Intelligent control method and system for production line mold machining

Abstract

The invention relates to the technical field of production line control, and discloses an intelligent control method and system for production line mold machining. According to the method, a pressure distribution state is determined by obtaining initial cutting parameters of a cutter and surface pressure data of a mold, mold node stress distribution is calculated by utilizing a three-dimensional finite element model, and a stress concentration position is identified. And predicting a mold deformation trend through a linear regression model according to the thermal expansion coefficient and the environment temperature data. Meanwhile, the initial cutting parameters are optimized by combining a cutting parameter optimization objective function, and the objective function comprehensively considers factors of tool wear, surface quality and residual stress. And on the basis of a die deformation prediction result, a dynamic compensation algorithm is adopted to adjust supporting structure parameters and optimize cutting parameters, and finally optimal cutting parameters and a target tool path are generated. According to the method, precise coordination control over pressure distribution and mold deformation under the complex working condition is achieved, the residual deformation risk is reduced, and the mold machining precision is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of production line control, and in particular to an intelligent control method and system for production line mold processing. Background Art

[0002] With the advancement of industrialization, production is shifting toward refinement and efficiency. Molds have emerged as tools for mass-producing products and improving production efficiency. From simple plastic products to complex automotive parts, molds precisely transform raw materials into finished products based on pre-designed patterns, ensuring consistent size and shape for each product. They are crucial for achieving large-scale production and ensuring consistent product quality in modern industry.

[0003] In a traditional control technology for mold processing on a production line, the specific operational process is as follows: the operator first sets fixed cutting speed and feed parameters based on experience or process manuals, then starts the CNC machine tool for high-speed cutting. During the machining process, the system uses contact sensors to collect real-time temperature data from the cutting area. When the temperature exceeds a preset threshold, the PID control module triggers linear adjustments to the coolant flow rate to reduce the temperature rise. The operator then manually adjusts the cutting parameters or stops the machine for tool compensation based on the interim machining results, visually inspecting the mold dimensions or using simple gaging tools to spot-check the mold dimensions. To address stress variations caused by material inhomogeneity, a symmetrical machining path planning strategy combined with a fixed retraction frequency is typically adopted to periodically interrupt the cutting process to release stress. After machining, manual aging is required, placing the mold in a constant temperature chamber for 8-12 hours to eliminate residual stress. This operating method, based on static parameter settings, segmented linear adjustment, and manual intervention, lacks real-time coordinated control capabilities to address dynamic coupling, resulting in a positive feedback effect between pressure fluctuations and thermal deformation.

[0004] In conventional control technologies for production line mold processing, dynamic changes in cutting speed and feed rate during high-speed machining generate local pressure fluctuations. Simultaneously, mold deformation due to thermal expansion evolves dynamically with temperature. This dynamic nature makes the coupling relationship between pressure distribution and mold deformation difficult to predict. Furthermore, the non-uniformity of mold materials and the accumulation of residual stresses during machining further complicate the control of pressure distribution and deformation.

[0005] In summary, it is difficult to achieve precise coordinated control of pressure distribution and mold deformation under complex working conditions using traditional control technologies, which increases the risk of residual deformation and leads to low mold processing accuracy. Summary of the Invention

[0006] The present invention provides an intelligent control method and system for production line mold processing, so as to achieve precise coordinated control of pressure distribution and mold deformation under complex working conditions, reduce the risk of residual deformation, and improve mold processing accuracy.

[0007] In a first aspect, in order to solve the above technical problems, the present invention provides an intelligent control method for production line mold processing, comprising:

[0008] Obtain the initial cutting parameters of the tool and the surface pressure data of the mold to determine the pressure distribution state;

[0009] According to the pressure distribution state, the stress distribution of each node of the mold is calculated by a preset three-dimensional finite element model to determine the stress concentration location;

[0010] Obtaining the thermal expansion coefficient and ambient temperature of the mold at the stress concentration location;

[0011] Inputting the thermal expansion coefficient and the ambient temperature into a pre-established linear regression model to predict the mold deformation trend;

[0012] According to the initial cutting parameters, combined with a cutting parameter optimization objective function, an optimized cutting parameter is obtained; wherein the cutting parameter optimization objective function includes a relationship between the cutting parameters and tool wear, surface quality and residual stress;

[0013] According to the deformation trend of the mold, a dynamic compensation algorithm is used to adjust the supporting structure parameters of the mold and the optimized cutting parameters to obtain the optimal cutting parameters;

[0014] A target tool trajectory path is generated according to the support structure parameters and the optimal cutting parameters, so as to process the production line mold according to the target tool trajectory path.

[0015] In an optional embodiment, obtaining the initial cutting parameters of the tool and the surface pressure data of the mold and determining the pressure distribution state includes:

[0016] After initializing the tool processing machine, obtaining the initial cutting parameters of the tool;

[0017] The surface pressure data of the mold is obtained by using multiple pressure sensors covering various areas of the mold surface;

[0018] The surface pressure data are classified and marked with different colors to generate a mold surface pressure distribution map, wherein the different colors in the mold surface pressure distribution map are used to represent the pressure distribution state of the mold surface.

[0019] In an optional embodiment, calculating the stress distribution of each node of the mold using a preset three-dimensional finite element model according to the pressure distribution state and determining the stress concentration location includes:

[0020] Performing a stress calculation operation on each node of the pressure distribution state in combination with the three-dimensional finite element model to obtain a stress value of each node;

[0021] According to the stress values ​​of each node, a drawing operation is performed on the stress values ​​of each node to generate a stress distribution diagram;

[0022] Performing a coordinate extraction operation on the stress distribution map to obtain the coordinates of a stress region having a stress value higher than a preset stress threshold as a stress concentration location;

[0023] The calculation formula of the node stress value is as follows:

[0024]

[0025] Where, σ j represents the stress value of the jth node, n represents the total number of nodes, K ij The influence coefficient of the unit pressure of the i-th node on the stress of the j-th node is obtained through the three-dimensional finite element model, P i Indicates the pressure value of the i-th node, dA j The area element representing the j-th node is obtained by the three-dimensional finite element model;

[0026] The process of setting up the three-dimensional finite element model includes:

[0027] Scan and measure the mold using a scanner to obtain the surface geometric dimensions of the mold;

[0028] Establishing an actual three-dimensional model of the mold according to the surface geometric dimensions;

[0029] According to the actual three-dimensional model, a finite element meshing operation is performed on the actual three-dimensional model to obtain a three-dimensional finite element meshing model; wherein the finite element meshing includes tetrahedral element meshing;

[0030] According to the three-dimensional finite element partition model, a boundary condition setting operation is performed on the three-dimensional finite element partition model to obtain a three-dimensional finite element model; wherein the boundary condition setting includes a fixed support condition and a pressure load condition.

