time monitored temperature data.

[0043] Exemplarily, when the temperature is 60° C., the upper and lower limits may be determined as 2.5-4 μm in combination with the range table.

[0044] Preferably, if the temperature continues to rise to 80°C, the upper limit of the oil film thickness may be adjusted to 3.5 microns to avoid lubrication failure. The advantage of this combined analysis is that it takes into account both theoretical and actual working conditions and improves applicability. A machine learning algorithm is used to analyze historical working condition data and friction coefficient to determine the dynamic adjustment value of the thickness range.

[0045] In one embodiment, by analyzing the past 1,000 stamping data, the machine learning model found that when the temperature exceeds 70°C, the oil film thickness needs to increase by 0.5 microns to maintain stability.

[0046] Specifically, the model may be based on a random forest algorithm, with inputs including temperature, pressure and friction coefficient, and outputs an adjustment value. This method can mine the implicit rules in the data and make the thickness range closer to actual needs. The final optimal oil film thickness range is obtained based on the degree of match between the dynamic adjustment value and the current working conditions.

[0047] For example, under the current working conditions, the temperature is 60°C and the adjustment value is 0.3 microns, then the final range may be 2.8 - 4.3 microns.

[0048] It can be understood that this dynamic matching ensures that the oil film thickness can not only adapt to real-time changes but also maintain process stability.

[0049] In one embodiment, if the working condition pressure increases, the range may be further narrowed to 3 - 4 microns to balance lubrication and precision. The advantage of this method is that it improves the consistency of stamping quality and reduces die wear and scrap rate.

[0050] Step S102, using a regression model in machine learning algorithms, based on historical data of friction coefficient and die temperature, train to obtain an oil film thickness prediction model, and predict the target oil film thickness value by inputting the current friction coefficient and die temperature.

[0051] Obtain historical data, where the historical data includes records of friction coefficient and die temperature; use a linear regression algorithm to train the historical data to obtain initial model parameters; obtain current values, where the current values include real-time data of friction coefficient and die temperature; input the real-time data into the initial model parameters to obtain a preliminary oil film thickness prediction value; determine whether the preliminary oil film thickness prediction value exceeds a preset threshold, if so, use a support vector regression algorithm to adjust the initial model parameters to obtain optimized model parameters; input the real-time data into the optimized model parameters to obtain an adjusted oil film thickness value; obtain the deviation between the adjusted oil film thickness value and the historical data, determine the stability of the prediction model to obtain a deviation analysis result; update the training process according to the deviation analysis result, use a linear regression algorithm to retrain the historical data and the real-time data to obtain updated model parameters; input the real-time data into the updated model parameters to obtain the final oil film thickness prediction value.

[0052] Exemplarily, obtaining records of friction coefficient and die temperature through historical data is the basis for constructing the prediction model.

[0053] Exemplarily, data can be extracted from the stamping production records of the past year. Assume that the friction coefficient range of a certain material is between 0.12 and 0.15, and the die temperature is recorded between 50°C and 80°C. These data provide a reliable basis for subsequent analysis.

[0054] In a possible implementation, when training a prediction model using a linear regression algorithm, the friction coefficient and the die temperature can be used as input variables, and the oil film thickness can be used as the output variable. The initial model parameters may show that when the friction coefficient is 0.13 and the die temperature is 60°C, the predicted value of the oil film thickness is approximately 0.05 mm. This method captures the linear relationship between variables through historical trends.

[0055] It should be noted that when extracting real-time data from the current value, assume that in a certain production, the friction coefficient is 0.14 and the die temperature is 65°C. After inputting into the initial model, the preliminary predicted value of the oil film thickness may be 0.06 mm. If this value exceeds the preset threshold (such as 0.04 mm to 0.055 mm), it indicates that the model needs to be further optimized.

[0056] Specifically, when adjusting the prediction model using the support vector regression algorithm, complex working conditions can be processed by introducing non-linear features.

[0057] For example, at a high die temperature, the influence of the friction coefficient on the oil film thickness may increase. After optimization, the model may be adjusted to: when the friction coefficient is 0.14 and the temperature is 65°C, the predicted value of the oil film thickness becomes 0.053 mm, which is closer to the actual demand.

[0058] In one embodiment, after inputting real-time data with optimized model parameters, the adjusted value of the oil film thickness may be 0.052 mm. At this time, through the deviation analysis of historical data, it can be found that the deviation between the predicted value and the actual value is only 0.002 mm, indicating that the model has high stability. This analysis ensures the reliability of the prediction.

[0059] Preferably, the results of the deviation analysis can be used to update the training process.

[0060] For example, if it is found that the deviation increases under high-temperature working conditions, the weight of relevant data can be increased, and the model can be retrained using linear regression. The updated model parameters may show that when the friction coefficient is 0.14 and the temperature is 70°C, the predicted value of the oil film thickness is adjusted to 0.051 mm. This dynamic update improves the adaptability of the model.

[0061] In one embodiment, the final predicted value of the oil film thickness can be obtained by combining real-time data.

[0062] For example, the current friction coefficient is 0.13 and the die temperature is 68°C. After inputting into the updated model, the predicted value is 0.05 mm. This result can directly guide the adjustment of the lubricating oil amount, avoiding material waste or stamping defects.

[0063] It can be understood that the core of this method lies in optimizing the prediction through data driving.

[0064] Exemplarily, if the temperature suddenly rises to 85°C in the historical data, the friction coefficient becomes 0.15, and the oil film thickness needs to increase to 0.06 mm, the new model can quickly adapt to this change. This flexibility helps to improve production efficiency.

[0065] For example, in actual production, if the predicted value is too low, it may lead to increased mold wear; while if the predicted value is too high, it may waste lubricating oil. Through deviation analysis and model update, the two can be effectively balanced, extending the mold life and reducing costs. The application value of this method is particularly obvious under dynamic working conditions.

