Stamping process optimization method for seat backrest connecting bracket

Abstract

The application discloses a stamping processing process optimization method for a seat backrest connecting support, relates to the technical field of stamping processing process optimization, collects real-time heat treatment process data in a heat treatment process of the support, combines environment condition data, uses a trained coupling model of phase transition and grain growth to predict corresponding microstructure, acquires structure prediction data, predicts performance data of the support according to heat treatment process parameters and microstructure characteristics of the material, generates performance values from the performance data, if the performance values do not exceed a performance threshold, uses a pre-trained genetic algorithm to optimize the current heat treatment process parameters, and executes the optimization by an automatic control module of a heat treatment device, real-time predicts microstructure characteristics, and iteratively optimizes the heat treatment process parameters when the deviation degree exceeds the expectation; the historical optimization scheme and experience are called at any time, the response speed of production decision is improved, and it is ensured that the heat treatment process is always in an optimal state.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of stamping process optimization, in particular to a stamping process optimization method for a seat backrest connecting bracket. BACKGROUND

[0002] With the rapid development of the automotive industry, higher requirements are placed on the design and manufacture of seat components, especially seat backrest connecting brackets. As a key load-bearing component in the seat system, the performance of the seat connecting bracket directly affects the comfort, safety, and durability of the seat. Therefore, how to ensure the high strength, good toughness, and reliable long-term use performance of the bracket through precise processing technology has become a problem that needs to be solved in the manufacturing process.

[0003] Traditional stamping processing technology often relies on experience and usually does not monitor and intervene in real-time environmental factors during the processing, which makes the product quality greatly affected by environmental changes. Especially in the stamping process, factors such as environmental temperature and humidity, dust in the air, etc. may pose potential risks to forming accuracy and product quality. In the heat treatment link, the microstructure and performance of the material are also significantly affected by process parameters such as temperature, heating time, and cooling rate. Therefore, combining real-time environmental monitoring, microstructure prediction, and process optimization has become a necessary means to improve the manufacturing quality of seat connecting brackets.

[0004] In the Chinese invention patent with the authorization announcement number CN109332482B, a kind of accumulator membrane box membrane piece optimization stamping forming process method is disclosed, including the following steps: the optimization of stamping blank, improve the flatness of blank;Stamping process optimization, forming twice;Determine reasonable die springback value.Before the part processing, the bending angle of the two stages of blanking-preforming and final forming parts is reasonably distributed. Before forming processing, simulation calculation can be carried out using finite element analysis tool, and the distribution value is obtained. Then carry out die test verification, evaluate the test results, and appropriately correct the die size. Through a series of iterative analysis, the structure parameters of the die can be finally determined, mainly the springback angle value of bending. Through the corrected springback, the final size of the membrane piece part can be guaranteed.

[0005] During the heat treatment process of the bracket, especially in critical process stages such as quenching and tempering, different heat treatment parameters such as heating temperature, heating time, cooling rate, etc. will significantly affect the microstructure changes of the material. Specifically, during the quenching process, the phase transition, grain growth and martensite formation of the material at different temperatures will directly affect the strength and hardness of the material, while the tempering process optimizes the toughness and internal stress distribution of the material by controlling the combination of temperature and time. However, due to the complex changes in the microstructure of the material caused by changes in these parameters, it often causes large fluctuations in the mechanical properties of the final product. Existing technical methods cannot effectively achieve precise control of these heat treatment processes, especially cannot adjust the heat treatment parameters in real time combined with environmental conditions, so that the mechanical properties of the product cannot stably reach the expected value.

[0006] To this end, the present application provides a stamping process optimization method for a seat backrest connecting bracket. SUMMARY

[0007] (1) Technical problems solved

[0008] In view of the deficiencies of the prior art, the present application provides a stamping process optimization method for a seat backrest connecting bracket, which acquires real-time heat treatment process data during the heat treatment process of the bracket and combines environmental condition data, uses a trained coupling model of phase transition and grain growth to predict the corresponding microstructure, obtains structure prediction data, and predicts the performance data of the bracket according to the heat treatment process parameters and the microstructure characteristics of the material, generates performance values from the performance data, and if the performance values do not exceed the performance threshold, uses a pre-trained genetic algorithm to optimize the current heat treatment process parameters, and executes them by the automatic control module of the heat treatment equipment, real-time predicts the microstructure characteristics, and iteratively optimizes the heat treatment process parameters when their deviation degree exceeds the expectation; historical optimization schemes and experience are called at any time to improve the response speed of production decision-making and ensure that the heat treatment process is always in the optimal state; thereby solving the technical problems recorded in the background art.