[0031] In an optional embodiment, obtaining the thermal expansion coefficient and the ambient temperature of the mold at the stress concentration location includes:

[0032] Based on the three-dimensional coordinate information of the stress concentration position, the lattice thermal expansion characteristic parameters of the corresponding material are matched in a preset material database, and the thermal expansion coefficient in the stress gradient direction is calculated in combination with an anisotropic interpolation algorithm;

[0033] A distributed optical fiber temperature sensor array collects ambient temperature in real time. The sensor array is configured with a high-density temperature measurement array in the stress concentration area and a low-density temperature measurement array in the remaining areas. In an optional embodiment, the thermal expansion coefficient and the ambient temperature are input into a pre-established linear regression model to predict the mold deformation trend, including:

[0034] Inputting the thermal expansion coefficient and the ambient temperature into a preset linear regression model to predict deformation trend data of the mold at the stress concentration location;

[0035] The deformation trend data is quantified using a preset deformation trend analysis algorithm, and the quantified deformation trend data is classified using a gradient boosting tree algorithm to determine the stability level of the deformation trend and obtain a stability level.

[0036] According to the stability level, combined with a preset material property database, query the elastic modulus and Poisson's ratio that match the mold material, and determine the deformation constraint conditions of the mold material at different stability levels;

[0037] According to the deformation constraint conditions, a predetermined Bayesian optimization algorithm is used to perform a probability prediction on the deformation distribution of the mold at the stress concentration position to obtain a deformation distribution probability;

[0038] Using the deformation trend data and the deformation distribution probability as a mold deformation trend;

[0039] The linear regression model is as follows:

[0040] ΔD=α·ΔT·(β1k+β2)+β3(ΔT) 2 +∈

[0041]

[0042] Among them, ΔD represents the deformation trend data, α represents the thermal expansion coefficient, ΔT represents the ambient temperature, k represents the stress concentration factor, σ max represents the maximum nodal stress value, σ avg represents the weighted average stress of the node, n represents the total number of nodes, σ j represents the stress value of the jth node, dA j A represents the area element of the jth node. totalrepresents the total surface area of ​​the mold, β1, β2, and β3 represent regression coefficients, and the units of β1, β2, and β3 are m, m, and m / K, respectively. 2 , β1, β2, and β3 are obtained by fitting the Levenberg-Marquardt algorithm, and ∈ represents the preset model residual term. In an optional embodiment, the optimized cutting parameters are obtained based on the initial cutting parameters in combination with a cutting parameter optimization objective function; wherein the cutting parameter optimization objective function includes the relationship between the cutting parameters and tool wear, surface quality, and residual stress, including:

[0043] The cutting parameter optimization objective function is as follows:

[0044]

[0045] Among them, V c Indicates the cutting speed in the cutting parameters, f indicates the feed rate in the cutting parameters, R a Represents the surface roughness of the mold, measured by white light interferometer, σ res represents the residual stress, which is measured by X-ray diffraction, vb represents the tool wear, ω1, ω2, ω3, ω4 and ω5 represent the preset dynamic weight coefficients, V c0 represents the initial cutting speed in the initial cutting parameters, f0 represents the initial feed in the initial cutting parameters, F opt Represents the output value of the objective function, R a,max , σ yickl and VB max Indicates the preset normalized standard value.

[0046] In an optional embodiment, adjusting the mold support structure parameters and the optimized cutting parameters using a dynamic compensation algorithm according to the mold deformation trend to obtain the optimal cutting parameters includes:

[0047] Smoothing and feature extraction are performed on the mold deformation trend to obtain deformation features;

[0048] Using a principal component analysis algorithm to perform correlation calculation on the deformation characteristics and pre-stored historical data of the support structure, a quantitative mapping relationship model between the mold deformation trend and pre-stored support structure parameters is established;

[0049] Based on the quantitative mapping relationship model, the support structure parameters are adjusted through the PID algorithm to obtain the preliminary adjusted support structure parameters;

[0050] Based on the initially adjusted support structure parameters and the optimized cutting parameters, the cutting parameters are updated and calculated using a linear interpolation method, and the performance of the updated cutting parameter set is predicted using a random forest algorithm to obtain a performance prediction result;

[0051] When the performance prediction result does not meet the preset performance threshold, iteratively updating the cutting parameters that do not meet the conditions by using a gradient descent algorithm;

[0052] When the performance prediction result meets a preset performance threshold, the most recently updated cutting parameters are used as the optimal cutting parameters.

[0053] In an optional embodiment, generating a target tool trajectory path according to the support structure parameters and the optimal cutting parameters, so as to process the production line mold according to the target tool trajectory path, includes:

[0054] Performing mesh analysis on the three-dimensional finite element model based on the support point density and stiffness distribution data in the support structure parameters to divide the force-constrained area and the free processing area;

[0055] Based on the three-dimensional geometric features of the free machining area and the cutting speed, feed rate and cutting depth in the optimal cutting parameters, an initial tool trajectory path is generated using a B-spline curve algorithm;

[0056] Performing collision detection on the initial tool trajectory path and the force-constrained area, and if a tool support interference area exists, optimizing and reconstructing the trajectory of the tool support interference area through a local path adjustment algorithm;

[0057] Importing the optimized tool trajectory path into the finite element simulation model, and generating a tool load dynamic curve in combination with the support structure parameters;

[0058] When the tool load exceeds the preset stress threshold, the trajectory curvature radius and feed angle are iteratively adjusted through a genetic algorithm to generate a smooth path that meets the residual stress constraint.

[0059] Finally, the path segments within each optimization parameter domain are integrated to generate a target tool trajectory path with continuous tangent vectors, which is then converted into NC code for machining equipment to perform machining.