[0066] Step S103, for the predicted oil film thickness value, in combination with the control parameter range of the lubrication system, calculate the injection volume and injection frequency of the lubricant, generate a lubrication parameter adjustment instruction and send it to the lubrication system execution module.

[0067] Obtain the predicted value of the oil film thickness representing the lubrication state. According to the predicted value of the oil film thickness, in combination with the preset lubrication parameter range, use the linear regression algorithm to calculate the corresponding injection volume and injection frequency to obtain the preliminary lubrication parameters; obtain the constraint conditions corresponding to the preliminary lubrication parameters, and determine whether the injection volume exceeds the preset range. If so, reduce the injection volume to the preset range according to the preset ratio to obtain the adjusted lubrication parameters; according to the adjusted lubrication parameters, in combination with the change trend of the oil film thickness, generate a lubrication parameter adjustment instruction; send the lubrication parameter adjustment instruction to the lubrication execution module, obtain the reception status of the lubrication system for the adjustment instruction, and determine whether the instruction transmission delay exceeds the preset threshold. If so, regenerate and send the lubrication parameter adjustment instruction until the final execution status is determined; obtain the oil film thickness change value feedback by the lubrication system, determine the consistency between the oil film thickness change value and the predicted value, and optimize the lubrication control parameters based on the consistency analysis result to obtain the optimized lubrication control parameters.

[0068] Exemplarily, after obtaining the oil film thickness data through the predicted value and analyzing it in combination with the preset parameter range is an important basis for ensuring the rationality of subsequent calculations.

[0069] For example, in the mold processing scenario, assume that the predicted oil film thickness is 0.05 mm, and the preset range is 0.03 - 0.07 mm. The data falls within the range, indicating that the initial calculation basis is reliable.

[0070] It can be understood that this process depends on the accuracy of historical data. If the predicted value deviates from the range, it may indicate that data collection or model training needs to be adjusted.

[0071] In a possible implementation manner, according to the oil film thickness data and the parameter range, use the linear regression algorithm to calculate the injection volume and injection frequency.

[0072] For example, when the oil film thickness is 0.05 mm, based on historical trends, linear regression yields an injection rate of 2 ml / s and an injection frequency of 3 times per minute, forming preliminary lubrication parameters.

[0073] It should be noted that here, through a simple mapping relationship, the thickness is associated with the lubrication requirements to avoid overly complex calculations. When obtaining the control parameter constraint conditions for the preliminary lubrication parameters.

[0074] Specifically, if the upper limit of the injection rate is 1.8 ml / s and the calculated value is 2 ml / s, it is scaled down to 1.8 ml / s proportionally.

[0075] Exemplarily, this adjustment can be achieved through a proportionality coefficient of 0.9 to ensure that the parameters meet the equipment capabilities. The advantage of this method is that it not only meets the constraints but also gets as close as possible to the original requirements. When generating the lubrication parameter adjustment instruction based on the adjusted injection rate and injection frequency.

[0076] Preferably, the changing trend of the oil film thickness can be combined.

[0077] For example, if the thickness slowly drops from 0.05 mm to 0.04 mm, the instruction may increase the injection frequency to 4 times per minute. This dynamic adjustment can better adapt to the changes in the actual working conditions.

[0078] In one embodiment, the adjustment instruction is sent to the execution module. After the lubrication system receives the data, if the transmission delay is 0.8 s, exceeding the preset threshold of 0.5 s, the instruction is regenerated and sent again. This retry mechanism ensures that the instruction takes effect in a timely manner and avoids insufficient lubrication caused by delays. After obtaining the change in the oil film thickness feedback from the lubrication system, judging the consistency with the predicted value is the key to optimization.

[0079] For example, if the feedback thickness stabilizes at 0.05 mm, which is consistent with the predicted value, it indicates that the control parameters are effective. If a deviation occurs, such as the feedback value being 0.06 mm, it may indicate that the injection rate needs to be fine-tuned. This consistency analysis helps to continuously improve the parameters.

[0080] Specifically, the injection rate adjustment is considered from multiple aspects.

[0081] For example, the equipment capabilities limit the upper injection rate to 1.8 ml / s. An increase in the ambient temperature may require more lubrication, and the downward trend of the thickness also requires a quick response. After comprehensive consideration, it is adjusted to 1.7 ml /

[0082] s, which not only meets the constraints but also responds to the changes. In another embodiment, if the frequency adjustment is the dominant factor, the frequency can be increased to 5 times per minute while keeping the injection rate at 1.5 ml / s, also achieving the effect of a stable thickness. These solutions support each other to jointly ensure the lubrication effect.

[0083] It is understandable that the advantage of these methods lies in forming a closed-loop control through real-time data and feedback, keeping the oil film thickness within the target range, and improving the die life and processing quality.

[0084] Step S104, during the execution of the lubrication system, by monitoring the changing trend of the die temperature in real time, determine whether the temperature exceeds a preset threshold. If it exceeds the threshold, recalculate the lubrication parameters and update the execution instruction.

[0085] Obtain die temperature data and determine the temperature fluctuation pattern; for the temperature fluctuation pattern, obtain the current lubrication parameters, input them into a preset regression model, and calculate the lubrication parameter adjustment value; update the lubrication parameters according to the adjustment value, generate a new execution instruction; use the new execution instruction to drive the lubrication system to operate and obtain real-time feedback data; extract the operation status indicators from the feedback data and determine whether the system is stable; if the system is not stable, adjust the monitoring frequency according to the feedback data and return to the step of obtaining die temperature data; if the system is stable, maintain the current lubrication parameter setting and execute the step of obtaining die temperature data at a preset frequency.

