Continuous punch forming method for high-hardness alloy thin-wall lock case

Abstract

The invention relates to a continuous punch forming method of a high-hardness alloy thin-wall lock case in the field of intelligent manufacturing, which comprises the following steps: according to a stress distribution diagram, adjusting a die gap by adopting a self-adaptive algorithm, dynamically optimizing gap parameters for a local stress concentration area, and determining real-time gap configuration; if the temperature deviation of any area in the temperature field exceeds 10 DEG C, recalculating heat source distribution through a heat flow simulation algorithm, and determining optimized heating parameters; acquiring surface topography data of the annealed thin-wall lock case through an optical scanner, and analyzing fracture and wrinkling defects by adopting an image processing algorithm to obtain defect distribution characteristics; if the proportion of the fracture area in the defect distribution characteristics exceeds 2%, the forming failure probability is predicted through a neural network model, and whether the stamping speed is adjusted or not is determined; and according to the defect distribution characteristics and the softening consistency data, a data fusion algorithm is adopted to comprehensively evaluate the forming stability, and optimized operation parameters of the continuous stamping production line are obtained.

Description

Technical Field

[0001] The present invention relates to the field of intelligent manufacturing, and in particular to a continuous stamping method for a high-hardness alloy thin-wall lock shell. Background Art

[0002] Precision metal stamping technology is crucial for manufacturing high-strength, lightweight, and complex-shaped security lock cases. As core components housing the lock key, these cases directly impact safety and production efficiency. However, existing processing methods struggle to meet the complex forming requirements of thin-walled alloy materials, necessitating breakthroughs to improve product quality and production efficiency.

[0003] Traditional processes face significant limitations when processing high-hardness iron alloys (such as Fe-Cr-Ni-Mo). Single annealing processes are time-consuming and inefficient, making them difficult to integrate with continuous stamping lines. Fixed-gap dies cannot adapt to localized deformation in thin-walled parts, resulting in low yield rates. Solutions like staged annealing or hydraulic compensation are fragmented, energy-intensive, and result in short die life. These shortcomings make stable production of thin-walled lock cases a challenge for the industry.

[0004] Work hardening of high-hardness alloy materials during thin-wall stretching presents a primary challenge. Due to the material's high hardening index, stress accumulates significantly during continuous stretching, leading to frequent cracking and wrinkling. Work hardening further complicates mold clearance adjustment, making conventional molds unable to dynamically adjust to material deformation, resulting in insufficient dimensional accuracy and excessive springback. Inaccurate control of the annealing process leads to uneven softening of the material, and temperature deviations directly impact forming stability. These interrelated factors collectively limit the efficient and precise forming of thin-walled lock cases.

[0005] Therefore, how to achieve stable forming of high-hardness alloy thin-walled parts through precise annealing temperature control and dynamic die gap adjustment in the continuous stamping production line has become a key issue that needs to be solved urgently in this study. Summary of the Invention

[0006] The present invention provides a continuous stamping method for a high-hardness alloy thin-wall lock shell, comprising the following steps:

[0007] The stress distribution data of the high-hardness alloy thin-walled lock shell during continuous stamping is collected in real time by sensors, and the stress accumulation degree caused by work hardening is calculated using a finite element analysis model to obtain a stress distribution diagram;

[0008] Based on the stress distribution diagram, an adaptive algorithm is used to adjust the mold gap, dynamically optimize the gap parameters for the local stress concentration area, and determine the real-time gap configuration;

[0009] If the real-time gap configuration deviates from the preset threshold by more than 5%, the mold adjustment mechanism is driven by the servo motor to obtain the adjusted mold gap data and determine the gap adaptability;

[0010] The real-time temperature distribution of the thin-walled lock shell during the annealing process is obtained by an infrared thermometer, and the PID control algorithm is used to adjust the heat source output of the annealing furnace to obtain a uniform temperature field;

[0011] If the temperature deviation in any area of ​​the temperature field exceeds 10°C, the heat source distribution is recalculated using the heat flow simulation algorithm to determine the optimized heating parameters;

[0012] According to the optimized heating parameters, a closed-loop control system is used to adjust the annealing furnace operation status, obtain the softening degree data of the material after annealing, and judge the softening consistency;

[0013] The surface morphology data of the thin-walled lock shell after annealing is collected by an optical scanner, and the crack and wrinkle defects are analyzed using image processing algorithms to obtain the defect distribution characteristics.

[0014] If the crack area in the defect distribution characteristics exceeds 2%, the probability of forming failure is predicted through the neural network model to determine whether to adjust the stamping speed;

[0015] According to the defect distribution characteristics and softening consistency data, a data fusion algorithm is used to comprehensively evaluate the forming stability and obtain the optimized operating parameters of the continuous stamping production line.