[0060] In a second aspect, the present invention provides an intelligent control system for production line mold processing, comprising:

[0061] The data acquisition module is used to obtain the initial cutting parameters of the tool and the surface pressure data of the mold to determine the pressure distribution state;

[0062] A stress analysis module is used to calculate the stress distribution of each node of the mold through a preset three-dimensional finite element model according to the pressure distribution state and determine the stress concentration location;

[0063] A data matching module is used to obtain the thermal expansion coefficient and ambient temperature of the mold at the stress concentration position;

[0064] a deformation trend analysis module, configured to input the thermal expansion coefficient and the ambient temperature into a pre-established linear regression model to predict the mold deformation trend;

[0065] a cutting parameter optimization module, configured to obtain optimized cutting parameters based on the initial cutting parameters and a cutting parameter optimization objective function; wherein the cutting parameter optimization objective function includes the relationship between the cutting parameters and tool wear, surface quality, and residual stress;

[0066] An optimal parameter analysis module, configured to adjust the mold support structure parameters and the optimized cutting parameters using a dynamic compensation algorithm according to the mold deformation trend to obtain optimal cutting parameters;

[0067] The processing path generation module is used to generate a target tool trajectory path according to the support structure parameters and the optimal cutting parameters, so as to process the production line mold according to the target tool trajectory path.

[0068] In a third aspect, the present invention also provides an electronic device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the intelligent control method for production line mold processing described in any one of the above items is implemented.

[0069] In a fourth aspect, the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned intelligent control methods for production line mold processing.

[0070] Compared with the prior art, the present invention has the following beneficial effects:

[0071] 1. Through the three-dimensional finite element model and pressure sensor network, the mold surface pressure distribution map is generated in real time and the node stress value is quantified. Combined with the linear regression model, the deformation trend is predicted to overcome the problem of coupling pressure fluctuation and thermal deformation.

[0072] 2. Dynamic compensation mechanism for multi-parameter collaborative optimization: Based on Bayesian optimization and gradient boosting tree algorithms, support structure parameters and cutting parameters are dynamically corrected to achieve a multi-objective balance among tool wear rate, surface roughness, and residual stress during machining.

[0073] 3. Precise suppression of thermal-mechanical coupling effects: A thermocouple sensor array collects ambient temperature in real time, and a dynamic evaluation system for elastic modulus and Poisson's ratio is constructed in combination with a material database and deformation constraints. This improves the matching degree between the thermal expansion coefficient and material strength, suppresses processing offset caused by thermal deformation, and reduces mold flatness deviation.

[0074] 4. Use B-spline curves and genetic algorithms to optimize cutting paths, introduce support structure interference detection and local path reconstruction technology to reduce tool load fluctuations, improve tool path curvature smoothness, and ensure the overall mold processing quality.

[0075] In summary, the present invention solves the problem of difficulty in coordinating pressure distribution and mold deformation in traditional technologies through full-closed-loop intelligent control of pressure-stress-deformation, multi-parameter dynamic compensation mechanism and tool path optimization algorithm, and significantly improves processing accuracy, processing efficiency and mold service life. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 This is a flow chart of an intelligent control method for mold processing on a production line provided by a first embodiment of the present invention;

[0077] Figure 2 It is a structural diagram of an intelligent control system for production line mold processing provided by the second embodiment of the present invention. DETAILED DESCRIPTION

[0078] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0079] Reference Figure 1 The first embodiment of the present invention provides an intelligent control method for mold processing of a production line, comprising the following steps:

[0080] S11, obtaining initial cutting parameters of the tool and surface pressure data of the mold to determine the pressure distribution state;

[0081] S12, calculating the stress distribution of each node of the mold using a preset three-dimensional finite element model according to the pressure distribution state, and determining the stress concentration location;

[0082] S13, obtaining the thermal expansion coefficient and ambient temperature of the mold at the stress concentration location;

[0083] S14, inputting the thermal expansion coefficient and the ambient temperature into a pre-established linear regression model to predict a mold deformation trend;

[0084] S15, obtaining optimized cutting parameters based on the initial cutting parameters and a cutting parameter optimization objective function; wherein the cutting parameter optimization objective function includes a relationship between the cutting parameters and tool wear, surface quality, and residual stress;

[0085] S16, adjusting the mold support structure parameters and the optimized cutting parameters using a dynamic compensation algorithm according to the mold deformation trend to obtain optimal cutting parameters;

[0086] S17 , generating a target tool trajectory path according to the support structure parameters and the optimal cutting parameters, so as to process the production line mold according to the target tool trajectory path.

[0087] In step S11 , the initial cutting parameters of the tool and the surface pressure data of the mold are obtained to determine the pressure distribution state.

[0088] In a specific embodiment, obtaining the initial cutting parameters of the tool and the surface pressure data of the mold and determining the pressure distribution state includes:

[0089] After initializing the tool processing machine, obtaining the initial cutting parameters of the tool;

[0090] The surface pressure data of the mold is obtained by using multiple pressure sensors covering various areas of the mold surface;

[0091] The surface pressure data are classified and marked with different colors to generate a mold surface pressure distribution map, wherein the different colors in the mold surface pressure distribution map are used to represent the pressure distribution state of the mold surface.

[0092] Specifically, the tool's initial cutting parameters, including the initial cutting speed, feed rate, and depth of cut, are derived from the tool's initial settings during machining. These parameters record the tool's initial operating state during machining. Surface pressure data for the mold is collected via multiple pressure sensors located across various areas of the mold surface. These sensors monitor pressure changes on the mold surface in real time during machining.

[0093] The acquired surface pressure data needs to be classified and processed. By classifying the pressure data and marking different colors according to different pressure value ranges, a pressure distribution map of the mold surface is generated. The different colored areas in the map intuitively reflect the pressure distribution state of the mold surface. For example, in a mold milling operation, the area near the tool cutting path will appear dark red (corresponding to the pressure peak exceeding 200MPa), the transition zone that spreads outward will show an orange-yellow gradient (pressure values ​​between 50-200MPa), and the mold support structure area will be marked with a uniform blue (pressure below 50MPa). This distribution map that intuitively displays the high-pressure area, transition area, and low-pressure area through color gradient can clearly show the stress state of the mold surface. When the tool cuts into the mold corner, the dark red patch will extend toward the weak part of the structure. However, after adjusting the cutting parameters, the orange-yellow area will significantly shrink, and the blue safe zone will expand accordingly. This dynamic visualization of the pressure distribution map provides intuitive data support for subsequent stress analysis and deformation prediction. For example, if an engineer discovers an unexpected red streak on the mold sidewall, he can immediately determine whether there is a clamping anomaly or tool runout problem in the area, and quickly intervene and make adjustments.

[0094] Determining the pressure distribution reveals the stress conditions on the mold surface during machining, helping to identify potential areas of stress concentration or uneven pressure. By analyzing the pressure distribution, we can more accurately understand the mechanical behavior of the mold during machining, providing reliable data for subsequent stress calculations, deformation predictions, and parameter optimization.

[0095] In step S12, according to the pressure distribution state, the stress distribution of each node of the mold is calculated through a preset three-dimensional finite element model to determine the stress concentration position.