[0086] Exemplarily, obtaining die temperature data through sensors and recording the changing trend is a common industrial monitoring method.

[0087] Exemplarily, during the operation of a stamping die, sensors can be installed at the key heat-receiving points of the die, such as the contact area between the punch and the workpiece, to collect temperature data in real time and record it in minutes.

[0088] For example, the initial temperature is 50 degrees, and after 10 minutes of operation, it rises to 75 degrees, showing a slow upward trend. When using time series analysis to determine the temperature fluctuation pattern, the changing rule of temperature over time can be analyzed with the help of historical data.

[0089] Specifically, assuming that the records within a week show that the temperature fluctuates periodically between 70 - 80 degrees, then the pattern of "rising periodically and then falling" can be extracted as a feature. Extracting the feature value from the changing trend and comparing it with the preset threshold is the key to judging the system state.

[0090] For example, setting the temperature threshold to 85 degrees, when the feature value shows that the temperature peak reaches 87 degrees during a certain operation, the temperature over-standard state is triggered.

[0091] In a possible implementation, the feature value is not limited to the peak value, but can also include the rising rate. If the temperature rises from 60 degrees to 80 degrees within 5 minutes with too fast a rate, it can also be regarded as an over-standard signal. This multi-dimensional feature judgment can more comprehensively reflect the die state. For the temperature over-standard state, obtaining the current lubrication parameters and inputting them into the regression model to calculate the adjustment value is the core link for optimizing lubrication.

[0092] Preferably, the current parameters may be an injection volume of 5 ml per time and a frequency of 1 time per minute. By analyzing the relationship between temperature and lubrication amount through a regression model, an adjustment value of increasing the injection volume to 6 ml per time is obtained.

[0093] It can be understood that this adjustment is based on historical data training to ensure matching with the temperature change trend, thereby effectively reducing frictional heat. After updating the lubrication parameters with the adjustment value and generating a new execution instruction, the lubrication system operates accordingly.

[0094] For example, the adjusted instruction may be "injection volume 6 ml per time, frequency 1 time per minute", and is sent to the injection device through the control unit.

[0095] In one embodiment, the feedback data shows that the temperature drops from 87 degrees to 82 degrees, proving that the adjustment is effective. Obtaining real-time feedback data usually relies on sensors to synchronously monitor the mold temperature and lubrication effect. Extracting the operating state indicators from the feedback data and judging whether the system is stable is an important step in the process closed-loop.

[0096] Specifically, the indicators may include the temperature fluctuation range and the lubricant coverage uniformity. If the temperature fluctuation is controlled within ±2 degrees and the lubrication coverage rate reaches more than 90%, it is considered stable.

[0097] It should be noted that if the feedback shows that the temperature still fluctuates between 83 - 87 degrees, it indicates that the system is not in a stable state. At this time, the monitoring frequency can be increased, such as from 1 time per minute to 1 time per 30 seconds, to obtain finer-grained data and repeat the adjustment.

[0098] In one embodiment, if the temperature drop trend is not obvious after the initial adjustment, the reasons can be analyzed in combination with the feedback data.

[0099] For example, it is found that the injection position of the lubricant is offset, resulting in insufficient coverage, then the nozzle angle is adjusted and a new instruction is generated. This dynamic optimization can significantly improve the system adaptability.

[0100] Preferably, through repeated iteration, the mold temperature is finally stabilized below 80 degrees, which not only prolongs the mold life but also reduces the workpiece quality problems caused by overheating.

[0101] For example, the surface roughness drops from 0.8 microns to 0.6 microns, reflecting the actual benefits of process improvement.

[0102] Step S105, compare the measured data of the oil film thickness after the lubrication system executes with the predicted value. If the error exceeds the preset range, adjust the weight parameters of the regression model to optimize the prediction accuracy.

[0103] Obtain the measured data of the oil film thickness generated after the lubrication system is executed. Calculate the difference between the measured data and the predicted value output by the regression model. Extract the error range from the difference, and determine whether the error range exceeds a preset threshold. If it exceeds, determine the set of parameters that need to be adjusted. For the set of parameters, use the gradient descent algorithm to adjust the weight parameters in the regression model to obtain updated model parameters. According to the updated model parameters, recalculate the predicted value of the oil film thickness to obtain a new prediction result. After obtaining the new prediction result, compare it with the measured data again to determine whether the error range is reduced to within the preset threshold. If it still exceeds the preset threshold, repeat the process of adjusting the weight parameters until a prediction result that meets the preset accuracy condition is obtained. According to the regression model that meets the preset accuracy condition, output the optimized predicted value of the oil film thickness to complete the accuracy optimization of the system execution.

[0104] In a possible implementation, when obtaining the measured data of the oil film thickness generated after the lubrication system is executed, the oil film distribution on the mold surface can be collected in real time through a high-precision sensor.

[0105] For example, in a certain operation, the measured oil film thickness is 20 microns, while the predicted value of the regression model is 22 microns, and the difference between the two is 2 microns. The extraction of this difference can be achieved by collecting data multiple times and taking the average to ensure the stability of the result.

[0106] It should be noted that the judgment of the error range requires a preset threshold.

[0107] Preferably, assuming the threshold is ±1.5 microns, a difference of 2 microns exceeds the range, indicating that the prediction model needs to be adjusted.

[0108] Specifically, for the case where the error range exceeds the standard, when determining the set of parameters that need to be adjusted, parameters such as lubrication flow rate, injection pressure, and execution cycle can be considered.

[0109] Exemplarily, in one adjustment, it is found that the lubrication flow rate is too low, resulting in insufficient oil film thickness, so it is included in the set of parameters. When using the gradient descent algorithm to adjust the weight parameters of the regression model.

