Process optimization method and system for multi-part adaptive self-adjusting stamping die

Abstract

The present application relates to the technical field of die parameter optimization, and discloses a process optimization method and system for a multi-part adaptive self-adjusting stamping die, which comprises the following steps: generating a structured process input parameter set through input parameters and storing the set in a central database; searching for relevant cases and rules from a knowledge graph according to the input parameters; calculating the similarity of the current parameters and historical cases, using a hybrid algorithm, and weighting and averaging to indicate the weight of each parameter, while processing special constraints, and finally normalizing to obtain a comprehensive similarity index. The present application realizes real-time monitoring and dynamic adjustment through an intelligent control system, combines adaptive die design optimization and centralized material property database management, and uses an integrated quality detection system to automatically identify and classify defects using image recognition and machine learning technology, effectively solving the problem of low production efficiency and unstable product quality caused by the inability to adjust key parameters in time under actual working conditions.

Description

Technical Field

[0001] This invention relates to the field of mold parameter optimization, and more specifically, to a process optimization method and system for multi-part adaptive self-adjusting stamping dies. Background Technology

[0002] As the manufacturing industry's demand for efficient and flexible production continues to increase, traditional stamping dies face numerous challenges. Traditional die design and production processes lack effective real-time monitoring and dynamic adjustment mechanisms. Die design relies on experienced engineers and lacks systematic and automated optimization tools, making it difficult to quickly respond to changes in market demands. Traditional quality inspection methods are outdated, with low accuracy and efficiency, failing to meet the requirements of modern high-precision manufacturing.

[0003] This results in the inability to adjust key parameters in a timely manner according to actual working conditions, leading to low production efficiency and unstable product quality. Summary of the Invention

[0004] This invention provides a process optimization method and system for multi-part adaptive self-adjusting stamping dies, solving technical problems in related technologies.

[0005] This invention provides a process optimization method for multi-part adaptive self-adjusting stamping dies, comprising the following steps:

[0006] S100, receiving input parameters: It receives four types of core input data through a standardized interface, and finally generates a structured process input parameter set and stores it in the central database;

[0007] S200, Knowledge Graph Retrieval: Based on the input parameters in S100, search for relevant cases and rules from the knowledge graph;

[0008] S300, Similarity Calculation: Calculate the similarity between the current parameter and historical cases, use a hybrid algorithm and weighted average, explain the weight of each parameter, handle special constraints, and finally normalize to obtain the comprehensive similarity index;

[0009] S400, Candidate Set Filtering: Cases are filtered based on similarity index and constraints. Case filtering includes threshold filtering, constraint filtering, time-sensitive weighting, and diversity control to form an optimized candidate set.

[0010] S500, Parameter Fusion: Integrates current parameters with candidate cases to generate an optimized parameter set. Parameter fusion includes main parameter inheritance, auxiliary parameter weighted averaging, dynamic compensation, constraint conflict resolution, generating a fused parameter set, and performing feasibility pre-check.

[0011] S600, Process Constraint Verification: Verify the fusion parameters from multiple dimensions, including equipment compatibility, process stability, economic feasibility, and quality assurance. Use a quantitative model for evaluation, identify anomalies, calculate the PFI index, generate a verification report, and provide conclusions and handling suggestions.

[0012] Furthermore, S100 includes the following steps:

[0013] S110, Parameter Reception and Vectorization: Receives 8 material property parameters input by the user and constructs a standardized parameter vector;

[0014] S120, Parameter integrity check: Verify the integrity of input parameters and calculate the proportion of missing parameters;

[0015] S130, Abnormal parameter handling: Detect and process abnormal input values;

[0016] S140 outputs a normalized parameter vector, which has undergone integrity checks and anomaly handling.

[0017] Furthermore, in S110, the eight material property parameters include elastic modulus, yield strength, Poisson's ratio, hardening index, strength coefficient, elongation, density, and coefficient of thermal expansion.

[0018] Furthermore, S200 includes the following steps:

[0019] S210, Feature Vector Construction: Convert the input parameters into a standardized retrieval feature vector;

[0020] S220, Similarity Calculation: Calculate the similarity between the query vector and the cases in the knowledge graph;

[0021] S230, Candidate Set Screening: Applying dual constraints to screen valid candidate cases;

[0022] S240, Search Result Ranking: Ranking candidate cases in multiple dimensions;

[0023] S250, weight allocation: allocated according to the ratio of 60% similarity, 30% forming performance, and 10% temperature adaptability.

[0024] Furthermore, the S300 includes the following steps:

[0025] S310, Feature standardization processing: Dimensionless processing of input parameters and knowledge graph cases;

[0026] S320, Weight Allocation Matrix: Establishes the weight configuration for the importance of material parameters;

[0027] S330, Multi-dimensional similarity calculation: Calculate the similarity of each parameter dimension and synthesize them with weights;

[0028] S340, Process adaptability correction: Consideration of process parameter matching degree correction;

[0029] S350, Output Final Similarity: The final output is the corrected final similarity.