[0096] In a specific embodiment, the step of calculating the stress distribution of each node of the mold using a preset three-dimensional finite element model according to the pressure distribution state and determining the stress concentration location includes:

[0097] Performing a stress calculation operation on each node of the pressure distribution state in combination with the three-dimensional finite element model to obtain a stress value of each node;

[0098] According to the stress values ​​of each node, a drawing operation is performed on the stress values ​​of each node to generate a stress distribution diagram;

[0099] Performing a coordinate extraction operation on the stress distribution map to obtain the coordinates of a stress region having a stress value higher than a preset stress threshold as a stress concentration location;

[0100] The calculation formula of the node stress value is as follows:

[0101]

[0102] Where, σ j represents the stress value of the jth node, n represents the total number of nodes, K ij The influence coefficient of the unit pressure of the i-th node on the stress of the j-th node is obtained through the three-dimensional finite element model, P i Indicates the pressure value of the i-th node, dA j The area element representing the j-th node is obtained through the three-dimensional finite element model; wherein the setting process of the three-dimensional finite element model includes:

[0103] Scan and measure the mold using a scanner to obtain the surface geometric dimensions of the mold;

[0104] Establishing an actual three-dimensional model of the mold according to the surface geometric dimensions;

[0105] According to the actual three-dimensional model, a finite element meshing operation is performed on the actual three-dimensional model to obtain a three-dimensional finite element meshing model; wherein the finite element meshing includes tetrahedral element meshing;

[0106] According to the three-dimensional finite element partition model, a boundary condition setting operation is performed on the three-dimensional finite element partition model to obtain a three-dimensional finite element model; wherein the boundary condition setting includes a fixed support condition and a pressure load condition.

[0107] Specifically, first, the stress calculation operation of each node is performed on the pressure distribution state through a three-dimensional finite element model to obtain the stress value of each node. The calculation formula of the node stress value is:

[0108]

[0109] Where, σ j represents the stress value of the jth node, n represents the total number of nodes, K ij The influence coefficient of the unit pressure of the i-th node on the stress of the j-th node is obtained through the three-dimensional finite element model, P i Indicates the pressure value of the i-th node, dA j The area element representing the j-th node is obtained by the three-dimensional finite element model;

[0110] The process of setting up a three-dimensional finite element model first involves scanning and measuring the mold with a scanner to obtain the surface geometric dimensions of the mold, including length, width, and height. Secondly, an actual three-dimensional model of the mold is established based on the surface geometric dimensions to ensure that the model can accurately reflect the physical shape of the mold. Next, a finite element meshing operation is performed on the actual three-dimensional model to obtain a three-dimensional finite element partitioning model, wherein the finite element meshing adopts tetrahedral element meshing to divide the mold into multiple small units for subsequent stress calculations. Finally, a boundary condition setting operation is performed on the three-dimensional finite element partitioning model to obtain a three-dimensional finite element model, wherein the boundary condition setting includes fixed support conditions and pressure load conditions. The fixed support conditions simulate the constraints of the mold being fixedly supported during the processing process, and the pressure load conditions simulate the pressure distribution applied by the tool to the mold surface.

[0111] The purpose of boundary condition setting is to simulate the constraints and external forces that the mold is subject to during the actual processing process, so as to ensure that the finite element analysis can accurately reflect the actual situation. The specific setting process includes the following two aspects: fixed support conditions and pressure load conditions. Fixed support conditions are used to simulate the constraints of the mold being fixedly supported during the processing process. In the model, the bottom surface of the mold or a specific support area is selected as the fixed boundary, and the displacement degrees of freedom of all nodes in these areas are set to zero, that is, their displacement in the X, Y, and Z directions is constrained. Through this setting, the simulated mold is firmly fixed during the processing process and will not move or rotate. The pressure load condition is used to simulate the pressure distribution applied by the tool to the mold surface. In the model, the area on the mold surface that is subject to pressure is selected as the load boundary, and the pressure distribution data is imported and assigned to the corresponding nodes. Specifically, the pressure value P of each node is set. j As load input, it is applied to the corresponding surface nodes to simulate the pressure effect of the tool on the mold surface during the actual machining process.

[0112] After obtaining the stress values ​​for each node, the system plots the stress distribution map of the mold surface using different colors or grayscale values, visually demonstrating the differences in stress intensity across regions. For example, in a typical machining scenario, the cutting area at the mold edge might appear bright red (corresponding to high stresses above 500 MPa), while the central support area appears dark blue (lower than 100 MPa). Coordinate extraction does not directly analyze image colors, but rather relies on the raw node data stored when generating the stress distribution map. Each node has precise coordinates in the 3D model, and its stress values ​​are categorized and stored according to threshold ranges. When the preset stress threshold is 300 MPa, the system automatically iterates through all node data, filters out nodes with stress values ​​exceeding this threshold, and extracts their spatial coordinates (e.g., X: 25.3mm, Y: 48.7mm, Z: 12.1mm). For example, if high stresses of 420-480 MPa are detected at 10 adjacent nodes at a mold corner, the system will label these node coordinates as "Stress Concentration Location A," highlighting them with a flashing red box in the 3D model and generating a coordinate list for engineers to locate and analyze. This data-driven approach allows for the identification of stress concentration areas while retaining the intuitiveness of visualization and ensuring accurate coordinate positioning. This process allows for the accurate identification of areas where stress concentration may occur during mold processing.

[0113] This step can reveal the stress characteristics of the mold during the processing process, help identify potential stress concentration problems, and provide a basis for optimizing processing parameters, adjusting support structures, and reducing the risk of residual deformation.

[0114] In step S13 , the thermal expansion coefficient and the ambient temperature of the mold at the stress concentration location are obtained.

[0115] In a specific embodiment, obtaining the thermal expansion coefficient and the ambient temperature of the mold at the stress concentration location includes:

[0116] Based on the three-dimensional coordinate information of the stress concentration position, the lattice thermal expansion characteristic parameters of the corresponding material are matched in a preset material database, and the thermal expansion coefficient in the stress gradient direction is calculated in combination with an anisotropic interpolation algorithm;

[0117] The ambient temperature is collected in real time by a distributed optical fiber temperature sensor array. The sensor array is arranged with a high-density temperature measurement point array in the area where the stress is concentrated, and a low-density temperature measurement point array in the remaining areas.