[0110] It can be understood that the gap between the predicted value and the measured value is gradually reduced through multiple iterations.

[0111] For example, the initial weight may be biased towards the injection pressure, while the lubrication flow rate has a greater actual impact. The algorithm will gradually increase the weight ratio of the flow rate and finally generate updated model parameters.

[0112] In one embodiment, after the updated model parameters are used to recalculate the predicted oil film thickness value, a predicted result of 21 microns may be obtained, and the difference from the measured value of 20 microns is reduced to 1 micron. At this time, when comparing the error range again, assuming that it is still necessary to reach within ±0.8 microns, further adjustment is required.

[0113] It can be understood that this iterative process continuously optimizes the weights, making the model closer to the actual working conditions and effectively improving the prediction accuracy.

[0114] For example, when judging whether the error range is shrinking, the volatility of the measured data can be analyzed from multiple aspects. In one possible implementation, if the measured data is 19 microns, 20 microns, and 21 microns in different die areas respectively, it indicates that the oil film distribution uniformity is good, and the error mainly comes from the systematic deviation of the model prediction rather than random noise. At this time, adjusting the weight parameters is more targeted and can quickly converge to the target accuracy.

[0115] Preferably, the predicted value output by the final model may be stabilized at 20.5 microns, and the error is controlled within ±0.5 microns, significantly improving the reliability of the system.

[0116] It should be noted that when the optimized predicted oil film thickness value is output through the final regression model, its consistency can be verified by combining historical data.

[0117] For example, in multiple tests, the matching degree between the optimized predicted value and the measured value is increased from 80% to 95%, indicating that the accuracy of the system execution is optimized. This method can not only reduce the problems of insufficient or excessive lubrication, but also extend the die life and improve the production efficiency.

[0118] Specifically, in a certain production, due to the more accurate predicted value, the lubrication system avoids die wear caused by too thin oil film and reduces the maintenance cost.

[0119] In one embodiment, the repeated adjustment process of the error range can also reflect the dynamic characteristics of the lubrication system.

[0120] For example, if the environmental temperature rises and causes a change in the lubricating oil viscosity, the measured oil film thickness may drop from 20 microns to 18 microns, and the optimized model can adapt to this change in a timely manner, adjusting the predicted value to 18.5 microns and keeping the error within a controllable range. This self-adaptive ability enables the system to maintain stable operation under complex working conditions, reflecting the practicality in technology.

[0121] Step S106, during the production process of the progressive die, automatically switch the lubrication parameter configuration according to the change of material characteristics at different stations to ensure that the lubrication effect at each station meets the requirements.

[0122] Obtain the material characteristic data of each station to get the characteristic change information; according to the characteristic change information, use the preset mapping rules to determine the adjustment direction of the lubrication parameters; for the adjustment direction, obtain the corresponding parameter configuration from the preset database to generate the adjusted configuration set; if the difference between the adjusted configuration set and the current configuration exceeds the preset threshold, update the lubrication parameters through the control system; obtain the lubrication effect data and determine whether the lubrication effect data meets the preset requirements of the differences between stations; if not, optimize the adjusted configuration set using the support vector machine algorithm to obtain the optimized configuration set; update the production process through the optimized configuration set to complete the automatic switching of the lubrication parameters.

[0123] Exemplarily, obtaining the material characteristic data of each station in real time through sensors is the basis for ensuring the dynamic adjustment of lubrication parameters.

[0124] Exemplarily, at the steel plate stamping station, the sensor can detect characteristics such as the hardness, surface roughness, and temperature of the material. Suppose the hardness of a certain batch of steel plates increases from the standard value of 200HB to 220HB, and the surface roughness increases from Ra1.6 to Ra2.0. This indicates that the material characteristics have changed and stronger lubrication support may be required. The characteristic change information is transmitted to the system in the form of a data stream by the sensor, providing a basis for subsequent decisions. According to the characteristic change information, using the preset mapping rules to determine the adjustment direction of the lubrication parameters is the key link to transform data into actions.

[0125] In a possible implementation, the mapping rules can be established based on historical data. For example, for every 10HB increase in hardness, the lubricant spraying amount increases by 5ml / min, and for every 10°C increase in temperature, the lubrication frequency increases by 10%.

[0126] Specifically, if it is detected that the hardness increases to 220HB and the temperature rises from 25°C to 35°C, the mapping rules may prompt an increase in the spraying amount to 15ml / min and an increase in the lubrication frequency at the same time. This rule design comes from the correlation between material characteristics and lubrication requirements, ensuring that the adjustment direction conforms to the actual working conditions. For the adjustment direction, obtaining the corresponding parameter configuration from the database to generate the adjusted configuration set is the specific step to achieve parameter matching.

[0127] For example, the database stores the configuration corresponding to a hardness of 220HB and a temperature of 35°C: a spraying amount of 15ml / min, a spraying pressure of 3bar, and a lubricant type of high-viscosity oil.

[0128] It should be noted that the configuration set is not limited to a single parameter, but also includes combination schemes to ensure that the lubrication effect fully covers the requirements of the workstations. After the adjusted configuration set is generated, it is compared with the current configuration. If the difference exceeds the preset threshold, such as the spraying volume changes by more than 20%, an update is triggered. The lubrication parameters are updated through the control system to ensure that the adjustment takes effect in real time.

[0129] In one embodiment, the control system sends the new configuration to the spraying equipment through the PLC to complete the parameter switching.

[0130] Preferably, the system will complete the update within 5 seconds to avoid production interruption. The real-time monitoring of the lubrication effect data is used to verify the effectiveness of the adjustment.

[0131] For example, when monitoring the uniformity of the oil film, if the deviation of the oil film thickness at a certain workstation decreases from 10 μm to 5 μm, it indicates that the effect meets the requirements; otherwise, further optimization is required. According to the judgment result, using the support vector machine algorithm to optimize the parameter configuration is an intelligent means to improve the accuracy.