[0016] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0017] The present invention discloses a method for continuous stamping of thin-walled lock shells made of high-hardness alloys. The method collects stress distribution data in real time, calculates stress accumulation using finite element analysis, and dynamically optimizes the mold gap based on the stress distribution diagram. At the same time, infrared temperature measurement and PID control algorithms are used to adjust the temperature field during the annealing process to ensure uniform softening. In addition, the present invention also uses optical scanning to analyze surface defects and combines a neural network model to predict the probability of forming failure. Finally, the forming stability is comprehensively evaluated through a data fusion algorithm to obtain optimized operating parameters. This method realizes real-time monitoring and intelligent control during the continuous stamping process of thin-walled lock shells made of high-hardness alloys, effectively improving the forming quality and production efficiency, and providing a new technical idea for similar high-difficulty stamping processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 The present invention is a flow chart of a continuous stamping method for a high-hardness alloy thin-wall lock shell.

[0019] Figure 2 Schematic diagram of a continuous stamping method for a high-hardness alloy thin-wall lock shell according to the present invention.

[0020] Figure 3 This is another schematic diagram of the continuous stamping method of the present invention for a high-hardness alloy thin-wall lock housing. DETAILED DESCRIPTION

[0021] To further understand the content of the present invention, the present invention is described in detail with reference to the accompanying drawings and examples. The present application is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are intended only to illustrate the relevant invention and are not intended to limit the invention. It should also be noted that, for ease of description, only portions relevant to the invention are shown in the accompanying drawings.

[0022] like Figure 1-3 In this embodiment, a continuous stamping method for a high-hardness alloy thin-wall lock shell may specifically include:

[0023] Step S101 , using sensors to collect real-time stress distribution data of a high-hardness alloy thin-walled lock housing during continuous stamping, and using a finite element analysis model to calculate the stress accumulation degree caused by work hardening to obtain a stress distribution diagram.

[0024] Real-time stress data of a high-hardness alloy thin-walled lock housing during continuous stamping is collected using sensors to obtain an original stress dataset. Data preprocessing methods are used to denoise and standardize the original stress dataset to obtain a processed stress dataset. If outliers exist in the processed stress dataset, these are removed using a median filter algorithm to obtain a filtered stress dataset. The filtered stress dataset is calculated using a finite element analysis model to determine the stress accumulation value caused by work hardening. The stress accumulation value is compared with a preset stress threshold. If the stress accumulation value exceeds the threshold, it is marked as a high-stress area to obtain a stress distribution feature. A visualization algorithm is used to map the stress distribution feature into a two-dimensional stress distribution map to generate a stress distribution image. The stress distribution image is contrast-adjusted using image enhancement technology to obtain an optimized stress distribution map.

[0025] Specifically, sensors were used to collect real-time stress distribution data for a thin-walled lock case made of a high-hardness alloy during continuous stamping. A finite element analysis model was then used to calculate the degree of stress accumulation caused by work hardening, generating a stress distribution diagram. First, during the stamping process, strain gauge sensors were used to monitor stress changes on the lock case surface in real time at a sampling frequency of 1000 times per second. The collected data was transmitted to a data processing center via a wireless transmission module. The data processing center then filtered the raw signal using a fast Fourier transform (FFT) algorithm to remove noise and extract valid stress data. Next, based on the finite element analysis model, a three-dimensional model of the lock case was constructed using ABAQUS software. The mesh was meshed using tetrahedral elements with a cell size of 0.1 mm to ensure accuracy. In the model, the material properties were set to a high-hardness alloy with an elastic modulus of 210 GPa, a Poisson's ratio of 0.3, and a yield strength of 1200 MPa. By simulating the continuous stamping process, the stress distribution at each time step was calculated, and the degree of stress accumulation was evaluated using the Von Mises stress criterion. The simulation results show that after 1000 stamping cycles, the maximum stress on the lock case surface reaches 850 MPa, concentrated primarily at the edge of the case. Finally, the simulation results were compared with sensor data, and the error was kept within 5%, verifying the accuracy of the model. The stress distribution diagram allows for intuitive visualization of stress concentration areas during the stamping process, providing data support for optimizing the stamping process.

[0026] Step S102 : According to the stress distribution diagram, an adaptive algorithm is used to adjust the mold gap, dynamically optimize the gap parameters for the local stress concentration area, and determine the real-time gap configuration.

[0027] An initial stress distribution diagram is obtained, wherein the initial stress distribution diagram includes data of multiple stress concentration areas. The initial stress distribution diagram is meshed using an adaptive algorithm to extract stress gradient characteristics of local stress concentration areas to obtain a stress gradient distribution. The stress gradient distribution is iteratively calculated using a finite element analysis model to determine the mold gap adjustment amount and obtain a gap adjustment parameter. If the gap adjustment parameter exceeds a preset gap threshold, the gap adjustment parameter is smoothed using a linear interpolation algorithm to obtain a smoothed gap parameter. Real-time gap configuration data is generated based on the smoothed gap parameter and mapped to mold control instructions to obtain a gap control instruction set. The gap control instruction set is transmitted to the mold control system via a real-time data interface, and the gap configuration is dynamically updated to obtain an updated gap state. Stress distribution data corresponding to the updated gap state is obtained and compared with the initial stress distribution diagram to determine the change trend of the stress concentration area. If the change trend shows that the stress concentration area has not decreased, the weight parameter of the adaptive algorithm is adjusted and the stress gradient distribution is recalculated.