[0118] Specifically, first, based on the three-dimensional coordinate information of the stress concentration location, the system retrieves the microscopic thermal expansion characteristic parameters of the material corresponding to that location from a preset material database, specifically the thermal expansion differential data of the material lattice structure in different directions. Because the mold material exhibits anisotropic thermal expansion behavior in different stress gradient directions, the system dynamically correlates the geometric relationship between the stress distribution direction and the principal axis of the material lattice, using an interpolation algorithm to comprehensively calculate the actual thermal expansion coefficient of the region. This process fully considers the coupling relationship between the material's microstructure and the macroscopic stress direction, ensuring that the calculated thermal expansion coefficient accurately reflects the material's true deformation characteristics under complex loads.

[0119] Secondly, the ambient temperature is collected through a distributed fiber optic temperature sensor array. The sensor network layout strategy is deeply coupled with the stress distribution characteristics: in areas with concentrated stress and significant gradient changes, a high-density dot matrix layout is adopted to capture subtle temperature fluctuations. The spacing between each sensor node is extremely small, forming a dense temperature measurement network; in areas with relatively flat stress distribution, a low-density dot matrix layout is adopted, which can not only cover the full-area monitoring needs but also optimize the allocation of sensor resources. The fiber optic sensor scans the mold surface temperature field in real time and inputs the collected raw temperature data into a dynamic calibration model. This model combines the material's thermal conductivity characteristics and the spatial distribution characteristics of the stress gradient to compensate for the sensor response delay, ultimately generating high-precision continuous temperature field distribution data.

[0120] This technical approach, which integrates material properties with real-time sensing, effectively solves the insufficient matching accuracy of thermal expansion coefficient and ambient temperature data in traditional methods.

[0121] In step S14, the thermal expansion coefficient and the ambient temperature are input into a pre-established linear regression model to predict the mold deformation trend.

[0122] In a specific embodiment, inputting the thermal expansion coefficient and the ambient temperature into a pre-established linear regression model to predict the mold deformation trend includes:

[0123] Inputting the thermal expansion coefficient and the ambient temperature into a preset linear regression model to predict deformation trend data of the mold at the stress concentration location;

[0124] The deformation trend data is quantified using a preset deformation trend analysis algorithm, and the quantified deformation trend data is classified using a gradient boosting tree algorithm to determine the stability level of the deformation trend and obtain a stability level.

[0125] According to the stability level, combined with a preset material property database, query the elastic modulus and Poisson's ratio that match the mold material, and determine the deformation constraint conditions of the mold material at different stability levels;

[0126] According to the deformation constraint conditions, a predetermined Bayesian optimization algorithm is used to perform a probability prediction on the deformation distribution of the mold at the stress concentration position to obtain a deformation distribution probability;

[0127] Using the deformation trend data and the deformation distribution probability as a mold deformation trend;

[0128] The linear regression model is as follows:

[0129] ΔD=α·ΔT·(β1κ+β2)+β3(ΔT) 2 +∈

[0130]

[0131] Among them, ΔD represents the deformation trend data, α represents the thermal expansion coefficient, ΔT represents the ambient temperature, κ represents the stress concentration factor, σ max represents the maximum nodal stress value, σ avg represents the weighted average stress of the node, n represents the total number of nodes, σ j represents the stress value of the jth node, dA j A represents the area element of the jth node. total represents the total surface area of ​​the mold, β1, β2, and β3 represent regression coefficients, and the units of β1, β2, and β3 are m, m, and m / K, respectively. 2 , β1, β2 and β3 are fitted by the Levenberg-Marquardt algorithm, and ∈ represents the preset model residual term.

[0132] Specifically, a linear regression model is first used to establish a mathematical relationship between the thermal expansion coefficient α, the ambient temperature ΔT, and the mold deformation trend ΔD. The deformation trend is then quantitatively predicted using the stress concentration factor κ. The linear regression model formula is:

[0133] ΔD=α·ΔT·(β1κ+β2)+β3(ΔT) 2 +∈

[0134] Among them, ΔD represents the deformation trend data, α represents the thermal expansion coefficient, ΔT represents the ambient temperature, and κ represents the stress concentration factor. The calculation formula is: σ max represents the maximum nodal stress value, σ avg Represents the weighted average stress of the node, and the calculation formula is Where n represents the total number of nodes, σ j represents the stress value of the jth node, dA j A represents the area element of the jth node. totalrepresents the total surface area of ​​the mold. β1, β2, and β3 are regression coefficients obtained by fitting the Levenberg-Marquardt algorithm, and ∈ represents the preset model residual term.

[0135] During the implementation, the thermal expansion coefficient α and ambient temperature ΔT are first substituted into a linear regression model to calculate the deformation trend data ΔD. Next, ΔD is quantified using a pre-set deformation trend analysis algorithm. The quantified data is then classified using a gradient boosting tree algorithm to determine the stability level of the deformation trend. The gradient boosting tree algorithm constructs multiple decision trees to gradually optimize classification accuracy and ultimately outputs a stability level. The stability level characterizes the severity of the mold deformation trend and is categorized as low, medium, and high.

[0136] Based on the stability level and a pre-set material property database, the elastic modulus E and Poisson's ratio ν that match the mold material are searched to determine the deformation constraints of the mold material at different stability levels. The deformation constraints include the maximum allowable deformation and the deformation distribution range.

[0137] Then, a pre-set Bayesian optimization algorithm is used to probabilistically predict the mold deformation distribution at the stress concentration location. By maximizing the objective function, the Bayesian optimization algorithm finds the optimal solution for the deformation distribution and ultimately outputs the deformation distribution probability. Finally, the deformation trend data ΔD and the deformation distribution probability are used as the output of the mold deformation trend.

[0138] This step accurately predicts mold deformation trends at stress concentration locations, providing a scientific basis for optimizing processing parameters and mold design. By combining a linear regression model with a Bayesian optimization algorithm, we can comprehensively consider thermodynamic and mechanical factors and precisely quantify the mold's deformation behavior, thereby reducing deformation risks during processing and improving mold accuracy and service life.

[0139] In step S15, the optimized cutting parameters are obtained based on the initial cutting parameters and in combination with a cutting parameter optimization objective function; wherein the cutting parameter optimization objective function includes the relationship between the cutting parameters and tool wear, surface quality and residual stress.

[0140] In a specific embodiment, the optimized cutting parameters are obtained based on the initial cutting parameters in combination with a cutting parameter optimization objective function; wherein the cutting parameter optimization objective function includes the relationship between the cutting parameters and tool wear, surface quality and residual stress, including:

[0141] The cutting parameter optimization objective function is as follows:

[0142]

[0143] Among them, V c Indicates the cutting speed in the cutting parameters, f indicates the feed rate in the cutting parameters, R a Represents the surface roughness of the mold, measured by white light interferometer, σ res represents the residual stress, which is measured by X-ray diffraction, vb represents the tool wear, ω1, ω2, ω3, ω4 and ω5 represent the preset dynamic weight coefficients, V c0 represents the initial cutting speed in the initial cutting parameters, f0 represents the initial feed in the initial cutting parameters, F opt Represents the output value of the objective function, R a,max , σ yickl and VB max Indicates the preset normalized standard value.