[0030] Furthermore, the S400 includes the following steps:

[0031] S410, Similarity threshold filtering: Initial screening based on corrected similarity;

[0032] S420, key parameter constraints: hard constraints on the performance of applied materials;

[0033] S430, Multi-objective optimization ranking: Constructing a comprehensive benefit evaluation function;

[0034] S440, Result Set Truncation: Output the optimal subset according to project requirements.

[0035] Furthermore, the S500 includes the following steps:

[0036] S510, Case Weight Allocation: Dynamically allocate fusion weights based on screening results;

[0037] S520, Multi-source parameter integration: Performs weighted fusion to generate a new parameter set;

[0038] S530, Process Constraint Correction: Ensure parameters meet production feasibility requirements;

[0039] S540, Fusion Validation: Evaluating the self-consistency of the parameter set.

[0040] Furthermore, the S600 includes the following steps:

[0041] S610, Equipment Capability Verification: Detecting whether the parameters exceed the physical limits of the production line;

[0042] S620, Process window check: Verify that the parameters are within the feasible process range;

[0043] S630, Economic Verification: Assessing the feasibility of production costs;

[0044] S640, Comprehensive Assessment: Calculate the process feasibility index;

[0045] S650, the final output is a verification report: the verification report includes a list of abnormal events, a set of optimization suggestions, and a feasibility index.

[0046] This invention also provides a process optimization system for multi-part adaptive self-adjusting stamping dies, used to perform one or more steps in the aforementioned process optimization method for multi-part adaptive self-adjusting stamping dies, including:

[0047] Input parameter receiving and processing module: Receives four types of core input data through a standardized interface, and finally generates a structured process input parameter set and stores it in the central database;

[0048] Knowledge graph retrieval module: Based on the input parameters received and processed by the input parameter receiving module, it searches for relevant cases and rules from the knowledge graph;

[0049] Similarity Calculation and Correction Module: Calculates the similarity between the current parameter and historical cases, uses a hybrid algorithm and weighted average, explains the weight of each parameter, handles special constraints, and finally normalizes to obtain the comprehensive similarity index;

[0050] Candidate case filtering module: Filters cases based on similarity index and constraints. Case filtering includes threshold filtering, constraint filtering, time-sensitive weighting, and diversity control to form an optimized candidate set.

[0051] Parameter fusion optimization module: Integrates current parameters and candidate cases to generate an optimized parameter set. Parameter fusion includes main parameter inheritance, auxiliary parameter weighted averaging, dynamic compensation, constraint conflict resolution, generating a fused parameter set, and performing feasibility pre-check.

[0052] Process constraint verification module: This module performs multi-dimensional verification of fusion parameters, including equipment adaptability, process stability, economic feasibility, and quality assurance. It uses quantitative models for evaluation, identifies anomalies, calculates the PFI index, generates a verification report, and provides conclusions and handling recommendations.

[0053] The present invention also provides a storage medium storing non-transitory computer-readable instructions for performing one or more steps in the aforementioned process optimization method for multi-part adaptive self-adjusting stamping dies.

[0054] The beneficial effects of this invention are as follows:

[0055] The optimization method proposed in this invention improves the production flexibility and efficiency of stamping die process. It achieves real-time monitoring and dynamic adjustment through intelligent control system, combined with adaptive die design optimization and centralized material property database management, ensuring accurate setting and efficient response of production process parameters. The integrated quality inspection system uses image recognition and machine learning technology to automatically identify and classify defects, which greatly improves inspection accuracy and efficiency. It effectively solves the problem of low production efficiency and unstable product quality caused by the inability to adjust key parameters in a timely manner under actual working conditions. Attached Figure Description

[0056] Figure 1 This is a flowchart of the process optimization method for the multi-part adaptive self-adjusting stamping die of the present invention;

[0057] Figure 2 This is the verification dimension weight allocation table proposed in this invention;

[0058] Figure 3 This is a detailed breakdown of the verification report proposed in this invention;

[0059] Figure 4 This is a structural block diagram of the process optimization system for the multi-part adaptive self-adjusting stamping die of the present invention.