[0028] Specifically, based on the analysis results of the stress distribution diagram, an adaptive genetic algorithm was used to dynamically optimize the die gap to reduce local stress concentration. The algorithm's initial population size was set to 50, with a crossover probability of 0.8, a mutation probability of 0.1, and 100 iterations. The objective function was to minimize the Von Mises stress value. During the optimization process, stress concentration in the lock shell edge area was addressed by using a finite element analysis model to calculate the stress distribution under different gap parameters in real time. The gap adjustment range was 0.05mm to 0.15mm, with a step size of 0.01mm. After each iteration, the algorithm updated the gap parameters based on the stress calculation results and re-evaluated the stress distribution. During the optimization process, the relationship between the gap parameters and stress values ​​was modeled using the Kriging interpolation method to improve computational efficiency. After optimization, the maximum stress in the lock shell edge area was reduced from 850MPa to 720MPa, significantly improving the stress concentration. Furthermore, the optimized gap parameters were fed back to the die control system through a real-time monitoring system, enabling dynamic gap adjustment and ensuring the stability and consistency of the stamping process.

[0029] Step S103: If the real-time gap configuration deviates from the preset threshold by more than 5%, the mold adjustment mechanism is driven by the servo motor to obtain the adjusted mold gap data and determine the gap adaptability.

[0030] Obtain the current gap data of the mold mechanism and the preset gap threshold. If the deviation between the gap data and the preset threshold exceeds the specified range, a deviation value is generated by the deviation calculation module to obtain deviation data. Based on the deviation data, a control algorithm is used to generate control instructions for the servo motor and determine the motor drive parameters. The servo motor is driven by the motor drive parameters, the mold mechanism is adjusted, and the adjusted gap data is obtained. If the deviation between the adjusted gap data and the preset threshold is within the specified range, a feedback signal is obtained by the data acquisition module to determine the gap adaptability. If the feedback signal indicates that the gap adaptability is insufficient, the PID algorithm is used to optimize the control instructions to obtain optimized drive parameters. The servo motor is re-driven by the optimized drive parameters, the mold mechanism is adjusted, new gap data is collected, and the adjustment accuracy is determined. If the adjustment accuracy meets the preset requirements, the final gap data is recorded by the data processing module to determine that the mold adjustment is complete.

[0031] Specifically, during real-time monitoring of the mold gap, a high-precision laser sensor collects gap data at a resolution of 0.01mm. A moving average filter algorithm is used to process the raw data (with a window size of 10 sampling points). When the deviation between the current gap value (e.g., 3.52mm) and the preset threshold (3.35mm) reaches 5.07%, the control logic is triggered. The servo motor outputs a speed of 1275rpm based on a PID control algorithm (proportional coefficient Kp = 1.2, integral time Ti = 0.5s, differential time Td = 0.1s), driving the ball screw mechanism to adjust the mold position at a feed rate of 0.8mm / s. After adjustment, the sheet metal forming process is simulated using finite element analysis. The ratio of the maximum stress value (e.g., 285MPa) to the material yield strength (310MPa) is calculated (91.9%). Compatibility is then determined based on the gap uniformity index (standard deviation ≤ 0.05mm). If the stress ratio exceeds 95% or the standard deviation exceeds the standard, a secondary adjustment cycle is initiated until the process requirements are met. During the entire process, the data acquisition frequency is maintained at 1kHz and the control period is set to 5ms to ensure the real-time response of the system.

[0032] Step S104: obtaining the real-time temperature distribution of the thin-walled lock shell during the annealing process by using an infrared thermometer, and adjusting the heat source output of the annealing furnace by using a PID control algorithm to obtain a uniform temperature field.

[0033] The surface temperature of the thin-walled lock shell is collected by an infrared thermometer to generate real-time temperature distribution data. If there is an area in the temperature distribution data where the deviation exceeds a preset threshold, a filtering algorithm is used to process the temperature distribution data to obtain smoothed temperature distribution data. Based on the smoothed temperature distribution data, the deviation between the temperature of each area and the target temperature is calculated to generate a deviation matrix. A PID control algorithm is used to adjust the output power of the annealing furnace heat source according to the deviation matrix to obtain an optimized heat source distribution. The annealing furnace is driven to operate by the optimized heat source distribution to generate new temperature field data. If the temperature deviation in the new temperature field data is lower than the preset threshold, the temperature field is determined to be uniform and stable process parameters are output. Based on the stable process parameters, the operating state of the annealing furnace is adjusted to obtain a continuously uniform temperature field.

[0034] Specifically, during the annealing process, the temperature distribution of the thin-walled lock shell is monitored in real time by an infrared thermometer. The measurement accuracy of the infrared thermometer is ±0.5°C and the sampling frequency is 10Hz, ensuring the real-time and accuracy of the temperature data. The collected temperature data is transmitted to the control unit through the data acquisition system, and the control unit uses the PID control algorithm for temperature adjustment. The parameters of the PID control algorithm are set to the proportional coefficient Kp = 2.5, the integral time Ti = 10s, and the differential time Td = 2s. These parameters are adjusted using the Ziegler-Nichols method to ensure the rapid response and stability of the system. The control unit calculates the adjustment amount of the heat source output power based on the deviation between the real-time temperature data and the set temperature value (for example, 600°C), and adjusts the heating power of the annealing furnace through the actuator.