[0144] Specifically, the cutting parameter optimization objective function is as follows:

[0145]

[0146] Among them, V c Indicates the cutting speed in the cutting parameters, f indicates the feed rate in the cutting parameters, R a Represents the surface roughness of the mold, measured by white light interferometer, σ res represents the residual stress, which is measured by X-ray diffraction, vb represents the tool wear, ω1, ω2, ω3, ω4 and ω5 represent the preset dynamic weight coefficients, V c0 represents the initial cutting speed in the initial cutting parameters, f0 represents the initial feed in the initial cutting parameters, F opt Represents the output value of the objective function, R a,max , σ yickl and VB max Indicates the preset normalized standard value.

[0147] This objective function comprehensively considers the influence of cutting speed, feed rate, surface roughness, residual stress and tool wear on machining quality, and weights the contribution of each factor through dynamic weight coefficients ω1, ω2, ω3, ω4 and ω5, and finally obtains the optimized cutting parameters. The dynamic weight coefficients ω1, ω2, ω3, ω4 and ω5 are pre-set according to the actual machining needs and process requirements to adjust the relative importance of each factor. For example, if the surface quality R a If tool life is the main concern, the value of ω3 is relatively large; if tool life is the main concern, the value of ω5 is relatively large.

[0148] First, according to the initial cutting parameters V c0 and f0, measuring the surface roughness R under the current processing conditions a , residual stress σ resand tool wear vb. Then, V c0 ,f0,R a , σ res and vb are substituted into the cutting parameter optimization objective function F opt Then, the objective function F is calculated using the gradient descent method. opt Perform minimization to obtain the optimized cutting speed V c And feed rate f. In the optimization process, the objective function F opt The smaller the value, the better the comprehensive performance of the cutting parameters.

[0149] By establishing a cutting parameter optimization objective function and solving it using an optimization algorithm, we can comprehensively consider multiple factors, including tool wear, surface quality, and residual stress, to determine the optimal cutting parameter combination. This not only improves machining efficiency and mold quality, but also extends tool life. The optimized model is highly robust and versatile, making it suitable for optimizing cutting parameters across a wide range of materials and machining conditions.

[0150] In step S16, according to the deformation trend of the mold, a dynamic compensation algorithm is used to adjust the supporting structure parameters of the mold and the optimized cutting parameters to obtain the optimal cutting parameters.

[0151] In a specific embodiment, the method of adjusting the mold support structure parameters and the optimized cutting parameters using a dynamic compensation algorithm according to the mold deformation trend to obtain the optimal cutting parameters includes:

[0152] Smoothing and feature extraction are performed on the mold deformation trend to obtain deformation features;

[0153] Using a principal component analysis algorithm to perform correlation calculation on the deformation characteristics and pre-stored historical data of the support structure, a quantitative mapping relationship model between the mold deformation trend and pre-stored support structure parameters is established;

[0154] Based on the quantitative mapping relationship model, the support structure parameters are adjusted through the PID algorithm to obtain the preliminary adjusted support structure parameters;

[0155] Based on the initially adjusted support structure parameters and the optimized cutting parameters, the cutting parameters are updated and calculated using a linear interpolation method, and the performance of the updated cutting parameter set is predicted using a random forest algorithm to obtain a performance prediction result;

[0156] When the performance prediction result does not meet the preset performance threshold, iteratively updating the cutting parameters that do not meet the conditions by using a gradient descent algorithm;

[0157] When the performance prediction result meets a preset performance threshold, the most recently updated cutting parameters are used as the optimal cutting parameters.

[0158] Specifically, the mold deformation trend is first smoothed and feature extracted to obtain deformation features. Smoothing eliminates noise interference, and feature extraction captures characteristic information from the deformation for subsequent analysis. Deformation feature extraction effectively captures deformation trends and patterns.

[0159] Next, a principal component analysis (PCA) algorithm is used to correlate deformation characteristics with pre-stored historical support structure data, establishing a quantitative mapping model between mold deformation trends and support structure parameters. The core of the PCA algorithm lies in revealing the essential relationships between multidimensional parameters through data dimensionality reduction and feature extraction. The system first integrates support structure parameters from historical machining data (such as support point spatial coordinates, stiffness distribution, and dynamic damping coefficient) with corresponding deformation monitoring data (such as regional displacement, strain gradient, and vibration spectrum) into a unified multidimensional dataset. The data is then standardized and preprocessed to eliminate dimensional differences. The algorithm then reconstructs the data space through an orthogonal transformation, calculates the intrinsic correlations between parameters in each dimension, and identifies the potential characteristic dimensions that contribute most to the overall data variability. These dimensions are essentially mathematical representations of the nonlinear coupling between support parameters and deformation characteristics. For example, the spatial distribution pattern of support point density may be strongly coupled with lateral displacement in the mold's midsection, while stiffness gradient changes have a higher weight on torsional deformation at the edges. The algorithm calculates the variance contribution of each latent dimension and selects principal components whose cumulative explanatory power exceeds a set threshold. Based on this, it constructs a weighted mapping matrix between support parameters and deformation indicators. This matrix transforms complex multivariate interactions into quantifiable linear combination patterns. For example, a principal component can clearly represent the physical law that "for every proportional increase in support stiffness, the corresponding regional deformation variable exponentially decays." Finally, the system uses backprojection to map the abstract relationships in the principal component space back to the original parameter dimensions, forming a predictive model for the effect of support parameter adjustment on deformation trends. This model allows any new set of support structure parameters to be quickly derived from the expected deformation distribution, providing directional guidance for parameter optimization in the dynamic compensation algorithm. Then, based on the quantified mapping relationship model, the support structure parameters are adjusted using the PID algorithm to obtain preliminary adjusted support structure parameters. The PID algorithm dynamically adjusts the support structure parameters by calculating the error signal and its integral and differential functions to suppress deformation trends and achieve stability control.