[0060] In the diagram: 101, Input parameter receiving and processing module; 102, Knowledge graph retrieval module; 103, Similarity calculation and correction module; 104, Candidate case screening module; 105, Parameter fusion and optimization module; 106, Process constraint verification module. Detailed Implementation

[0061] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0062] like Figure 1 As shown, the process optimization method for multi-part adaptive self-adjusting stamping dies includes the following steps:

[0063] S100, receiving input parameters: It receives four types of core input data through a standardized interface, and finally generates a structured process input parameter set and stores it in the central database;

[0064] In one embodiment of the present invention, the core input data includes the following:

[0065] ① Equipment performance parameters (including operating ranges such as pressure, temperature, and speed);

[0066] ② Process specifications (including tolerance standards / quality indicators);

[0067] ③ Production order requirements (including batch size / delivery time / cost constraints);

[0068] ④ Material property data (including physical property parameters / rheological properties), the system automatically performs data integrity verification (verification rule base contains 32 constraints), unit system conversion (supports 6 standards such as SI / Imperial), and outlier correction (based on historical data interpolation);

[0069] In one embodiment of the present invention, the material property data specifically includes the following steps:

[0070] S110, Parameter Reception and Vectorization: Receives 8 material property parameters input by the user and constructs a standardized parameter vector;

[0071] The calculation formula is as follows:

[0072]

[0073] in Let E represent the input vector, and σ represent the elastic modulus. y ν represents yield strength, ν represents Poisson's ratio, n represents hardening index, K represents strength coefficient, δ represents elongation, ρ represents density, and α represents coefficient of thermal expansion.

[0074] It should be noted that the material property parameters include:

[0075] Elastic modulus: The ratio of stress to strain in a material during the elastic deformation stage, which measures the material's ability to resist deformation;

[0076] Yield strength: The stress value at which a material begins to undergo plastic deformation, marking the critical point at which a material transitions from elastic deformation to plastic deformation;

[0077] Tensile strength: The maximum stress a material can withstand during tension, indicating the material's maximum load-bearing capacity before fracture;

[0078] Poisson's ratio: The ratio of transverse strain to longitudinal strain of a material under stress, reflecting the volume change characteristics of the material under stress;

[0079] Hardening index: describes the hardening behavior of the stress-strain curve of a material during plastic deformation, and reflects the change in the degree of hardening of the material during deformation;

[0080] Strength factor: Used together with the hardening index to describe the stress-strain relationship of a material, especially in the plastic deformation stage;

[0081] Elongation at break: The maximum elongation of a material before fracture, representing the material's ductility;

[0082] Density: Mass per unit volume, which affects the mass distribution and energy absorption characteristics of a material;

[0083] Coefficient of thermal expansion: The proportionality of material dimensional changes caused by temperature variations, affecting the dimensional stability of materials at different temperatures.

[0084] It should be noted that all four types of core data corresponding to each material property parameter exist. For example, when inputting the elastic module, in addition to its material property data, it also includes the corresponding parameters of the equipment that produces the finished product, the required tolerance standards, and the parameters corresponding to the production order of the finished product. Therefore, the input is divided into four types of core input data.

[0085] S120, Parameter integrity check: Verify the integrity of input parameters and calculate the proportion of missing parameters;

[0086] Calculation formula:

[0087]

[0088] Where N received C represents the number of parameters received. complete Represents the percentage of completeness (threshold ≥ 75%), and I(·) represents the indicator function (1 for presence, 0 for absence);

[0089] S130, Abnormal parameter handling: Detect and process abnormal input values;

[0090] The verification rules for handling abnormal parameters are as follows:

[0091]

[0092] If the strategy for handling abnormal parameters is as follows:

[0093] 1. Out-of-range parameters: trigger knowledge graph correction;

[0094] 2. Missing parameters: filled with the average value of similar materials;

[0095]

[0096] Where M missing M represents the filler value for missing parameters. k N represents the parameter value corresponding to the k-th case. KG This represents the total number of process cases stored in the knowledge graph;

[0097] S140 outputs a normalized parameter vector, which has undergone integrity verification and anomaly handling.

[0098]

[0099] in This represents the standardized parameter vector, where adj represents the adjusted parameter state flags. arrive These represent eight adjusted material parameters;

[0100] S200, Knowledge Graph Retrieval: Based on the input parameters in S100, the system needs to search for relevant cases and rules from the knowledge graph;

[0101] The knowledge graph includes nodes such as equipment parameters, process parameters, historical cases, and material properties;

[0102] The search employs multi-dimensional index matching, such as equipment type, process type, and material type, as well as weight allocation strategies, such as 40% for equipment matching, 30% for process, 20% for material, and 10% for exception handling. It also involves dynamic expansion of the search, such as similar processes or alternative materials.

[0103] In one embodiment of the present invention, knowledge graph retrieval includes the following steps:

[0104] S210, Feature Vector Construction: Convert the input parameters into a standardized retrieval feature vector;

[0105] Calculation formula:

[0106]

[0107] in This represents the standardized retrieval feature vector, with the denominator being the baseline value: E0 = 200 GPa, σ0 = 300 MPa, n0 = 0.3, δ0 = 40.