[0035] For example, if the temperature in a certain area is detected to be below the set value, the control unit will increase the heating power in that area by 10% of the current power to ensure temperature uniformity. Simultaneously, the control unit analyzes temperature distribution data to identify areas with large temperature gradients and adjusts the heating strategy accordingly, for example increasing the heating power by 5% in areas with large temperature gradients to reduce temperature differences. This process ultimately achieves a uniform temperature distribution for thin-walled lock shells during annealing, with temperature differences controlled within ±5°C to ensure annealing quality.

[0036] Step S105: If the temperature deviation of any area in the temperature field exceeds 10°C, the heat source distribution is recalculated by a heat flow simulation algorithm to determine the optimized heating parameters.

[0037] The current temperature field distribution is obtained from the real-time temperature monitoring data to obtain the regional temperature deviation value. If the temperature deviation of any area exceeds the preset temperature deviation threshold, the temperature field distribution is analyzed by the heat flow transfer model, and the heat source distribution is calculated by the heat flow simulation algorithm to obtain the preliminary heat source distribution result. According to the preliminary heat source distribution result and the parameter optimization target, the heat source position is adjusted to obtain the optimized heat source position data. If the deviation between the optimized heat source position data and the current heat source position exceeds the preset heat source position adjustment threshold, the input parameters of the heat flow simulation algorithm are recalculated to obtain the updated heat source distribution. The optimized heating parameters are determined by the updated heat source distribution to obtain the heating parameter adjustment scheme. According to the heating parameter adjustment scheme, combined with the regional division method, the temperature field distribution is recalculated to obtain new temperature field distribution data. If the temperature deviation of any area in the new temperature field distribution data still exceeds the preset temperature deviation threshold, the number of algorithm iterations is increased, and the heat flow simulation algorithm is repeatedly executed to obtain the final optimized heating parameters.

[0038] Specifically, if the temperature deviation in a certain area of ​​the temperature field exceeds 10°C, the system will automatically trigger the heat flow simulation algorithm to recalculate. First, the temperature field is meshed using finite element analysis (FEA). The temperature data of each grid cell is collected in real time and input into the heat flow simulation algorithm. The algorithm uses the heat conduction equation based on Fourier's law, combined with boundary conditions and initial conditions, to calculate the heat flux density of each grid cell.

[0039] For example, if the temperature deviation in a certain area is 12°C, the algorithm will recalculate the location and intensity of the heat source based on the heat flux density distribution. Through optimization algorithms such as genetic algorithms or particle swarm optimization algorithms, the system can determine the optimal heating parameters.

[0040] For example, optimized heating parameters might include adjusting the power of a heat source from 1000W to 1200W, or moving the heat source's position from coordinates (x1, y1) to (x2, y2). These optimized parameters are updated in real time to the control system, ensuring a more uniform temperature distribution within the temperature field, with deviations within 10°C. This entire process is automated, eliminating the need for human intervention and ensuring accurate and efficient temperature field control.

[0041] Step S106: According to the optimized heating parameters, a closed-loop control system is used to adjust the operating state of the annealing furnace, obtain the softening degree data of the material after annealing, and determine the softening consistency.

[0042] Acquire the real-time operating status data of the annealing furnace, wherein the operating status data includes temperature and heating power. Determine the deviation between the operating status data and the preset optimized heating parameters. If the deviation exceeds the preset threshold, the heating power of the annealing furnace is adjusted through the closed-loop control system to obtain a stable operating state. According to the stable operating state, obtain the softening degree data of the material after annealing, and judge the integrity and accuracy of the softening degree data. Use the K-means clustering algorithm to classify the softening degree data to obtain the distribution characteristics of the softening degree. If the distribution characteristics do not meet the preset uniformity standards, adjust the heating parameters through the parameter optimization algorithm to obtain an optimized heating scheme. According to the optimized heating scheme, update the control instructions of the closed-loop control system to obtain new annealing furnace operating status data. Through consistency judgment, analyze the degree of matching between the softening degree data under the new operating state and the target uniformity standard to determine the softening consistency.

[0043] Specifically, in the optimized heating parameters, the temperature of the annealing furnace is set to 750°C, the holding time is 2 hours, and the heating rate is 10°C / min. The closed-loop control system monitors the temperature in the furnace in real time and uses the PID control algorithm to adjust the heating power to ensure that the temperature fluctuation is within the range of ±5°C. The system collects temperature data through thermocouples, samples every 10 seconds, and inputs the data into the control algorithm to calculate the deviation between the current temperature and the target temperature and adjust the output power of the heating element. After annealing is completed, the system automatically collects the hardness data of the material and uses a Rockwell hardness tester for measurement. 10 samples are randomly selected for each batch, and the measurement points are distributed at the center and edge of the material. The collected hardness data is processed by statistical analysis software to calculate the average and standard deviation to determine the softening consistency.