[0009] (2) Technical solutions

[0010] To achieve the above purpose, the present application is implemented by the following technical solutions: a stamping process optimization method for a seat backrest connecting bracket, comprising, real-time monitoring the environmental condition data in the processing area and performing abnormality detection, constructing an environmental abnormality degree Aop from the obtained abnormality detection data, and if the environmental abnormality degree Aop exceeds the expectation, issuing a monitoring instruction to the outside;

[0011] After acquiring real-time heat treatment process data during the heat treatment process of the bracket and combining environmental condition data, a trained coupling model of phase transition and grain growth is used to predict the corresponding microstructure, and structure prediction data is obtained;

[0012] According to the heat treatment process parameters and the microstructure characteristics of the material, the performance data of the bracket is predicted, the performance value Pot is generated from the performance data, and if the performance value Pot does not exceed the performance threshold, a control optimization instruction is sent to the outside;

[0013] The current heat treatment process parameters are optimized by using a pre-trained genetic algorithm, and the microstructure characteristics are predicted in real time by the automatic control module of the heat treatment equipment. When the deviation degree exceeds the expectation, the heat treatment process parameters are iteratively optimized;

[0014] The performance of the processed bracket is tested, and the test data is used as feedback to output process optimization suggestions from the bracket heat treatment process optimization knowledge graph.

[0015] Further, the sensor network monitors the changes of environmental conditions in the processing area in real time and obtains corresponding environmental condition data; using the trained anomaly detection model for anomaly detection, and obtaining corresponding anomaly detection data.

[0016] Further, the environmental anomaly degree Aop is constructed from the anomaly detection data as follows:

[0017]

[0018] In the formula: Ω is the integral region, φ(x, y, t) is the abnormal distribution function, is the partial derivative of the abnormal distribution function with respect to time; α is the diffusion coefficient.

[0019] Further, real-time monitoring of temperature change data during heating and cooling, receiving monitoring instructions, using online X-ray diffraction equipment to monitor and collect material grain size, phase change microstructure changes of materials in the heat treatment process, and obtaining real-time heat treatment process data.

[0020] Further, collect microstructure characteristic data of materials under different heat treatment parameters, combine it with real-time heat treatment process data and environmental condition data in the processing area, and obtain labeled sample data after preprocessing and labeling.

[0021] Further, using the heat treatment process parameters and the microstructure characteristics of the material as input, using the trained mechanical property prediction model to predict the mechanical properties of the material, obtaining the corresponding mechanical property data and generating the performance feature data set from the performance feature data in the performance feature data set.

[0022] Further, after receiving the control optimization instruction, taking improving the performance value as the optimization goal, using a pre-trained genetic algorithm to optimize the current heat treatment process parameters, and obtaining the optimized heat treatment process parameters;

[0023] The optimized heat treatment process parameters are simulated and tested using finite element analysis to obtain corresponding simulation test data.

[0024] Further, the simulation prediction result is compared with the design target value, if it is inconsistent with the expectation, the optimization is re-performed, if it is consistent with the expectation, the optimized heat treatment process parameters are executed by the automatic control module of the heat treatment equipment.

[0025] Further, the real-time predicted microstructure features are compared with the preset target value to obtain the deviation degree of each feature and the target value, the deviation degree Pok is generated from the deviation degree, if the obtained deviation degree Pok exceeds the deviation threshold, the genetic algorithm is re-called for optimization based on the deviation degree and the current heat treatment parameter, new process parameters are generated and executed, or an alarm instruction is sent to the outside.