[0108] S220, Similarity calculation between query vector and knowledge graph: Calculate the similarity between the query vector and the cases in the knowledge graph;

[0109] Calculation formula:

[0110]

[0111] Where S k ∈[0,1],S k V represents the similarity score of the k-th case. k Let represent the feature vector of the k-th case in the knowledge graph. This represents the i-th component of the query vector. This represents the i-th component of the k-th case vector;

[0112] S230, Candidate Set Screening: Applying dual constraints to screen valid candidate cases;

[0113] Filtering rules:

[0114]

[0115] Where |n input -n k | represents the difference in hardening exponents, δ kδ represents the extension rate of the k-th case. input Indicates the input elongation, n input Indicates the hardening exponent, n k This represents the hardening index of the k-th case;

[0116] S240, Search Result Ranking: Ranking candidate cases in multiple dimensions;

[0117] Sorting rules:

[0118]

[0119] Where T k RankScore represents the process temperature of the k-th case. k S represents the ranking score of the k-th candidate case. k δ represents the similarity score. k Indicates elongation;

[0120] S250, weight allocation: allocated according to the ratio of 60% similarity, 30% forming performance, and 10% temperature adaptability;

[0121] CandidateSet = {(C j S j RankScore j |j=1,...,N valid};

[0122] Where N valid CandidateSet represents the number of cases that passed the screening, and C represents the sorted set of candidate cases. j S represents the cost metric for the j-th case. j RankScore represents the similarity metric for the j-th case. j This represents the ranking score of the j-th case;

[0123] S300, Calculation of similarity between current parameters and historical cases: Calculate the similarity between current parameters and historical cases using a hybrid algorithm, such as Euclidean distance, cosine similarity, and Jaccard coefficient, and then use a weighted average. The weight of each parameter needs to be specified, for example, equipment parameters account for 40%, process parameters for 30%, quality results for 20%, and economic indicators for 10%.

[0124] At the same time, special constraints, such as mandatory matching items (material grades must be consistent) and degraded matching items (acceptable similar processes), are handled, and finally normalized to obtain the comprehensive similarity index;

[0125] In one embodiment of the present invention, similarity calculation specifically includes the following steps:

[0126] S310, Feature standardization processing: Dimensionless processing of input parameters and knowledge graph cases;

[0127] Calculation formula:

[0128]

[0129] Where μ i Let σ represent the mean of the i-th parameter in the knowledge graph. i Let X represent the standard deviation of the i-th parameter, N represent the total number of cases in the knowledge graph, and X represent the standard deviation of the i-th parameter. i Let X′ represent the i-th input parameter. i X represents the standardized value of the i-th parameter. i (k) represents the value of the i-th parameter in the k-th case;

[0130] S320, Weight Allocation Matrix: Establishes the weight configuration for the importance of material parameters;

[0131] Weight configuration:

[0132] W=[0.30, 0.25, 0.15, 0.10, 0.08, 0.07, 0.03, 0.02];

[0133] Where W1 = 0.30, is the weight of the elastic modulus; W2 = 0.25, is the weight of the yield strength; W3 = 0.15, is the weight of the hardening index; W4 = 0.10, is the weight of Poisson's ratio; W5 = 0.08, is the weight of the strength coefficient; W6 = 0.07, is the weight of the elongation; W7 = 0.03, is the weight of the density; and W8 = 0.02, is the weight of the coefficient of thermal expansion.

[0134] S330, Multi-dimensional similarity calculation: Calculate the similarity of each parameter dimension and synthesize them with weights;

[0135] Calculation formula:

[0136]

[0137] Where β i S represents the sensitivity coefficient of the i-th parameter (default β = 2.5). i ∈(0,1],S i S represents the similarity of the i-th parameter. total ∈[0,1],S total X′ represents the overall similarity. i Y′ represents the standardized value of the i-th input parameter. i W represents the standardized value of the parameter for the i-th case. i This represents the weight of the i-th parameter;

[0138] S340, Process adaptability correction: Consideration of process parameter matching degree correction;

[0139] Correction formula:

[0140]

[0141] Where S final T represents the corrected overall similarity. input Indicates the input process temperature, T case Indicates the process temperature in the case study;

[0142] S350, Output final similarity: The final similarity is adjusted;

[0143]

[0144] Where ResultSet represents the final similarity result set with corrections, N candidate Indicates the number of candidate cases, CaseID j This represents the ID of the j-th case. Let represent the final similarity of the j-th case;

[0145] S400, Candidate Set Filtering: Filtering cases based on similarity index and constraints;

[0146] The filtering of cases includes threshold screening (e.g., similarity ≥ 0.7), constraint filtering (excluding cases with quality incidents), time-based weighting (cases from the last two years + 15%), and diversity control (retaining the top 3 cases of different equipment models), forming an optimized candidate set;

[0147] In one embodiment of the present invention, candidate set screening specifically includes the following steps:

[0148] S410, Similarity threshold filtering: Initial screening based on corrected similarity;

[0149] Filtering criteria:

[0150] S final ≥0.75;