[0044] For example, the average hardness of a batch of material is HRC25, with a standard deviation of 0.8, indicating relatively uniform softening. If the standard deviation exceeds 1.5, the system will automatically trigger an alarm, indicating a possible process anomaly. The system also compares the hardness data with historical data and uses regression analysis algorithms to predict softening trends for future batches, providing data support for process optimization. This combination of closed-loop control and data analysis ensures the stability of the annealing process and the consistency of material properties.

[0045] Step S107 , collecting surface morphology data of the thin-walled lock shell after annealing by an optical scanner, analyzing cracking and wrinkling defects by an image processing algorithm, and obtaining defect distribution characteristics.

[0046] An optical scanner is used to acquire surface topography data of the annealed thin-walled lock housing to generate a first topography image. An image denoising algorithm is used to process the first topography image to obtain a second topography image. If the pixel grayscale value of the second topography image exceeds a preset threshold value T1, it is marked as a potential crack defect area, and a crack defect candidate set is obtained. An edge detection algorithm is used to extract the contour features of the crack defect from the crack defect candidate set, and the crack defect distribution is determined. If the pixel gradient change rate of the second topography image exceeds a preset threshold value T2, it is marked as a potential wrinkling defect area, and a wrinkling defect candidate set is obtained. The wrinkling defect candidate set is processed through morphological analysis to extract the texture features of the wrinkling defects and determine the wrinkling defect distribution. A clustering algorithm is used to fuse the crack defect distribution and the wrinkling defect distribution to obtain a comprehensive defect distribution feature.

[0047] Specifically, when collecting the surface morphology data of the thin-walled lock shell after annealing by an optical scanner, a line laser scanner with a resolution of 10μm is used to obtain surface point cloud data at a scanning speed of 200mm / s. The point cloud density is set to 1000 points per square millimeter to ensure that micron-level defect features can be captured. After the acquisition is completed, a pre-processing algorithm based on Gaussian filtering is used to reduce the noise of the point cloud data. The Gaussian kernel size is set to 3×3 pixels, and the standard deviation σ=0.5, which effectively eliminates random noise during the scanning process. In the defect detection stage, a multi-scale morphological analysis method is used. First, a circular structuring element with a radius of 50μm is used for opening operation to separate the background, and then a top hat transformation is performed through a structuring element with a radius of 20μm to enhance the contrast between the cracked and wrinkled areas. For rupture defects, an algorithm based on curvature analysis is used, and the calculated surface curvature threshold is set to 0.05mm -1 When the local curvature exceeds the threshold, it is marked as a broken area. For wrinkling defects, Fourier transform is applied to extract surface corrugation features, and the spatial frequency is analyzed in the range of 0.1-0.5mm. -1 Energy distribution within the range, the area where the energy exceeds the background value by 30% is judged as wrinkling. The defect distribution feature analysis uses the DBSCAN clustering algorithm, setting the neighborhood radius to 100μm and the minimum number of samples to 5. The detected defect points are clustered into different areas, and the geometric features such as area and perimeter of each area are counted. For example, the average area of ​​the cracked area is 0.2mm 2 The average wavelength of the wrinkled area is 0.3 mm. Finally, a Kriging interpolation algorithm is used to generate a defect distribution heat map. The interpolation grid size is set to 50 μm × 50 μm, which intuitively displays the spatial distribution of defects.

[0048] Step S108: If the proportion of the cracked area in the defect distribution characteristics exceeds 2%, the forming failure probability is predicted by the neural network model to determine whether to adjust the stamping speed.

[0049] Acquire the stamping equipment parameters and real-time data acquisition results to obtain initial production process monitoring data. Extract the rupture area ratio from the initial production process monitoring data. If the rupture area ratio exceeds the preset threshold, predict it through a pre-established neural network model to obtain the probability of forming failure. According to the comparison between the forming failure probability and the quality control standard, determine whether to trigger the stamping speed adjustment. If the stamping speed adjustment is triggered, determine the adjusted stamping speed value according to the stamping equipment parameters and the real-time data acquisition results. Update the stamping equipment parameters with the adjusted stamping speed value to obtain new production process monitoring data. Extract the defect distribution characteristics from the new production process monitoring data to determine whether the rupture area ratio still exceeds the preset threshold. If the rupture area ratio still exceeds the preset threshold, optimize the neural network model through the data processing process to obtain an updated neural network model. Predict it through the updated neural network model to obtain an updated forming failure probability. According to the updated forming failure probability, cyclically adjust the stamping equipment parameters to determine the final production process monitoring data.