[0026] Further, the performance of the connecting bracket after heat treatment is detected, and the obtained detection data is summarized as sample detection data; the sample detection data, environmental condition data, heat treatment process monitoring data and process parameters are integrated for feature extraction to obtain corresponding optimization features; the bracket heat treatment optimization is taken as a target word to obtain a bracket heat treatment process optimization knowledge graph.

[0027] (Three) beneficial effects

[0028] The present application provides a stamping processing process optimization method for a seat back connecting bracket, which has the following beneficial effects:

[0029] 1. Continuous monitoring of environmental changes can discover potential environmental problems, effectively prevent environmental abnormalities from affecting the processing process, improve the stability of the production process, ensure that the processing process is always within the predetermined environmental condition range, improve the consistency of product quality, and avoid quality fluctuations caused by environmental abnormal fluctuations.

[0030] 2. The machine learning algorithm is used to predict the microstructure structure in the heat treatment process, and the accurate prediction of the microstructure structure helps to optimize the heat treatment scheme, ensures that the grain size, phase composition and other factors are within the optimal range, and further optimizes the mechanical properties of the product.

[0031] 3. The prediction result provides a reliable basis for process optimization, ensures that the material reaches the ideal mechanical properties under different heat treatment conditions; through real-time prediction of microstructure features, personalized heat treatment schemes can be customized according to specific material and product requirements.

[0032] 4、When the predicted performance value Pot does not reach the preset standard, the dynamic adjustment of the heat treatment process makes the process parameters flexible to adapt to environmental changes, reduces the performance fluctuations caused by external changes; through the continuous monitoring of the performance value Pot, timely adjustment of the heat treatment process can ensure that all performance indicators of the product are always in the ideal range, thereby reducing the production of unqualified products and improving the production efficiency and product qualification rate.

[0033] 5、Through the genetic algorithm to optimize the process parameters, determine the optimal parameter combination (such as temperature, time, cooling rate, etc.) in the heat treatment process, improve the optimization efficiency of the heat treatment process, ensure that the optimized parameters can achieve the expected effect in actual production, reduce the production risk.

[0034] 6、Through the comparison of microstructure characteristics and preset target value, any deviation in the process can be monitored in real time, and automatic adjustment can be made to ensure that the production process always remains on the predetermined optimal path, improving the adaptability and flexibility of the process; continuously optimize the process parameters according to the adjustment results of each deviation to form a closed-loop improvement mechanism, which can improve the stability and consistency of the production process, reduce the production of defective products caused by production fluctuations, and ultimately improve the overall production efficiency.

[0035] 7、According to the optimization characteristics, automatically generate process optimization suggestions, and at any time retrieve historical optimization schemes and experience to improve the response speed of production decision-making; through the automatically generated process optimization suggestions, the process optimization can be quickly executed to ensure that the heat treatment process is always in the optimal state, improving the flexibility and adaptability of the production process. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 The figure is a flowchart of the stamping process optimization method of the application. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0038] Please refer to Figure 1 The application provides a stamping process optimization method for a seat backrest connecting bracket, which comprises,

[0039] Step one, after real-time monitoring of the environmental condition data in the processing area, abnormality detection is performed, the environmental abnormality degree Aop is constructed from the obtained abnormality detection data, and if the environmental abnormality degree Aop exceeds the expectation, a monitoring instruction is sent to the outside.

[0040] The step one includes the following contents:

[0041] Step 101, after determining the support stamping processing area, arrange a sensor network in the processing area, such as temperature sensor, humidity sensor, dust sensor and the like, and real-time monitor the change of environmental conditions in the processing area by the sensor network, and obtain the corresponding environmental condition data, such as temperature, humidity and dust amount in the air and the like;

[0042] Train the machine learning algorithm by the labeled sample data, obtain the trained anomaly detection model, use the trained anomaly detection model for anomaly detection with the environmental condition data as input, and obtain the corresponding anomaly detection data, such as abnormal area, abnormal time and abnormal degree and the like;