[0132] It can be understood that the support vector machine finds the optimal parameter combination by analyzing the historical lubrication effect data and material characteristics.

[0133] For example, in the case of a hardness of 220 HB, the algorithm may recommend adjusting the spraying volume to 16 ml / min and slightly adjusting the pressure to 3.2 bar. This optimization can better adapt to the differences between workstations and improve production consistency. Updating the production process through the optimized configuration set to achieve automatic switching of lubrication parameters is the closed-loop embodiment of the entire process.

[0134] In one embodiment, due to frequent changes in material characteristics at a certain workstation, the configuration is switched frequently, and the production efficiency is increased by 15% and the scrap rate is reduced by 8%.

[0135] Specifically, the automatic switching avoids the delay of manual intervention and ensures that the lubrication parameters always match the working conditions. This method not only improves the response speed but also enhances the adaptive ability of the system, providing technical support for stable production.

[0136] Step S107, through the data linkage between the lubrication system and the mold temperature sensor, establish a dynamic adjustment mechanism for lubrication parameters and mold temperature to achieve real-time optimization of the lubrication effect.

[0137] Obtain the sensor data of the lubrication system and establish the initial mapping relationship between the lubrication parameters and the die temperature; extract the data linkage eigenvalue from the initial mapping relationship and construct the core algorithm of the dynamic adjustment mechanism; use the linear regression algorithm to obtain the adjustment scheme of the lubrication parameters based on the data linkage eigenvalue; according to the output of the adjustment scheme, obtain the change trend of the die temperature and judge the real-time change of the lubrication effect; if the lubrication effect is lower than the preset threshold, update the lubrication parameters through the dynamic adjustment mechanism to obtain a new lubrication scheme; according to the updated lubrication scheme, obtain the feedback value of the sensor data and determine the stability of the effect feedback; through the analysis of the effect feedback stability, use the support vector machine algorithm to judge the correlation between the temperature change and the lubrication effect and obtain the optimized adjustment strategy; extract the real-time optimization key parameters from the optimized adjustment strategy and construct the closed-loop control process of the lubrication system and the die temperature.

[0138] Exemplarily, obtaining the sensor data through the lubrication system and determining the initial mapping relationship between the lubrication parameters and the die temperature is the basis of the whole process.

[0139] For example, in progressive die production, the sensor can collect the data of the lubricating oil flow rate, pressure and the die surface temperature in real time.

[0140] Exemplarily, assume that the initial mapping relationship shows that when the die temperature rises from 50 degrees to 80 degrees, the lubricating oil flow rate needs to increase from 10 ml / min to 15 ml / min to maintain a stable lubricating film. The establishment of this mapping relationship depends on the accumulation of historical production data, can intuitively reflect the correlation between temperature and lubrication requirements, and helps the accuracy of subsequent adjustments. Extracting the data linkage eigenvalue from the initial mapping relationship and constructing the core algorithm of the dynamic adjustment mechanism is often the key to technical implementation.

[0141] Specifically, the temperature change rate, the change of lubricating oil viscosity, etc. can be selected as the eigenvalues.

[0142] In a possible implementation, if the temperature change rate exceeds 5 degrees per minute, the system will consider that the lubrication parameters need to be adjusted quickly. When using the linear regression algorithm, by analyzing the data of the past 100 productions, it can be obtained that for every 10-degree increase in temperature, the lubricating oil flow rate needs to increase by 2 ml / min. This method is simple and efficient and can provide a reliable basis for dynamic adjustment. Obtaining the change trend of the die temperature and judging the real-time change of the lubrication effect for the output of the dynamic adjustment mechanism is an important link in the closed-loop control.

[0143] For example, when the die temperature continues to rise and the lubrication effect decreases, the system will record that the lubricating film thickness decreases from 0.1 mm to 0.08 mm.

[0144] Preferably, by comparing with a preset threshold value such as 0.09 mm, it is determined that the current lubrication is insufficient. This real-time judgment can detect problems in a timely manner and avoid the aggravation of die wear caused by improper lubrication. If the lubrication effect is lower than the preset threshold value, the lubrication parameters are updated through an adjustment mechanism to obtain a new lubrication plan.

[0145] It can be understood that the updated plan may increase the lubricating oil flow rate from 15 ml / min to 18 ml / min and at the same time increase the injection pressure to 2.5 bar to enhance the lubrication coverage rate. This adjustment can quickly respond to changes in production and improve the processing stability. According to the updated lubrication plan, obtaining the feedback value of the sensor data and determining the stability of the effect feedback are the key steps to verify the effectiveness of the adjustment.

[0146] In one embodiment, the feedback value shows that the lubricating film thickness has recovered to 0.1 mm and remains stable within the next 10 minutes. This indicates that the new plan can adapt to the current working conditions and avoid the production uncertainty caused by frequent fluctuations. Through the stability analysis of the effect feedback, using the support vector machine algorithm to judge the correlation between the temperature change and the lubrication effect can further optimize the adjustment strategy.

[0147] For example, the support vector machine may analyze that when the temperature change rate exceeds 8 degrees / minute, the probability of the lubrication effect decreasing is as high as 90%. This correlation analysis can provide a more accurate direction for parameter optimization and avoid the waste of resources caused by blind adjustment. Extracting the key parameters of real-time optimization from the optimized adjustment strategy and constructing a closed-loop control process for the lubrication system and the die temperature are the core value of the entire plan.

[0148] Specifically, the key parameters may include a lubricating oil flow rate of 18 ml / min, an injection pressure of 2.5 bar, and a temperature change rate threshold of 8 degrees / minute.