[0151]

[0152] in δ represents the elongation retention rate. case δ represents the extension rate of the case. input Input elongation;

[0153] S420, key parameter constraints: hard constraints on the performance of applied materials;

[0154] Constraint formula:

[0155]

[0156] Where σ case σ represents the yield strength of the case. input Indicates the input yield strength, |n case -n input | refers to the hardening index deviation, E case E represents the elastic modulus of the case. input Indicates the input elastic modulus;

[0157] S430, Multi-objective optimization ranking: Constructing a comprehensive benefit evaluation function;

[0158] The formula for the comprehensive benefit evaluation function is as follows:

[0159]

[0160] Where RankScore represents the ranking score, and Cost... base Cost represents the basic cost in a knowledge graph. case Let ΔT represent the case cost, and ΔT represent the temperature difference, where ΔT = T case -T input T case Indicates the process temperature in the case study, T input Indicates the input process temperature;

[0161] S440, Result Set Truncation: Output the optimal subset according to project requirements;

[0162] Truncation rules:

[0163]

[0164] Where outputSet represents the optimal subset of output, RankScore represents the ranking score, and Top represents the ranking order. For example, Top3 represents the top three optimal subsets.

[0165] The final output is:

[0166] FinalSet = {(CaseID)} j RankScore j KeyParams j |j=1,...,N output};

[0167] Where N output The FinalSet represents the number of output cases determined according to the truncation rule, and the KeyParams represents the optimal set of final outputs. j Let CaseID represent the set of key parameters for the j-th case. jLet RankSCore represent the ID of the j-th case. j This represents the ranking score of the j-th case;

[0168] S500, Parameter Fusion: Integrates current parameters with candidate cases to generate an optimized parameter set;

[0169] In one embodiment of the present invention, parameter fusion is achieved through the following steps:

[0170] S510, Case Weight Allocation: Dynamically allocate fusion weights based on screening results;

[0171] Calculation formula:

[0172]

[0173] Where γ = 0.02, γ is the distance attenuation coefficient, and d j Let rankScore be the Mahalanobis distance between case j and the input parameter space. j ∈[0,1], RankScore j w represents the ranking score of the j-th case. j RankScore represents the weight of the j-th case. k d represents the ranking score of the k-th case. k Let k be the Mahalanobis distance between case k and the input parameter space, and N represent the total number of cases.

[0174] S520, Multi-source parameter integration: Performs weighted fusion to generate a new parameter set;

[0175] Fusion formula:

[0176]

[0177] Where μ E Let represent the mean of the elastic modulus in the knowledge graph, i(·) represent the indicator function (takes 1 if the condition is met), and the exponents 0.7 / 0.3 are empirical adjustment coefficients. fused This represents the elastic modulus after fusion. n represents the yield strength after fusion. fused E represents the hardening index after fusion. j Let j represent the elastic modulus of the j-th case. Let n represent the yield strength of the j-th case. j This represents the hardening index of the j-th case;

[0178] S530, Process Constraint Correction: Ensure parameters meet production feasibility requirements;

[0179] Rule revision:

[0180]

[0181] Where ΔT is the maximum temperature difference in the candidate set, T fused T represents the process temperature after fusion. case The case process temperature is represented, where the correction priority is equipment limit > process window > theoretical value;

[0182] S540, Fusion Validation: Evaluating the self-consistency of the parameter set;

[0183] Verification metrics:

[0184]

[0185] Consistency represents the parameter consistency index, and Robustness represents the parameter stability index. Let |ΔParams| represent the mean of the i-th parameter in the knowledge graph. j | represents the degree of difference between the j-th case and the fusion parameters. This represents the i-th fusion parameter. Let represent the final similarity of the j-th case, denoted by the criteria of Consistency ≥ 0.85 and Robustness ≥ 0.7.

[0186] S600, Process Constraint Verification: Verify the fusion parameters in multiple dimensions, including equipment compatibility, process stability, economic feasibility, and quality assurance.

[0187] The evaluation is conducted using quantitative models (such as finite element analysis, cost simulation, and SPC prediction), outliers are identified, the PFI index is calculated, a verification report is generated, and conclusions and handling recommendations are given.