[0160] During the parameter optimization process, the system gradually approaches the optimal cutting parameter combination through the collaboration of multiple algorithms. Based on the initially adjusted support structure parameters (such as the stiffness distribution and damping characteristics of the support points), the linear interpolation method first constructs the spatial mapping relationship of the cutting parameters in the historical parameter database - according to the spatial relative position of the current optimized cutting parameters and the historical optimal parameters, a smooth transition parameter sequence is generated along the process boundaries such as material removal rate and cutting force constraints. For example, if the cutting speed and feed rate of adjacent parameter points in the historical data are linearly correlated, the algorithm will generate a new parameter combination along this trend to ensure that the adjustment range of core parameters such as cutting depth and feed rate is always within the physically feasible domain, avoiding tool chatter or thermal shock caused by parameter step changes.

[0161] The random forest algorithm then performs a multi-objective performance simulation on the updated parameter set. By integrating hundreds of pre-trained decision trees (each trained on machining data from different process scenarios, covering the cutting characteristics of a variety of materials such as aluminum alloys and titanium alloys), the algorithm simultaneously infers the potential performance of parameter combinations based on three dimensions: surface roughness, residual stress, and tool wear. Each decision tree independently analyzes the feasibility of the parameter combination—for example, one tree focuses on evaluating tool wear rate at high feed rates, while another predicts the nonlinear impact of sudden changes in cutting speed on residual stress. A weighted voting mechanism is used to output a comprehensive performance score. If the score falls below a preset threshold (e.g., surface roughness Ra must be below 0.8μm, tool wear must be controlled within 0.02mm / hour), a gradient descent algorithm immediately intervenes to optimize the parameters. In the three-dimensional parameter space composed of cutting speed, feed rate, and depth of cut, the algorithm searches along the negative gradient of the objective function (a composite indicator that comprehensively represents surface quality, stress level, and tool life). It avoids local extremes through an adaptive step-size adjustment strategy, while simultaneously monitoring process stability constraints (e.g., cutting force sudden change threshold and thermal load safety margin) in real time. After each parameter iteration, the system automatically triggers a new round of random forest prediction and gradient correction until the prediction results for three consecutive iterations meet all performance thresholds. At this point, the system dynamically binds the final parameter combination to the support structure parameters and generates a process package that includes parameter stability margin and anti-interference capability analysis, ensuring the robustness and repeatability of the optimal cutting parameters under complex working conditions. When the performance prediction results meet the preset performance threshold, the most recently updated cutting parameters are used as the optimal cutting parameters.

[0162] By optimizing and adjusting the support structure parameters and cutting parameters through dynamic compensation algorithms, mold deformation can be effectively suppressed and processing quality and stability can be improved.

[0163] In step S17 , a target tool trajectory path is generated according to the support structure parameters and the optimal cutting parameters, so as to process the production line mold according to the target tool trajectory path.

[0164] In a specific embodiment, generating a target tool trajectory path according to the support structure parameters and the optimal cutting parameters, so as to process the production line mold according to the target tool trajectory path, includes:

[0165] Performing mesh analysis on the three-dimensional finite element model based on the support point density and stiffness distribution data in the support structure parameters to divide the force-constrained area and the free processing area;

[0166] Based on the three-dimensional geometric features of the free machining area and the cutting speed, feed rate and cutting depth in the optimal cutting parameters, an initial tool trajectory path is generated using a B-spline curve algorithm;

[0167] Performing collision detection on the initial tool trajectory path and the force-constrained area, and if a tool support interference area exists, optimizing and reconstructing the trajectory of the tool support interference area through a local path adjustment algorithm;

[0168] Importing the optimized tool trajectory path into the finite element simulation model, and generating a tool load dynamic curve in combination with the support structure parameters;

[0169] When the tool load exceeds the preset stress threshold, the trajectory curvature radius and feed angle are iteratively adjusted through a genetic algorithm to generate a smooth path that meets the residual stress constraint.

[0170] Finally, the path segments within each optimization parameter domain are integrated to generate a target tool trajectory path with continuous tangent vectors, which is then converted into NC code for machining equipment to perform machining.

[0171] Specifically, a mesh analysis is first performed on the 3D finite element model based on the support point density and stiffness distribution data from the support structure parameters, dividing the force-constrained area and the free processing area. This mesh analysis clarifies the influence of the support structure during mold processing, identifies the force-constrained areas that need to be avoided and the areas that can be freely processed, and provides spatial constraint information for tool path planning.

[0172] Next, an initial tool path is generated using a B-spline algorithm based on the 3D geometric features of the free machining area and the optimal cutting parameters of cutting speed, feed rate, and depth of cut. The B-spline algorithm generates a smooth and continuous path based on the geometric features, ensuring tool stability and machining efficiency. This initial tool path fully considers the shape of the free machining area and the cutting parameters.

[0173] The initial tool path is then subjected to collision detection against the force-constrained area. If interference between the tool path and the support structure is detected, a local path adjustment algorithm reconstructs the trajectory through dynamic interpolation and geometric constraint solving in three-dimensional space. The algorithm first extracts the geometric boundary data of the interference area and, combined with the stiffness distribution of the support points, constructs a multi-level obstacle avoidance buffer zone. This buffer zone, centered around the support structure surface, extends outward, forming a three-dimensional protective layer that prohibits the tool center point from entering. Its thickness is determined by the elastic modulus of the support material, the tool radius, and a safety margin. Subsequently, leveraging the adjustable control points of the B-spline curve, the interfering segment is topologically deformed while maintaining overall path continuity. By introducing a virtual gravitational field model, the algorithm drives the control points of the original trajectory perpendicular to the interference direction while constraining the relative curvature change rate of adjacent control points to ensure a smooth transition around the support structure. During this process, the algorithm dynamically calculates the balance between cutting load and material removal rate in real time. It adaptively adjusts the feed gradient distribution to compensate for the loss in machining efficiency caused by path extension. It also utilizes the stiffness data of the support area to optimize the tool roll angle, avoiding mechanical collisions while maximizing cutting stability. The stress distribution characteristics of the reconstructed local path are verified through finite element simulation. If the residual stress peak still exceeds the threshold, a secondary optimization cycle is triggered until all process constraints are met. The final optimized trajectory appears as a continuous spline curve with a variable curvature radius on a microscopic scale, which not only avoids physical interference but also ensures the balanced optimization of machining surface quality and tool life through the coordinated adjustment of cutting parameters. The collision detection algorithm (Gilbert-Johnson-Keerthi algorithm) can accurately identify the interference area between the tool and the support structure. The local path adjustment algorithm combines geometric constraints and kinematic characteristics to re-plan the trajectory of the interference area to avoid collision between the tool and the support structure.