[0149] In one embodiment, the closed-loop control process will automatically calibrate the lubrication system every 5 minutes according to these parameters. This mechanism can significantly improve the production efficiency and at the same time reduce the need for manual intervention, providing stable support for progressive die production.

[0150] Step S108, based on the lubricant consumption data in the production process, combined with the change trends of the oil film thickness and the die temperature, analyze the use efficiency of the lubricant and optimize the lubricant supply strategy.

[0151] Obtain lubricant consumption data and oil film thickness data during the production process to obtain an initial data set; according to the initial data set, use the linear regression algorithm to determine the relationship model between lubricant consumption and oil film thickness; obtain die temperature data, and according to the relationship model, judge the influence of temperature change on the oil film thickness to obtain a temperature correction coefficient; obtain the temperature correction coefficient and lubricant consumption data, use the support vector machine algorithm to judge the lubricant use efficiency, and determine the efficiency distribution law; according to the efficiency distribution law and the preset supply strategy, use the decision tree algorithm to judge the optimization direction to obtain a strategy adjustment parameter; according to the strategy adjustment parameter and production process data, determine an optimized lubricant supply plan; obtain the lubricant consumption data corresponding to the optimized lubricant supply plan, combine the change trend of lubricant consumption, judge the improvement range of lubricant use efficiency, and output the optimization result.

[0152] Exemplarily, obtaining lubricant consumption data and oil film thickness data during the production process is the key to establishing the analysis foundation.

[0153] Exemplarily, in progressive die production, the sensor can record that the lubricant consumption per minute is 12 milliliters, and at the same time, the oil film thickness is stable at 0.1 millimeter. This data acquisition relies on high-precision flow meters and thickness gauges to provide reliable support for subsequent modeling. When using the linear regression algorithm to determine the relationship model between lubricant consumption and oil film thickness according to the initial data set.

[0154] It can be understood that this process aims to explore the linear correlation between the two.

[0155] Specifically, by analyzing the past 50 production data, it may be found that for every 5-milliliter increase in lubricant consumption, the oil film thickness increases by 0.02 millimeter. This relationship model is simple and intuitive, laying the foundation for subsequent analysis. Obtaining die temperature data and judging its influence on the oil film thickness often requires considering the changes in the production environment.

[0156] In a possible implementation, when the die temperature rises from 60 degrees to 90 degrees, the oil film thickness drops from 0.1 millimeter to 0.09 millimeter. Based on this, the temperature correction coefficient can be obtained through regression analysis. For example, the thickness decreases by 0.01 millimeter for every 10-degree increase. This coefficient reflects the direct effect of temperature on the lubrication effect. When using the support vector machine algorithm to judge the lubricant use efficiency, the temperature correction coefficient and consumption data are used as inputs together.

[0157] Preferably, if the lubricant consumption is 15 milliliters but the oil film thickness is only 0.08 millimeter during a certain production, the efficiency may be determined to be low.

[0158] In one embodiment, after the support vector machine analyzes 100 times of production data, the efficiency distribution law is obtained: when the temperature is higher than 80 degrees, the probability of efficiency decline reaches 85%. This provides data support for optimization. When using the decision tree algorithm according to the efficiency distribution law and the preset supply strategy.

[0159] It should be noted that the decision tree can clearly divide the adjustment direction.

[0160] For example, when the efficiency is lower than 70% and the temperature exceeds 85 degrees, it is recommended to increase the lubricant supply amount to 18 ml / min. At the same time, if the oil film thickness is lower than 0.09 mm, the injection pressure can be adjusted to 2.8 bar. This strategy adjustment parameter is intuitive and easy to implement. After determining the optimized lubricant supply plan, it is a key step to verify it in combination with the production process data.

[0161] In one embodiment, after adjustment, the lubricant consumption increases to 18 ml / min and the oil film thickness recovers to 0.1 mm. This solution can quickly adapt to the working condition changes and improve the stability. When obtaining the lubricant consumption data of the optimized solution and analyzing the change trend.

[0162] It can be understood that the consumption may be stable in the range of 17 - 18 ml / min within 10 minutes, and the fluctuation of the oil film thickness is less than 0.01 mm.

[0163] For example, by comparing the data before and after optimization, the efficiency is increased by about 20%. This analysis can verify the effectiveness of the adjustment and provide a direction for continuous improvement of production.

[0164] The present invention provides a stamping lubrication adaptive control system for metal stamping parts, mainly including:

[0165] A friction coefficient acquisition module, which is used to obtain the friction coefficient of the material from a pre-established database according to the material information of the stamping material, and combine the real-time monitoring data of the mold temperature sensor to determine the optimal oil film thickness range under the current working conditions;

[0166] An oil film thickness prediction module, which is used to train an oil film thickness prediction model based on the historical data of the friction coefficient and the mold temperature by using a regression model in machine learning algorithms, and predict the target oil film thickness value by inputting the current friction coefficient and the mold temperature;

[0167] A lubrication parameter calculation module, which is used to calculate the injection amount and injection frequency of the lubricant for the predicted oil film thickness value, combine the control parameter range of the lubrication system, generate a lubrication parameter adjustment instruction and send it to the lubrication system execution module;

[0168] A temperature monitoring module, which is used to judge whether the temperature exceeds the preset threshold by monitoring the change trend of the die temperature in real time during the execution of the lubrication system. If the threshold is exceeded, the lubrication parameters are recalculated and the execution instructions are updated;

[0169] A model optimization module, which is used to compare the measured data of the oil film thickness after the execution of the lubrication system with the predicted value. If the error exceeds the preset range, the weight parameters of the regression model are adjusted to optimize the prediction accuracy;

[0170] A lubrication parameter switching module, which is used to automatically switch the lubrication parameter configuration according to the change of material characteristics at different stations during the production of progressive dies, so as to ensure that the lubrication effect at each station meets the requirements;

[0171] A dynamic adjustment module, which is used to establish a dynamic adjustment mechanism between lubrication parameters and die temperature through the data linkage between the lubrication system and the die temperature sensor, so as to realize the real-time optimization of the lubrication effect;

[0172] A lubricant optimization module, which is used to analyze the use efficiency of the lubricant based on the lubricant consumption data during the production process, combined with the change trends of the oil film thickness and the die temperature, and optimize the lubricant supply strategy.