[0043] In use, through the arrangement of temperature sensor, humidity sensor and dust sensor and the like, comprehensive and real-time environmental monitoring can be realized. Continuous monitoring of environmental changes can discover potential environmental problems, which can effectively prevent the adverse effects of environmental anomalies on the processing process, improve the stability of the production process, help identify the impact of environmental anomalies on processing, and effectively prevent damage to production equipment caused by these anomalies, especially in the precision machining link;

[0044] Step 102, under the dimensionless condition, construct the environmental anomaly degree Aop from the anomaly detection data, in the following way:

[0045]

[0046] In the formula: Ω is the integral region, and φ(x, y, t) is the abnormal distribution function, which describes the environmental anomaly degree in the processing area (x, y) and at time t, is the partial derivative of the abnormal distribution function with respect to time; α is the diffusion coefficient, which is between 0 and 1, and are the second-order partial derivatives of the abnormal distribution function with respect to spatial coordinates x and y, respectively, which describe the spatial dispersion behavior of the abnormal degree; Laplacian operator, used to describe the diffusion characteristics of the abnormal degree in the whole space;

[0047] According to the historical data and the management expectation of the environmental conditions, set the abnormal threshold value in advance; if the obtained environmental anomaly degree Aop exceeds the abnormal threshold value, it means that the environmental anomaly degree in the processing area is high, which may affect the subsequent processing process, and needs to be targeted. At this time, send a monitoring instruction to the outside;

[0048] In use, combined with the contents in steps 101 and 102:

[0049] Real-time monitoring of environmental data changes and timely detection of environmental conditions can avoid the negative impact of environmental abnormalities on processing quality, improve product stability and consistency; ensure that the processing process is always within the predetermined environmental condition range, which can improve the consistency of product quality and avoid quality fluctuations caused by environmental abnormal fluctuations.

[0050] Step two, after collecting real-time heat treatment process data during the heat treatment process of the stent, combining with environmental condition data, using the trained phase transition and grain growth coupling model to predict the corresponding microstructure, and obtaining structure prediction data;

[0051] The step two includes the following contents:

[0052] Step 201, install temperature sensors at key positions in the tempering and quenching region of the stent, for example, in the heating and cooling zones, to monitor the temperature change data in the heating and cooling process in real time and record the time of each stage of heating, holding and cooling;

[0053] After receiving the monitoring instruction, use the online X-ray diffraction equipment to monitor and collect the phase change of the material in the heat treatment process, such as the process of austenite transforming into ferrite and pearlite, monitor the microstructure changes of the material such as grain size and phase transition, and obtain real-time heat treatment process data;

[0054] In use, by monitoring the heating and cooling temperature changes in the heat treatment process in real time, the temperature of each stage of heating, holding and cooling is accurately controlled, ensuring that the material heat treatment process is strictly in accordance with the process requirements, avoiding the microstructure unevenness caused by temperature fluctuations in the heat treatment process. It can also achieve rapid adjustment during heating and cooling, improve the adaptability of the heat treatment process, and thus improve the quality and performance of the final product.

[0055] Step 202, collect microstructure feature data of the material under different heat treatment parameters (temperature, time, cooling rate), including grain size distribution, phase composition ratio and martensite content, etc., combine it with real-time heat treatment process data and environmental condition data in the processing area, and obtain labeled sample data after preprocessing and labeling;

[0056] Based on the phase transition model (such as ferrite-pearlite transformation model and martensite transformation model) and grain growth model combined with machine learning algorithm; after training from labeled sample data, obtain the trained phase transition and grain growth coupling model; using the trained phase transition and grain growth coupling model to predict the corresponding microstructure with heat treatment process data as input, and obtain structure prediction data;

[0057] For example, grain size: according to the Hall-Petch relationship, the grain size directly affects the yield strength of the material; phase composition: different phases (such as pearlite, martensite, bainite, etc.) have different effects on hardness, strength and toughness; martensite content: in steel, the increase of martensite content usually increases the hardness and strength; carbon content, distribution of alloying elements: the type and distribution of alloying elements also affect the mechanical properties of the material, especially in high-alloy steel;

[0058] In use, the contents in steps 201 and 202 are combined:

[0059] Using machine learning algorithms to predict the microstructure during heat treatment, and understanding the effects of heat treatment on the microstructure of the material, such as grain size and phase composition, can effectively reduce the time and cost of a large number of experimental verification in traditional experiments; accurate prediction of microstructure helps to optimize the heat treatment scheme, ensures that factors such as grain size and phase composition are within the optimal range, and further optimizes the mechanical properties of the product, such as hardness, strength and toughness.