[0188] In one embodiment of the present invention, process constraint verification specifically includes the following steps:

[0189] S610, Equipment Capability Verification: Detecting whether the parameters exceed the physical limits of the production line;

[0190] Verification formula:

[0191]

[0192] Where P fused Indicates the forming pressure after fusion. Indicates the upper limit of the equipment's safe pressure. Indicates the lower limit of the safe temperature of the equipment. Indicates the upper limit of the safe temperature of the equipment, d fused D represents the diameter of the fused part. die Indicates the dimensions of the mold cavity;

[0193] S620, Process window check: Verify that the parameters are within the feasible process range;

[0194] Window constraints:

[0195] CRcool∈[μ CR -3σ CR μ CR +2σ CR ];

[0196]

[0197] Where CRcool represents the cooling rate, μ CR σ represents the average cooling rate. CR α represents the standard deviation of the cooling rate. safe Represents the safety factor (α) safe =0.25), Δt hold Indicates the holding time, h fused v represents the thickness of the fused part. min Minimum mold closing speed;

[0198] S630, Economic Verification: Assessing the feasibility of production costs;

[0199] Validate the model:

[0200]

[0201] Cost baseline Cost represents the historical average cost of the production line. fused The value represents the cost of the fused molded part, Q represents the current order quantity, Q0 represents the baseline order quantity (Q0 = 1000 pieces), and β represents the cost elasticity coefficient (β = 0.15).

[0202] S640, Comprehensive Assessment: Calculate the process feasibility index;

[0203] Evaluation formula:

[0204]

[0205] Where PFI represents the process feasibility index, k represents the sensitivity coefficient (k=2), and x i Let x represent the pass rate of the i-th key parameter. i0 This represents the baseline value of the i-th parameter, and ΔE represents the percentage increase in energy consumption.

[0206] The pass criterion for comprehensive evaluation is PFI ≥ 0.82;

[0207] Validation dimension weight allocation as follows Figure 2 As shown:

[0208] S650, final output verification report:

[0209] Report verify ={PFI, Violation List, Suggested Adjustment};

[0210] Report verify This refers to the validation report; PFI represents the process feasibility index; Violation List represents the list of abnormal events; and Suggested Adjustment represents the set of optimization suggestions.

[0211] In one embodiment of the present invention, the verification report includes the following:

[0212] 1. Overall Feasibility Index: PFI value (range 0-1);

[0213] 2. Overall conclusion: Approved / Conditionally approved / Rejected;

[0214] 3. Key Risk Warnings: List the top 3 high-risk items;

[0215] The verification report includes a breakdown of verification items as follows: Figure 3 As shown;

[0216] In one embodiment of the present invention, the list of abnormal events includes the following:

[0217] 1. Process parameter offset:

[0218] Cooling rate CR = 12℃ / s (allowable range 10±1.5℃ / s);

[0219] The pressure holding time t = 8.2s (theoretical value 7.5 ± 0.5s);

[0220] 2. Equipment load early warning:

[0221] The load rate of the mold clamping mechanism is 89% (threshold ≤ 85%).

[0222] The peak pressure of the hydraulic system is 92 MPa (safety limit 95 MPa).

[0223] In one embodiment of the present invention, the set of optimization suggestions includes the following:

[0224] 1. Adjustment of process parameters:

[0225] It is recommended to adjust the cooling rate to the range of 10.5-11.5℃ / s;

[0226] The recommended holding time is shortened to 7.8 ± 0.3 seconds;

[0227] 2. Production assurance measures:

[0228] The hydraulic system is recommended to be forced to cool every 4 hours;

[0229] Mold temperature compensation +5℃ gradient control;

[0230] In one embodiment of the present invention, the following example is given according to S100-S600 described above: S100, input parameters are received:

[0231] Processing material: DC04 cold-rolled steel sheet (thickness 1.2±0.1mm);

[0232] Mold type: Six-station self-adjusting progressive die (maximum stroke 120mm);

[0233] Target parts family: door hinge reinforcement plate / seat rail / seat belt mounting bracket;

[0234] Process requirements: stamping force ≤3500kN, forming speed ≥12spm;

[0235] Quality constraints: springback angle ≤ 0.8°, surface scratch depth < 15μm;

[0236] Economic indicators: Changeover time < 25 min, mold wear rate ≤ 0.03% / thousand cycles;

[0237] S200, Knowledge Graph Retrieval:

[0238] In the stamping process knowledge graph:

[0239] Position the self-adjusting mold node and extract its dynamic compensation mechanism parameters (guide pillar clearance).

[0240] 0.02-0.05mm);

[0241] Match the DC04 material node to obtain 12 sets of historical forming parameters;

[0242] By examining multiple component adaptation cases, three adaptive plate thickness solutions (1.0-1.5mm) were identified.