[0173] As mentioned above, the above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution of the present invention and its inventive concept, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. A stamping lubrication adaptive control method for metal stamping parts, characterized in that: The method comprises the following steps: Step S101, according to the material information of the stamping material, the friction coefficient of the material is obtained from a pre-established database, and combined with the real-time monitoring data of the mold temperature sensor, the optimal oil film thickness range under the current working condition is determined; Step S102: using a regression model in a machine learning algorithm, based on historical data of friction coefficient and mold temperature, to train an oil film thickness prediction model, and predicting a target oil film thickness value by inputting the current friction coefficient and mold temperature; Step S103, for the predicted oil film thickness value, combined with the control parameter range of the lubrication system, the injection amount and injection frequency of the lubricant are calculated, and a lubrication parameter adjustment instruction is generated and sent to the lubrication system execution module; Step S104: During the execution of the lubrication system, by real-time monitoring the change trend of the mold temperature, determine whether the temperature exceeds a preset threshold. If the temperature exceeds the threshold, recalculate the lubrication parameters and update the execution instructions; Step S105, comparing the measured data of the oil film thickness after the lubrication system is executed with the predicted value, and if the error exceeds a preset range, adjusting the weight parameter of the regression model to optimize the prediction accuracy; Step S106: During the continuous mold production process, according to the changes in material properties of different workstations, the lubrication parameter configuration is automatically switched to ensure that the lubrication effect of each workstation meets the requirements; Step S107: establishing a dynamic adjustment mechanism of lubrication parameters and mold temperature through data linkage between the lubrication system and the mold temperature sensor, so as to achieve real-time optimization of the lubrication effect; Step S108: Based on the lubricant consumption data during the production process, combined with the change trends of the oil film thickness and the mold temperature, the lubricant usage efficiency is analyzed and the lubricant supply strategy is optimized.

2. According to claim 1, a stamping lubrication adaptive control method for metal stamping parts is characterized in that: The step S101 includes: Acquire initial friction parameters of the stamping material, wherein the initial friction parameters are acquired from a pre-established material information database; Acquire current temperature data of the mold, where the current temperature data is obtained by real-time monitoring by a temperature sensor; Determine whether there is a preset matching relationship between the initial friction parameter and the current temperature data, and if so, obtain the adjusted friction coefficient from a preset mapping table; According to the adjusted friction coefficient, obtaining an oil film thickness range table associated with the current working condition; Determine the upper and lower limit thresholds of the optimal oil film thickness range according to the oil film thickness range table and the current temperature data; Acquire historical operating condition data and friction coefficient data, and use a machine learning algorithm to analyze the historical operating condition data and friction coefficient data to obtain a dynamic adjustment value of the oil film thickness range; The final optimal oil film thickness range is determined according to the matching degree between the dynamic adjustment value and the current working condition.

3. According to the method for self-adaptive control of stamping lubrication of metal stamping parts according to claim 1, it is characterized in that: The step S102 includes: Acquiring historical data, the historical data including records of friction coefficient and mold temperature; Using a linear regression algorithm to train the historical data to obtain initial model parameters; Acquiring current values, the current values ​​including real-time data of friction coefficient and mold temperature; Inputting the real-time data into the initial model parameters to obtain a preliminary oil film thickness prediction value; Determine whether the initial oil film thickness prediction value exceeds a preset threshold value, and if so, use a support vector regression algorithm to adjust the initial model parameters to obtain optimized model parameters; Inputting the real-time data into the optimization model parameters to obtain an adjusted oil film thickness value; Obtaining the deviation between the adjusted oil film thickness value and the historical data, determining the stability of the prediction model, and obtaining a deviation analysis result; Update the training process according to the deviation analysis result, and retrain the historical data and the real-time data using a linear regression algorithm to obtain updated model parameters; The real-time data is input into the updated model parameters to obtain a final predicted value of the oil film thickness.

4. According to claim 1, a stamping lubrication adaptive control method for metal stamping parts is characterized in that: The step S103 comprises: Obtaining a predicted value of oil film thickness characterizing the lubrication state, and calculating the corresponding injection amount and injection frequency using a linear regression algorithm based on the predicted value of oil film thickness and a preset lubrication parameter range to obtain preliminary lubrication parameters; Obtaining the constraint conditions corresponding to the preliminary lubrication parameters, determining whether the injection amount exceeds a preset range, and if so, reducing the injection amount to within a preset range according to a preset ratio to obtain an adjusted lubrication parameter; generating a lubrication parameter adjustment instruction according to the adjusted lubrication parameter and in combination with the change trend of the oil film thickness; Send the lubrication parameter adjustment instruction to the lubrication execution module, obtain the reception status of the lubrication system for the adjustment instruction, determine whether the instruction transmission delay exceeds a preset threshold, and if so, regenerate and send the lubrication parameter adjustment instruction until the final execution status is determined; The oil film thickness change value fed back by the lubrication system is obtained, the consistency between the oil film thickness change value and the predicted value is determined, and the lubrication control parameters are optimized based on the consistency analysis result to obtain the optimized lubrication control parameters.