[0060] Step three, according to the heat treatment process parameters and the microstructure characteristics of the material, the performance data of the bracket is predicted, and the performance value Pot is generated from the performance data; if the performance value Pot does not exceed the performance threshold, a control optimization instruction is sent to the outside;

[0061] The step three includes the following contents:

[0062] Step 301, combine the Hall-Petch relationship, Lüders formula and other theoretical relationship models with deep neural networks, and use labeled experimental data and other related data for training and iterative optimization to obtain a trained mechanical property prediction model;

[0063] Using the trained mechanical property prediction model, the mechanical properties of the material are predicted using the heat treatment process parameters (temperature, time, cooling rate, etc.) and the microstructure characteristics of the material (grain size, phase composition, etc.) as input, and the corresponding mechanical property data, such as strength, hardness and toughness, are obtained. After being summarized, the performance characteristic data set is generated;

[0064] By combining the Hall-Petch relationship, Lüders formula and other classical theoretical models with deep neural networks, high-precision prediction of the mechanical properties (such as strength, hardness, toughness, etc.) of the material under different heat treatment parameters is achieved, providing a reliable basis for process optimization and ensuring that the material achieves ideal mechanical properties under different heat treatment conditions; through real-time prediction of microstructure characteristics (such as grain size, phase composition, etc.), personalized heat treatment schemes can be customized for specific materials and product requirements; maximize the mechanical properties of the product to ensure that the product meets strict quality and strength requirements;

[0065] Among them, according to the effect of grain refinement on material strength, the Hall-Petch formula is used to describe the relationship between grain size and material strength:

[0066] σ y =σ0+k y d -1 / 2

[0067] Among them, σ y is the yield strength, σ0 is the intrinsic shear strength of the material, k y is the Hall-Petch constant, and d is the grain diameter; through this relationship, the grain size can be related to the strength of the material;

[0068] For martensitic transformation, empirical formulas or theoretical models are usually used to describe the relationship between its hardness and phase transformation. The common Lüders formula can be used to predict the hardness of martensitic transformation:

[0069] H=H0+k M ·(V M -V0)

[0070] Where H is the hardness, H0 is the matrix hardness, V M is the martensite content, k M is a constant, V0 is the initial martensite content. This formula can be used to predict the hardness of the material after heat treatment.

[0071] During heat treatment, the grain growth process has a significant impact on the mechanical properties of the material (such as strength and toughness). According to the classic JMAK (Johnson-Mehl-Avrami-Kolmogorov) model, the grain growth process can be described by the following equation:

[0072]

[0073] Where X is the degree of phase transition, t is time, τ is the characteristic time, and n is a constant representing the rate of grain growth. This equation can help us quantitatively predict the size of grains under different time and temperature conditions.

[0074] To establish the relationship between phase composition and toughness, an empirical formula of the following form can usually be used:

[0075] K I C=f(C,α,T)

[0076] Among them, K I C is the fracture toughness of the material, C is the carbon content, α is the phase composition parameter (such as martensite volume fraction, pearlite volume fraction), and T is the temperature;

[0077] Step 302, under the dimensionless condition, the performance value Pot is generated from the performance characteristic data in the performance characteristic data set in the following manner:

[0078]

[0079] In the formula, T is the time range of performance evaluation, F(t) is the performance characteristic vector, which contains a plurality of performance characteristic data at time t, and is defined as: f s (t) is the intensity data at time t, f h (t) is the hardness data at time t, f v (t) is the toughness data at time t; W is a weight vector, where w s , w h , and w v are the weights of intensity, hardness, and toughness, respectively, and w s +w h +w v = 1; K is a characteristic contribution matrix, where k ij represents the contribution of characteristic i to characteristic j, controlling the interaction between performance characteristics; exp(K·F(t)) is a nonlinear conversion function, representing the interaction of performance characteristics, for a specific time t r This term will capture the nonlinear coupling relationship between intensity, hardness, and toughness;