[0050] Specifically, during the stamping process, the surface image of the workpiece is collected in real time through a high-precision visual inspection system, and the pixel area of ​​the defective area is calculated using a contour detection algorithm based on OpenCV. If the crack area in a batch of workpieces reaches 2.3% (the threshold is set to 2%), the prediction process is triggered. The 12-dimensional features including stamping speed (1.2m / s), material thickness (0.8mm), and mold temperature (150°C) are input into the pre-trained BP neural network model. The network contains 3 hidden layers (the number of nodes is 64, 32, and 16 respectively). The ReLU activation function and Adam optimizer (learning rate 0.001) are used, and the output layer obtains a failure probability of 0.72 through the Sigmoid function. When the probability exceeds the warning line of 0.7, the control module automatically generates a speed adjustment plan: the quadratic polynomial regression equation y=0.15x fitted according to historical data 2 The optimal stamping speed was calculated to be -1.8x + 5.6 (x is the speed change rate), and the adjustment instruction was transmitted to the servo motor via the PLC. The process parameter change was also recorded in the MES system. A sliding window mechanism (window size of 50 samples) was used to continuously monitor the defect rate changes during the process, ensuring that the actual crack area ratio after adjustment remained stable below 1.8%.

[0051] Step S109 , based on the defect distribution characteristics and softening consistency data, a data fusion algorithm is used to comprehensively evaluate the forming stability to obtain the optimized operating parameters of the continuous stamping production line.

[0052] Real-time data of the continuous stamping production line during operation is acquired, and the real-time data is collected by sensors to obtain an original data set. The principal component analysis algorithm is used to perform dimensionality reduction processing on the original data set, and the main components of the defect distribution characteristics and softening consistency data are extracted to determine the feature vector set. If the variance contribution rate of the feature vector set is greater than the preset threshold, the defect distribution characteristics and the softening consistency data are integrated by the weighted fusion algorithm to obtain a forming stability evaluation index. According to the forming stability evaluation index, the support vector machine algorithm is used to classify the operating status of the continuous stamping production line to determine the operating status category. If the operating status category is a non-optimized state, the operating parameters are iteratively adjusted by the gradient descent algorithm to determine the optimized operating parameter scheme. According to the optimized operating parameter scheme, the operating parameters of the continuous stamping production line are adjusted by the real-time feedback mechanism to obtain updated production line operation data. According to the updated production line operation data, the data acquisition and processing process is repeated to determine whether the forming stability meets the preset standard.

[0053] Specifically, during the stamping process, a high-precision laser scanner was first used to collect surface defect distribution data for the sheet metal. For example, a microcrack with a depth of 0.12 mm was detected in the 25-30 mm area in the Y direction of a batch of sheet metal. A Gaussian kernel density estimation algorithm (with a bandwidth coefficient set to 0.8) was used to generate a defect heat map, and a feature matrix was established by combining the softening consistency test data (such as the hardness value fluctuation range of ±5 HB). The defect distribution characteristics and material performance data were integrated using the DS evidence theory, setting the basic probability distribution function m1 (defect) = 0.65 and m2 (softening) = 0.35. The combined formula calculated the comprehensive instability probability to be 0.72. Based on this result, a genetic algorithm was used to optimize the stamping parameters, with a population size of 50. After 100 iterations, the optimal solution was obtained: the stamping speed was adjusted from 12 times / minute to 10.5 times / minute, the blank holder force was increased from 150kN to 165kN, and the die gap was reduced by 0.03mm. Finite element simulation verified that the optimized forming limit diagram (FLD) showed an 8.3% reduction in strain in key areas, with springback controlled to within 0.05mm. Finally, the optimized parameters were programmed into the PLC control system to enable adaptive adjustment of the stamping line. After 300 consecutive stamping tests, the product qualification rate increased from 92% to 97.5%, and the vibration amplitude of the equipment decreased by 15%.

[0054] The present invention uses real-time monitoring and intelligent control: real-time data is collected through sensors, infrared thermometers, and optical scanners, and combined with finite element analysis, PID control, neural network models, etc. to achieve dynamic adjustment of mold gap, annealing temperature, and stamping parameters. This belongs to process optimization and intelligent decision-making in intelligent manufacturing. Data-driven process optimization: a data fusion algorithm is used to comprehensively evaluate forming stability, which meets the technical characteristics of "industrial big data analysis" in intelligent manufacturing. The technology of the present invention solves the problems of work hardening, mold adaptation, and annealing accuracy in the stamping of high-hardness alloy thin-walled parts through intelligent process control and data-driven dynamic optimization. It belongs to the field of intelligent manufacturing. At the same time, as an advanced process technology in high-end equipment manufacturing, its core innovation highlights the characteristics of intelligent control.

[0055] The above is only a preferred embodiment of the present invention. It should be pointed out that ordinary technicians in this technical field can make several improvements and supplements without departing from the principles of the present invention. These improvements and supplements should also be regarded as the scope of protection of the present invention.