5. A stamping lubrication adaptive control method for metal stamping parts according to any one of claims 1 to 4, characterized in that: The step S105 comprises: Obtaining measured data of oil film thickness generated after the lubrication system is executed, and calculating the difference between the measured data and the predicted value output by the regression model; Extracting an error range from the difference, determining whether the error range exceeds a preset threshold, and if so, determining a set of parameters that need to be adjusted; For the parameter set, a gradient descent algorithm is used to adjust the weight parameters in the regression model to obtain updated model parameters; Recalculating the predicted value of the oil film thickness according to the updated model parameters to obtain a new prediction result; After obtaining the new prediction result, compare it with the measured data again to determine whether the error range is reduced to within the preset threshold. If it still exceeds the preset threshold, repeat the process of adjusting the weight parameter until a prediction result that meets the preset accuracy condition is obtained; According to the regression model that meets the preset accuracy conditions, the optimized oil film thickness prediction value is output to complete the accuracy optimization of the system execution.

6. A stamping lubrication adaptive control method for metal stamping parts according to any one of claims 1 to 4, characterized in that: The step S106 comprises: Obtain material property data at each workstation and obtain property change information; According to the characteristic change information, a preset mapping rule is used to determine the adjustment direction of the lubrication parameters; According to the adjustment direction, corresponding parameter configuration is obtained from a preset database to generate an adjusted configuration set; If the difference between the adjusted configuration set and the current configuration exceeds a preset threshold, updating the lubrication parameters through the control system; Obtain lubrication effect data, and determine whether the lubrication effect data meets preset requirements for differences among various workstations; If not, a support vector machine algorithm is used to optimize the adjusted configuration set to obtain an optimized configuration set; The production process is updated by the optimized configuration set to complete the automatic switching of lubrication parameters.

7. A stamping lubrication adaptive control method for metal stamping parts according to any one of claims 1 to 4, characterized in that: The step S107 includes: Obtain sensor data from the lubrication system and establish an initial mapping relationship between lubrication parameters and mold temperature; Extracting data linkage feature values ​​from the initial mapping relationship and constructing a core algorithm for a dynamic adjustment mechanism; Using a linear regression algorithm, based on the data linkage characteristic values, an adjustment plan for lubrication parameters is obtained; According to the output of the adjustment plan, the mold temperature change trend is obtained to determine the real-time change of the lubrication effect; If the lubrication effect is lower than a preset threshold, the lubrication parameters are updated through the dynamic adjustment mechanism to obtain a new lubrication solution; According to the updated lubrication plan, obtain the feedback value of the sensor data and determine the stability of the effect feedback; By analyzing the stability of the effect feedback, a support vector machine algorithm is used to determine the correlation between temperature change and lubrication effect, and an optimized adjustment strategy is obtained; Real-time optimization key parameters are extracted from the optimized adjustment strategy to build a closed-loop control process for the lubrication system and mold temperature.

8. A stamping lubrication adaptive control method for metal stamping parts according to any one of claims 1 to 4, characterized in that: The step S108 includes: Obtain lubricant consumption data and oil film thickness data during the production process to obtain an initial data set; According to the initial data set, a linear regression algorithm is used to determine a relationship model between lubricant consumption and oil film thickness; Acquire mold temperature data, determine the effect of temperature change on oil film thickness based on the relationship model, and obtain a temperature correction coefficient; Obtain the temperature correction coefficient and lubricant consumption data, use a support vector machine algorithm to determine the lubricant usage efficiency, and determine the efficiency distribution law; According to the efficiency distribution law and the preset supply strategy, a decision tree algorithm is used to determine the optimization direction and obtain the strategy adjustment parameters; Determine an optimized lubricant supply plan based on the strategy adjustment parameters and production process data; The lubricant consumption data corresponding to the optimized lubricant supply scheme is obtained, and the improvement range of the lubricant use efficiency is determined based on the lubricant consumption change trend, and the optimization result is output.

9. A stamping lubrication adaptive control system for metal stamping parts, characterized in that: The system is used to implement the stamping lubrication adaptive control method of a metal stamping part according to any one of claims 1 to 8, and the system comprises: The friction coefficient acquisition module is used to obtain the friction coefficient of the material from the pre-established database according to the material information of the stamping material, and determine the optimal oil film thickness range under the current working conditions in combination with the real-time monitoring data of the mold temperature sensor; The oil film thickness prediction module is used to use the regression model in the machine learning algorithm to train the oil film thickness prediction model based on the historical data of the friction coefficient and the mold temperature. By inputting the current friction coefficient and the mold temperature, the target oil film thickness value is predicted; A lubrication parameter calculation module is used to calculate the injection amount and injection frequency of the lubricant based on the predicted oil film thickness value and the control parameter range of the lubrication system, generate a lubrication parameter adjustment instruction and send it to the lubrication system execution module; The temperature monitoring module is used to monitor the change trend of the mold temperature in real time during the execution of the lubrication system to determine whether the temperature exceeds the preset threshold. If it exceeds the threshold, the lubrication parameters are recalculated and the execution instructions are updated; The model optimization module is used to compare the measured data of the oil film thickness after the lubrication system is executed with the predicted value. If the error exceeds the preset range, the weight parameters of the regression model are adjusted to optimize the prediction accuracy. The lubrication parameter switching module is used to automatically switch the lubrication parameter configuration according to the changes in material properties of different workstations during the continuous mold production process to ensure that the lubrication effect of each workstation meets the requirements; The dynamic adjustment module is used to establish a dynamic adjustment mechanism for lubrication parameters and mold temperature through data linkage between the lubrication system and the mold temperature sensor, so as to achieve real-time optimization of the lubrication effect; The lubricant optimization module is used to analyze the lubricant usage efficiency and optimize the lubricant supply strategy based on the lubricant consumption data in the production process and the changing trends of oil film thickness and mold temperature.

Images