[0243] Returns 17 valid nodes (6 sets of mold structure parameters + 8 sets of material forming schemes + 3 changeover rules);

[0244] S300, Similarity Calculation:

[0245] Mold structure matching degree (35%): stroke range coverage 0.95, guide pillar clearance matching 0.88;

[0246] Material formability (30%): DC04 springback compensation empirical value matching degree 0.82;

[0247] Multi-part compatibility (25%): Part size distribution overlap 0.78;

[0248] Economic indicators (10%): historical best changeover time of 18 minutes;

[0249] Overall similarity: Case X = 0.91, Case Y = 0.84, Case Z = 0.73;

[0250] S400, Candidate Set Filtering:

[0251] Exclude single-part-specific solutions (similarity > 0.8 but no multi-station adaptation);

[0252] Cases with adaptive plate thickness compensation mechanisms should be prioritized:

[0253] The top 5 cases are retained, including:

[0254] Case X: Hydraulic dynamic pressure regulation (rapid switching between 3 types of parts);

[0255] Case Y: Elastomer-filled gap compensation (thickness adaptable to 1.0-1.8mm);

[0256] Case W: Electromagnetic assisted molding (reduces springback by 0.5°);

[0257] S500, parameter fusion:

[0258] Main parameter inheritance: Maintain the upper limit of 3500kN pressure;

[0259] Dynamic parameter optimization:

[0260] Take the blank holder force gradient control of case X (1200kN→800kN);

[0261] The pre-compression of the elastomer in integrated case Y (initial compression 15%);

[0262] The electromagnetic auxiliary parameters (pulse frequency 8Hz) of Case W are adopted;

[0263] The switching logic integrates three positioning methods (mechanical + laser + pneumatic);

[0264] S600, Process Constraint Verification:

[0265] Mold compatibility: Finite element analysis shows that the maximum stress is less than 80% of the material's yield strength;

[0266] Multi-part compatibility: Simulated springback <0.7° for thicknesses of 1.0-1.5mm;

[0267] Changeover efficiency: Digital twin verification changeover time is 23.5 minutes (including automatic positioning and calibration);

[0268] Economic verification: The predicted mold loss rate is 0.027% / thousand cycles (meets the standard);

[0269] Output optimization scheme: Enable electromagnetic assistance + elastomer compensation composite mode, and it is recommended to check the guide post gap every 5000 cycles.

[0270] like Figure 4 As shown, this invention also proposes a process optimization system for multi-part adaptive self-adjusting stamping dies, comprising the following modules:

[0271] Input parameter receiving and processing module 101: Receives four types of core input data through a standardized interface, and finally generates a structured process input parameter set and stores it in the central database;

[0272] Knowledge graph retrieval module 102: Based on the input parameters in the input parameter receiving and processing module, it searches for relevant cases and rules from the knowledge graph;

[0273] Similarity Calculation and Correction Module 103: Calculates the similarity between the current parameter and historical cases, uses a hybrid algorithm and weighted average, explains the weight of each parameter, handles special constraints, and finally normalizes to obtain the comprehensive similarity index.

[0274] Candidate case screening module 104: Filters cases based on similarity index and constraints. Case filtering includes threshold screening, constraint filtering, time-sensitive weighting, and diversity control to form an optimized candidate set.

[0275] Parameter fusion optimization module 105: Integrates current parameters and candidate cases to generate an optimized parameter set. Parameter fusion includes main parameter inheritance, auxiliary parameter weighted averaging, dynamic compensation, constraint conflict resolution, generating a fused parameter set, and performing feasibility pre-check.

[0276] Process constraint verification module 106: Performs multi-dimensional verification of fusion parameters, including equipment adaptability, process stability, economic feasibility, and quality assurance. It uses a quantitative model for evaluation, identifies anomalies, calculates the PFI index, generates a verification report, and provides conclusions and handling suggestions.

[0277] At least one embodiment of the present invention discloses a storage medium storing non-transitory computer-readable instructions for performing one or more steps in the aforementioned process optimization method for multi-part adaptive self-adjusting stamping dies.

[0278] Computer programs may be stored / distributed on suitable media, such as optical storage media or solid-state media supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunications systems. No reference numerals in the claims should be construed as limiting the scope.

[0279] The embodiments of this example have been described above. However, this example is not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of this example, and all of them are within the protection scope of this example.