Claims

1. A continuous stamping method for a high-hardness alloy thin-wall lock shell, characterized in that: The method comprises the following steps: Step S101: collecting stress distribution data of a high-hardness alloy thin-walled lock shell during continuous stamping in real time through a sensor, calculating the degree of stress accumulation caused by work hardening using a finite element analysis model, and obtaining a stress distribution diagram; Step S102: Based on the stress distribution diagram, an adaptive algorithm is used to adjust the mold gap, dynamically optimize the gap parameters for the local stress concentration area, and determine the real-time gap configuration; Step S103: If the real-time gap configuration deviates from the preset threshold by more than 5%, the mold adjustment mechanism is driven by the servo motor to obtain the adjusted mold gap data and determine the gap adaptability; Step S104: obtaining the real-time temperature distribution of the thin-walled lock shell during the annealing process by using an infrared thermometer, and adjusting the heat source output of the annealing furnace using a PID control algorithm to obtain a uniform temperature field; Step S105: If the temperature deviation of any area in the temperature field exceeds 10°C, the heat source distribution is recalculated using a heat flow simulation algorithm to determine the optimized heating parameters; Step S106: According to the optimized heating parameters, a closed-loop control system is used to adjust the operating state of the annealing furnace, obtain the softening degree data of the material after annealing, and determine the softening consistency; Step S107: collecting surface morphology data of the thin-walled lock shell after annealing by an optical scanner, analyzing cracking and wrinkling defects using an image processing algorithm, and obtaining defect distribution characteristics; Step S108: If the crack area in the defect distribution characteristics exceeds 2%, the forming failure probability is predicted by the neural network model to determine whether to adjust the stamping speed; Step S109: Based on the defect distribution characteristics and softening consistency data, a data fusion algorithm is used to comprehensively evaluate the forming stability to obtain the optimized operating parameters of the continuous stamping production line.

2. The continuous stamping method for a high-hardness alloy thin-wall lock shell according to claim 1, characterized in that: The step S101 includes: The sensor collects the real-time stress data of the high-hardness alloy thin-wall lock shell during continuous stamping to obtain the original stress data set; Using a data preprocessing method to denoise and standardize the original stress dataset to obtain a processed stress dataset; If there are outliers in the processed stress data set, the outliers are removed by a median filtering algorithm to obtain a filtered stress data set; Calculating the filtered stress data set using a finite element analysis model to determine a stress accumulation value caused by work hardening; Comparing the stress accumulation value with a preset stress threshold, if the stress accumulation value exceeds the threshold, marking it as a high stress area, and obtaining a stress distribution feature; Mapping the stress distribution characteristics into a two-dimensional stress distribution map using a visualization algorithm to generate a stress distribution image; The stress distribution image is contrast-adjusted by image enhancement technology to obtain an optimized stress distribution map.

3. The continuous stamping method for a high-hardness alloy thin-wall lock shell according to claim 1, characterized in that: The step S102 includes: obtaining an initial stress distribution map, wherein the initial stress distribution map includes data of a plurality of stress concentration areas; Using an adaptive algorithm to mesh the initial stress distribution map, extracting stress gradient characteristics of the local stress concentration area, and obtaining a stress gradient distribution; Iteratively calculating the stress gradient distribution using a finite element analysis model to determine the mold gap adjustment amount and obtain a gap adjustment parameter; If the gap adjustment parameter exceeds a preset gap threshold, the gap adjustment parameter is smoothed by a linear interpolation algorithm to obtain a smoothed gap parameter; Generating real-time gap configuration data according to the smooth gap parameters, mapping the data into mold control instructions, and obtaining a gap control instruction set; Transmitting the gap control instruction set to the mold control system via a real-time data interface, dynamically updating the gap configuration, and obtaining an updated gap state; Obtaining stress distribution data corresponding to the updated gap state, performing comparative analysis with the initial stress distribution diagram, and determining a change trend of the stress concentration area; If the change trend shows that the stress concentration area has not decreased, the weight parameters of the adaptive algorithm are adjusted and the stress gradient distribution is recalculated.

4. A continuous stamping method for a high-hardness alloy thin-wall lock housing according to any one of claims 1 to 3, characterized in that: The step S103 includes: Obtain the current gap data and preset gap threshold of the mold mechanism; If the deviation between the gap data and the preset threshold exceeds a limited range, a deviation value is generated by a deviation calculation module to obtain deviation data; Based on the deviation data, a control algorithm is used to generate a control instruction for the servo motor and determine the motor drive parameters; Driving the servo motor using the motor drive parameters to adjust the mold mechanism and obtain adjusted gap data; If the deviation between the adjusted gap data and the preset threshold is within a limited range, a feedback signal is obtained through a data acquisition module to determine the gap adaptability; If the feedback signal indicates that the gap adaptability is insufficient, the PID algorithm is used to optimize the control instructions to obtain optimized drive parameters; Re-driving the servo motor using the optimized drive parameters, adjusting the mold mechanism, collecting new gap data, and determining the adjustment accuracy; If the adjustment accuracy meets the preset requirements, the final gap data is recorded by the data processing module to determine that the mold adjustment is completed.