[0080] According to historical data and management expectations of the processing effect of the stent, the performance threshold is set in advance; if the obtained performance value Pot does not exceed the preset performance threshold, it means that if the current process parameters are continuously executed, the performance of the stent may not be able to achieve the expected effect, and therefore the process parameters need to be adjusted in real time, at which time the control optimization instruction is sent to the outside;

[0081] In use, the contents in steps 301 and 302 are combined:

[0082] The relationship between the process parameters and the performance value is fed back in real time, and when the predicted performance value Pot does not reach the preset standard, the dynamic adjustment of the heat treatment process is performed, the real-time performance feedback and adjustment mechanism can cope with the uncertainty in the complex production environment, so that the process parameters can flexibly adapt to environmental changes and reduce the performance fluctuations caused by external changes; through continuous monitoring of the performance value Pot, the heat treatment process can be adjusted in a timely manner, which can ensure that all performance indicators of the product are always in the ideal range, thereby reducing the generation of unqualified products and improving the production efficiency and product qualification rate.

[0083] Step four, using a pre-trained genetic algorithm to optimize the current heat treatment process parameters, and executed by the automatic control module of the heat treatment equipment, real-time prediction of microstructure characteristics, when the deviation degree exceeds the expected, iterative optimization of heat treatment process parameters;

[0084] The step four includes the following contents:

[0085] Step 401, after receiving the control optimization instruction, using the pre-trained genetic algorithm to optimize the current heat treatment process parameters (temperature, time, cooling rate, etc.) with the performance value as the optimization target, and obtaining the optimized heat treatment process parameters; using finite element analysis to simulate and test the optimized heat treatment process parameters, verifying the influence of the optimized parameters on the microstructure and mechanical properties of the material, and obtaining the corresponding simulation test data;

[0086] Through genetic algorithm to optimize the process parameters, the optimal parameter combination (such as temperature, time, cooling rate, etc.) in the heat treatment process is determined, which improves the optimization efficiency of the heat treatment process, realizes the optimal performance in a short time, avoids repeated experiments and time waste in the manual adjustment process; using finite element analysis to simulate the optimized process parameters, which can verify the influence of the optimization scheme on the microstructure and mechanical properties in advance, ensure that the optimized parameters can achieve the expected effect in actual production, and reduce the production risk.

[0087] Step 402, compare the simulation prediction results with the design target value, if they are inconsistent with the expected, re-optimize;

[0088] If they are consistent with the expected, execute the optimized heat treatment process parameters by the automatic control module of the heat treatment equipment, real-time adjust the heating rate, holding time, cooling rate, etc., realize the real-time feedback control of the heat treatment process;

[0089] When using, the performance deviation caused by optimization error can be avoided, the reliability of process execution and the qualified rate of products are improved. The optimized process parameters are adjusted in real time through the automatic control system, which speeds up the response speed of process parameter adjustment.

[0090] Step 403, compare the real-time predicted microstructure characteristics (such as grain size, phase composition, etc.) with the preset target value, obtain the deviation degree of each feature from the target value, and generate the deviation degree Pok by the deviation degree, as follows:

[0091] Pok=||W(t)·∑ -1 / 2 ·(F-T)|| p

[0092] In the formula: p=2, F is the microstructure feature vector, which contains all the feature values (such as grain size, phase composition ratio, martensite content, etc.) predicted in real time; Target feature vector, containing the preset target value of each feature, ∑ is the feature covariance matrix, capturing the correlation between features, defined as: ∑ ij = Cov(f i ,f j ); wherein Cov(f i ,f j ) represents the covariance of features f i and f j ; ∑ -1 / 2 is the inverse square root of the covariance matrix, used to normalize the correlation between features and eliminate the influence of feature dimension; W(t) = diag(w1(t), w2(t), …, w n (t)) is a dynamic weight matrix, representing the influence weight of each feature on the deviation degree, and the weight is dynamically adjusted over time;