Claims

1. A process optimization method for multi-part adaptive self-adjusting stamping dies, characterized in that, Includes the following steps: S100, receiving input parameters: It receives four types of core input data through a standardized interface, and finally generates a structured process input parameter set and stores it in the central database; The four types of core input data include equipment performance parameters, process specification requirements, production order requirements, and material property data. The material property data include elastic modulus, yield strength, Poisson's ratio, hardening index, strength coefficient, elongation, density, and coefficient of thermal expansion. S200, Knowledge Graph Retrieval: Based on the input parameters in S100, search for relevant cases and rules from the knowledge graph; S300, Similarity Calculation: Calculate the similarity between the current parameter and historical cases, use a hybrid algorithm and weighted average, explain the weight of each parameter, handle special constraints, and finally normalize to obtain the comprehensive similarity; Hybrid algorithms include Euclidean distance, cosine similarity, and Jaccard coefficient; Among them, handling special constraints refers to mandatory matching items and degraded matching items. Mandatory matching items are consistent material grades, and degraded matching items are acceptance of similar processes. Specifically, the following steps are included: S310, Feature standardization processing: Dimensionless processing of input parameters and knowledge graph cases; S320, Weight Allocation Matrix: Establishes the weight configuration for the importance of material parameters; S330, Multi-dimensional similarity calculation: Calculate the similarity of each parameter dimension and synthesize them with weights; S340, Process adaptability correction: Consideration of process parameter matching degree correction; S350, Output final similarity: The final output is the corrected final similarity; S400, Candidate Set Filtering: Cases are filtered based on comprehensive similarity and constraints. Case filtering includes threshold filtering, constraint filtering, time-weighted filtering, and diversity control to form an optimized candidate set. Among them, threshold screening is an initial screening based on corrected similarity. Among them, constraint filtering is used to exclude cases with quality incidents; Among them, the time-sensitivity weighting is the weight of cases that are two consecutive years apart + 15%; Among them, diversity control is represented by the top 3 cases of retaining different equipment models; S500, Parameter Fusion: Integrates current parameters with candidate cases to generate an optimized parameter set. Parameter fusion includes main parameter inheritance, auxiliary parameter weighted averaging, dynamic compensation, constraint conflict resolution, generating a fused parameter set, and performing feasibility pre-check. Dynamic compensation, through process constraint correction, ensures that parameters meet production feasibility requirements; ； in, The maximum temperature difference in the candidate set; Indicates the temperature of the formed part. Indicates the temperature of the case; S600, Process Constraint Verification: Verify the fusion parameters in multiple dimensions, including equipment adaptability, process stability, economic feasibility, and quality assurance. Use a quantitative model for evaluation, identify outliers, calculate the PFI index, generate a verification report, and provide conclusions and handling suggestions. The formula for calculating the PFI index is as follows: ； in Indicates the feasibility index. This represents the steepness coefficient of the Sigmoid function. , This represents the compliance rate of 5 key parameters. This represents the percentage increase in energy consumption. These serve as the baseline values ​​for five key parameters; The passing criteria for the comprehensive evaluation are: .

2. The process optimization method for multi-part adaptive self-adjusting stamping dies according to claim 1, characterized in that, S100 includes the following steps: S110, Parameter Reception and Vectorization: Receives 8 material property parameters input by the user and constructs a standardized parameter vector; S120, Parameter integrity check: Verify the integrity of input parameters and calculate the proportion of missing parameters; S130, Abnormal parameter handling: Detect and process abnormal input values; S140 outputs a normalized parameter vector, which has undergone integrity checks and anomaly handling.

3. The process optimization method for multi-part adaptive self-adjusting stamping dies according to claim 2, characterized in that, S200 includes the following steps: S210, Feature Vector Construction: Convert the input parameters into a standardized retrieval feature vector; S220, Similarity calculation between query vector and knowledge graph: Calculate the similarity between the query vector and the cases in the knowledge graph; S230, Candidate Set Screening: Applying dual constraints to screen valid candidate cases; S240, Search Result Ranking: Candidate cases are ranked in multiple dimensions, with ranking weights allocated according to the following proportions: similarity 60%, forming performance 30%, and temperature adaptability 10%. ； in, This represents the ranking score of the k-th candidate case. Indicates the similarity score. Elongation, i.e., forming performance, This represents the process temperature of the k-th case, i.e., temperature adaptability.

4. A process optimization system for multi-part adaptive self-adjusting stamping dies, characterized in that, The steps in the process optimization method for multi-part adaptive self-adjusting stamping dies as described in any one of claims 1-3 include: Input parameter receiving and processing module: Receives four types of core input data through a standardized interface, and finally generates a structured process input parameter set and stores it in the central database; Knowledge graph retrieval module: Based on the input parameters received and processed by the input parameter receiving module, it searches for relevant cases and rules from the knowledge graph; Similarity Calculation and Correction Module: Calculates the similarity between the current parameter and historical cases, uses a hybrid algorithm and weighted average, explains the weight of each parameter, handles special constraints, and finally normalizes to obtain the comprehensive similarity. Candidate case selection module: Filters cases based on comprehensive similarity and constraints. Case filtering includes threshold selection, constraint filtering, time-sensitive weighting, and diversity control to form an optimized candidate set. Parameter fusion optimization module: Integrates current parameters and candidate cases to generate an optimized parameter set. Parameter fusion includes main parameter inheritance, auxiliary parameter weighted averaging, dynamic compensation, constraint conflict resolution, generating a fused parameter set, and performing feasibility pre-check. Process constraint verification module: Performs multi-dimensional verification of fusion parameters, including equipment adaptability, process stability, economic feasibility, and quality assurance. It uses quantitative models for evaluation, identifies anomalies, calculates the PFI index, generates a verification report, and provides conclusions and handling suggestions.

5. A storage medium, characterized in that, It stores non-transitory computer-readable instructions for performing steps in the process optimization method for a multi-part adaptive self-adjusting stamping die as described in any one of claims 1-3.

Images