5. A continuous stamping method for a high-hardness alloy thin-wall lock housing according to any one of claims 1 to 3, characterized in that: The step S104 includes: The surface temperature of the thin-walled lock shell is collected by an infrared thermometer to generate real-time temperature distribution data; If there is an area in the temperature distribution data where the deviation exceeds a preset threshold, a filtering algorithm is used to process the temperature distribution data to obtain smoothed temperature distribution data; Calculating the deviation between the temperature of each region and the target temperature based on the smoothed temperature distribution data to generate a deviation matrix; Using a PID control algorithm, the output power of the annealing furnace heat source is adjusted according to the deviation matrix to obtain an optimized heat source distribution; The annealing furnace is driven to operate by optimizing the heat source distribution to generate new temperature field data; If the temperature deviation in the new temperature field data is lower than a preset threshold, the temperature field is determined to be uniform and stable process parameters are output; According to the stable process parameters, the operating state of the annealing furnace is adjusted to obtain a continuous and uniform temperature field.

6. A continuous stamping method for a high-hardness alloy thin-wall lock housing according to any one of claims 1 to 3, characterized in that: The step S105 includes: Obtain the current temperature field distribution from the real-time temperature monitoring data and obtain the regional temperature deviation value; If the temperature deviation of any area exceeds the preset temperature deviation threshold, the temperature field distribution is analyzed by the heat flow transfer model, and the heat source distribution is calculated using the heat flow simulation algorithm to obtain a preliminary heat source distribution result; According to the preliminary heat source distribution result and the parameter optimization target, the heat source position is adjusted to obtain optimized heat source position data; If the deviation between the optimized heat source position data and the current heat source position exceeds a preset heat source position adjustment threshold, the input parameters of the heat flow simulation algorithm are recalculated to obtain an updated heat source distribution; Determine the optimized heating parameters through the updated heat source distribution and obtain a heating parameter adjustment plan; According to the heating parameter adjustment scheme and in combination with the area division method, the temperature field distribution is recalculated to obtain new temperature field distribution data; If the temperature deviation of any area in the new temperature field distribution data still exceeds the preset temperature deviation threshold, the number of algorithm iterations is increased, and the heat flow simulation algorithm is repeatedly executed to obtain the final optimized heating parameters.

7. A continuous stamping method for a high-hardness alloy thin-wall lock housing according to any one of claims 1 to 3, characterized in that: The step S106 includes: Acquiring real-time operating status data of the annealing furnace, wherein the operating status data includes temperature and heating power; determining a deviation between the operating status data and preset optimized heating parameters; If the deviation exceeds a preset threshold, the heating power of the annealing furnace is adjusted through a closed-loop control system to obtain a stable operating state; According to the stable operating state, obtaining softening degree data of the material after annealing, and judging the completeness and accuracy of the softening degree data; The softening degree data are classified using a K-means clustering algorithm to obtain a distribution characteristic of the softening degree; If the distribution characteristics do not meet the preset uniformity standard, the heating parameters are adjusted through a parameter optimization algorithm to obtain an optimized heating solution; According to the optimized heating scheme, the control instructions of the closed-loop control system are updated to obtain new annealing furnace operation status data; Through consistency judgment, the matching degree between the softening degree data under the new operating state and the target uniformity standard is analyzed to determine the softening consistency.

8. A continuous stamping method for a high-hardness alloy thin-wall lock housing according to any one of claims 1 to 3, characterized in that: The step S107 includes: Acquiring surface topography data of the thin-walled lock shell after annealing by an optical scanner to generate a first topography image; Processing the first topographic image using an image denoising algorithm to obtain a second topographic image; If the pixel grayscale value of the second topographic image exceeds a preset threshold value T1, it is marked as a potential crack defect area, and a crack defect candidate set is obtained; For the crack defect candidate set, an edge detection algorithm is used to extract the contour features of the crack defects and determine the crack defect distribution; If the pixel gradient change rate of the second topographic image exceeds a preset threshold value T2, it is marked as a potential wrinkle defect area, and a wrinkle defect candidate set is obtained; Processing the wrinkle defect candidate set through morphological analysis, extracting texture features of the wrinkle defects, and determining the wrinkle defect distribution; A clustering algorithm is used to fuse the rupture defect distribution and the wrinkling defect distribution to obtain a comprehensive defect distribution feature.

9. A continuous stamping method for a high-hardness alloy thin-wall lock housing according to any one of claims 1 to 3, characterized in that: The step S108 includes: Obtain stamping equipment parameters and real-time data collection results to obtain initial production process monitoring data; Extracting the fracture area ratio from the initial production process monitoring data; If the proportion of the fracture area exceeds a preset threshold, a pre-established neural network model is used to predict the probability of forming failure; Determining whether to trigger a stamping speed adjustment based on a comparison of the molding failure probability and a quality control standard; If the stamping speed adjustment is triggered, the adjusted stamping speed value is determined according to the stamping equipment parameters and the real-time data acquisition results; Using the adjusted stamping speed value to update the stamping equipment parameters to obtain new production process monitoring data; Extracting defect distribution characteristics from the new production process monitoring data to determine whether the crack area ratio still exceeds the preset threshold; If the rupture area ratio still exceeds the preset threshold, optimizing the neural network model through a data processing process to obtain an updated neural network model; Perform prediction using the updated neural network model to obtain an updated molding failure probability; According to the updated forming failure probability, the stamping equipment parameters are cyclically adjusted to determine the final production process monitoring data.

Images