[0093] According to historical data and the expected processing performance of the bracket, a deviation threshold is set in advance; if the obtained deviation degree Pok exceeds the deviation threshold, based on the deviation degree and the current heat treatment parameters, the genetic algorithm is called again for optimization to generate new process parameters and execute, or an alarm instruction is sent externally;

[0094] In use, the contents in steps 401 to 403 are combined:

[0095] By comparing the microstructure features (such as grain size, phase composition, etc.) with the preset target values, any deviation in the process can be monitored in real time, and automatic adjustment can be made to ensure that the production process always remains on the predetermined optimal path, improving the adaptability and flexibility of the process;

[0096] According to the continuous optimization of process parameters based on each deviation adjustment result, a closed-loop improvement mechanism is formed, which can improve the stability and consistency of the production process, reduce the generation of defective products caused by production fluctuations, and ultimately improve the overall production efficiency.

[0097] Step five, test the performance of the processed bracket, and use the test data as feedback to output process optimization suggestions from the bracket heat treatment process optimization knowledge graph;

[0098] The step five includes the following contents:

[0099] Step 501, after completing the heat treatment of the bracket in the current stage, perform performance tests such as tensile test, hardness test, impact toughness test, microstructure detection, and surface quality detection on the heat-treated connecting bracket, and aggregate the obtained detection data as sample detection data; integrate the sample detection data, environmental condition data, heat treatment process monitoring data, and process parameters, and then perform feature extraction to obtain the corresponding optimization features;

[0100] In use, by integrating test data with heat treatment process monitoring data, environmental data, etc., and feature extraction, reference data is provided. Data integration not only improves the accuracy of process optimization, but also quickly identifies key factors affecting performance.

[0101] Step 502, taking the heat treatment optimization of the stent as the target word, after deep search and entity relationship building, the stent heat treatment process optimization knowledge graph is obtained; taking the optimization features as input, according to the correspondence between the optimization features and the process optimization suggestions, output the process optimization suggestions from the stent heat treatment process optimization knowledge graph, execute the process optimization suggestions to optimize the current heat treatment process, and obtain the optimized process;

[0102] In use, in combination with the contents in steps 501 and 502:

[0103] According to the optimization features, the process optimization suggestions are automatically generated, the historical optimization schemes and experience are recalled at any time, the response speed of production decision is improved; through the automatically generated process optimization suggestions, the process optimization can be quickly executed, the heat treatment process is always in the optimal state, and the flexibility and adaptability of the production process are improved.

[0104] It should be noted that: the construction method of the stent heat treatment process optimization knowledge graph takes data-driven and knowledge fusion as the core, by integrating multi-source heterogeneous data (such as experimental data, literature data, industry specifications and production history records) and the process experience of field experts, the systematic modeling and semantic expression of heat treatment process are realized.

[0105] Firstly, natural language processing (NLP) technology is used to extract key entities (such as quenching temperature, cooling rate, holding time, grain size, phase composition, etc.) and their associated relationships (such as causal relationship, parameter dependency) from literature and technical documents. Secondly, through multivariate data analysis, experimental data, real-time monitoring data and historical records are extracted and associated, and the potential relationship between data is supplemented. Then, using knowledge graph construction tools (such as RDF or Neo4j), entities and relationships are modeled as nodes and edges, and semantic labels in heat treatment process are assigned, forming a structured graph database. Finally, combined with graph algorithms (such as path search, centrality analysis and graph embedding) and machine learning models, intelligent reasoning, dynamic updating and accurate generation of process optimization schemes are supported, thus forming an extensible and interactive knowledge system, providing scientific basis and intelligent support for the design, optimization and quality improvement of stent heat treatment process.

[0106] Those skilled in the art can clearly understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0107] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

[0108] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the above-described device embodiments are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0109] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0110] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any skilled person in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for optimizing the stamping process of a seat back connecting bracket, characterized in that: include, Real-time monitoring of environmental condition data in the processing area is performed to detect anomalies. The environmental anomaly degree is constructed based on the acquired anomaly detection data. If the environmental anomaly degree Exceed expectations and issue monitoring instructions to external parties; After collecting real-time heat treatment process data during the stent heat treatment process and combining it with environmental condition data, the trained phase change and grain growth coupling model is used to predict the corresponding microstructure and obtain structural prediction data; Predict the performance data of the scaffold based on the heat treatment process parameters and the microstructure characteristics of the material, and generate performance values ​​from the performance data , if the performance value If the performance threshold is not exceeded, control optimization instructions are issued to the outside world; A pre-trained genetic algorithm is used to optimize the current heat treatment process parameters, which are then executed by the automatic control module of the heat treatment equipment to predict the microstructure characteristics in real time and iteratively optimize the heat treatment process parameters when the deviation exceeds the expected level. The performance of the processed bracket is tested, and the test data is used as feedback. The bracket heat treatment process optimization knowledge map outputs process optimization suggestions.

2. The method for optimizing the stamping process of the seat back connecting bracket according to claim 1, characterized in that: The sensor network monitors the changes in environmental conditions in the processing area in real time and obtains the corresponding environmental condition data; Taking environmental condition data as input, the trained anomaly detection model is used to perform anomaly detection and obtain the corresponding anomaly detection data.

3. The method for optimizing the stamping process of the seat back connecting bracket according to claim 2, characterized in that: Constructing environmental anomaly from anomaly detection data , as follows: ； Where: is the integration area, is the abnormal distribution function, is the partial derivative of the anomaly distribution function with respect to time; is the diffusion coefficient.

4. The method for optimizing the stamping process of the seat back connecting bracket according to claim 3, characterized in that: Real-time monitoring of temperature change data during heating and cooling processes. After receiving monitoring instructions, use online X-ray diffraction equipment to monitor and collect real-time changes in the grain size, phase transformation and microstructure of the material during the heat treatment process to obtain real-time heat treatment process data.

5. The method for optimizing the stamping process of the seat back connecting bracket according to claim 4, characterized in that: The microstructural characteristic data of the material under different heat treatment parameters are collected, combined with the real-time heat treatment process data and the environmental condition data in the processing area, and the labeled sample data are obtained after preprocessing and labeling.

6. The method for optimizing the stamping process of the seat back connecting bracket according to claim 5, characterized in that: Taking heat treatment process parameters and material microstructure characteristics as input, the trained mechanical property prediction model is used to predict the mechanical properties of the material. The corresponding mechanical property data are obtained and summarized to generate a performance characteristic data set. The performance value is generated from the performance characteristic data in the performance characteristic data set. .

7. The method for optimizing the stamping process of the seat back connecting bracket according to claim 6, characterized in that: After receiving the control optimization instruction, the pre-trained genetic algorithm is used to optimize the current heat treatment process parameters with the improvement of performance value as the optimization goal, and the optimized heat treatment process parameters are obtained; Finite element analysis is used to simulate and test the optimized heat treatment process parameters to obtain corresponding simulation test data.

8. The method for optimizing the stamping process of the seat back connecting bracket according to claim 7, characterized in that: The simulation prediction results are compared with the design target values. If they are inconsistent with the expectations, they are re-optimized; if they are consistent with the expectations, the automatic control module of the heat treatment equipment will execute the optimized heat treatment process parameters.

9. The method for optimizing the stamping process of the seat back connecting bracket according to claim 8, characterized in that: Compare the real-time predicted microstructure characteristics with the preset target values ​​to obtain the degree of deviation between each characteristic and the target value, and generate the deviation degree from the deviation degree. ; If the obtained deviation If the deviation threshold is exceeded, the genetic algorithm is called again for optimization based on the deviation degree and the current heat treatment parameters, and new process parameters are generated and executed, or an alarm instruction is issued to the outside.

10. The method for optimizing the stamping process of the seat back connecting bracket according to claim 9, characterized in that: Performing performance testing on the connection bracket after heat treatment, and summarizing the obtained test data as sample test data; Integrate sample test data, environmental condition data, heat treatment process monitoring data, and process parameters to perform feature extraction and obtain corresponding optimization features; Taking bracket heat treatment optimization as the target word, the knowledge graph of bracket heat treatment process optimization is obtained.

Images