[0174] The optimized tool path is imported into the finite element simulation model and combined with the support structure parameters to generate a dynamic tool load curve. The finite element simulation model can simulate the forces acting on the tool during machining and generate a dynamic curve of tool load changes over time, providing data support for further path optimization.

[0175] When tool loads exceed safety limits, a genetic algorithm optimizes path parameters by simulating biological evolution. The system encodes the path curvature radius and feed angle into a "genetic sequence," with each parameter combination forming a candidate solution. The algorithm first randomly generates hundreds of initial parameter combinations (populations) and calculates the fitness value of each group through finite element simulation. This value comprehensively evaluates the path's residual stress level, load fluctuation amplitude, and smoothness indicators.

[0176] During the iteration process, parameter combinations with high fitness (such as paths with residual stress below the threshold and continuously changing curvature) will be retained first, and the advantageous features will be passed to offspring through the "chromosome crossover" operation (such as fusing the curvature features of a low-stress trajectory with the feed angle of another smooth path). At the same time, controllable "gene mutations" are introduced to randomly fine-tune certain parameter values ​​with a specific probability (such as increasing the curvature radius by 5%) to avoid falling into local optimal solutions. After dozens of generations of evolution, the algorithm has screened out the optimal parameter combination (trajectory curvature radius, feed angle, cutting speed, feed rate, cutting depth, tool inclination angle) that can stabilize the cutting load in a safe range and significantly reduce residual stress, and finally form an optimized trajectory that takes into account both processing stability and quality constraints. This global search strategy based on population iteration is particularly suitable for complex path optimization scenarios with multi-parameter coupling.

[0177] Finally, the path segments within each optimized parameter domain are integrated to generate a target tool path with continuous tangent vectors. This is then converted into NC code for the machining equipment to execute the process. This integration of path segments ensures overall continuity and smoothness of the trajectory, avoiding sudden changes or pauses during tool movement. The target tool path, converted into NC code, directly drives the machining equipment to precisely complete the mold.

[0178] Among them, the process of converting the optimized tool trajectory path into NC code is realized through the post-processor: the system first converts the geometric features of the trajectory path (coordinate points, tangent vector direction, curvature parameters) into basic motion instructions (such as G01 linear interpolation, G02 / G03 circular interpolation) according to the machine tool coordinate system, and encodes the process parameters of cutting speed and feed rate into corresponding F / S values. Through the dynamic interpolation algorithm, the continuous B-spline path is discretized into micro-segments executable by the machine tool, and the trajectory offset is automatically corrected based on the tool radius compensation (G41 / G42). The final generated NC code strictly follows the syntax specifications of the target machine tool (such as FANUC or SIEMENS system), contains complete machining cycle instructions and safety verification parameters, and directly drives the CNC system to perform high-precision machining.

[0179] By combining support structure parameters with optimized cutting parameters, a highly accurate target tool path is generated. This effectively avoids interference between the tool and support structures, optimizes tool load distribution, reduces the effects of residual stress, and improves machining efficiency and quality. Finite element simulation models and genetic algorithms play a key role in the path optimization process, ensuring the rationality and feasibility of the path. The generated NC code can be directly used in actual machining, achieving a seamless transition from path planning to machining execution.

[0180] Reference Figure 2 The second embodiment of the present invention provides an intelligent control system for mold processing of a production line, comprising:

[0181] The data acquisition module is used to obtain the initial cutting parameters of the tool and the surface pressure data of the mold to determine the pressure distribution state;

[0182] A stress analysis module is used to calculate the stress distribution of each node of the mold through a preset three-dimensional finite element model according to the pressure distribution state and determine the stress concentration location;

[0183] A data matching module is used to obtain the thermal expansion coefficient and ambient temperature of the mold at the stress concentration position;

[0184] a deformation trend analysis module, configured to input the thermal expansion coefficient and the ambient temperature into a pre-established linear regression model to predict the mold deformation trend;

[0185] a cutting parameter optimization module, configured to obtain optimized cutting parameters based on the initial cutting parameters and a cutting parameter optimization objective function; wherein the cutting parameter optimization objective function includes the relationship between the cutting parameters and tool wear, surface quality, and residual stress;

[0186] An optimal parameter analysis module, configured to adjust the mold support structure parameters and the optimized cutting parameters using a dynamic compensation algorithm according to the mold deformation trend to obtain optimal cutting parameters;

[0187] The processing path generation module is used to generate a target tool trajectory path according to the support structure parameters and the optimal cutting parameters, so as to process the production line mold according to the target tool trajectory path.

[0188] It should be noted that the intelligent control device for production line mold processing provided by an embodiment of the present invention is used to execute all process steps of the intelligent control method for production line mold processing of the above embodiment. The working principles and beneficial effects of the two correspond one to one, so they will not be repeated here.

[0189] An embodiment of the present invention further provides an electronic device. The electronic device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as an intelligent control program for production line mold processing. When the processor executes the computer program, the steps of the above-mentioned intelligent control method for production line mold processing are implemented, such as Figure 1 Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are realized, such as the intelligent control module for mold processing of a production line.

[0190] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, and the instruction segments are used to describe the execution process of the computer program in the electronic device.

[0191] The electronic device may be a computing device such as a desktop computer, notebook, PDA, or smart tablet. The electronic device may include, but is not limited to, a processor and memory. Those skilled in the art will appreciate that the aforementioned components are merely examples of electronic devices and do not constitute a limitation of the electronic device. The electronic device may include more or fewer components than those described above, or a combination of certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, buses, and the like.

[0192] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the electronic device, connecting various parts of the entire electronic device using various interfaces and lines.

[0193] The memory can be used to store the computer programs and / or modules, and the processor realizes various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created based on the use of the mobile phone (such as audio data, a phone book, etc.). In addition, the memory can include a high-speed random access memory and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0194] Wherein, if the module / unit integrated in the electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by a computer program to instruct the relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, it can implement the steps of each of the above-mentioned method embodiments. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0195] It should be noted that the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive effort.

[0196] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. An intelligent control method for production line mold processing, characterized in that: include: Obtain the initial cutting parameters of the tool and the surface pressure data of the mold to determine the pressure distribution state; According to the pressure distribution state, the stress distribution of each node of the mold is calculated by a preset three-dimensional finite element model to determine the stress concentration location; Obtaining the thermal expansion coefficient and ambient temperature of the mold at the stress concentration location; Inputting the thermal expansion coefficient and the ambient temperature into a pre-established linear regression model to predict the mold deformation trend; According to the initial cutting parameters, combined with a cutting parameter optimization objective function, an optimized cutting parameter is obtained; wherein the cutting parameter optimization objective function includes a relationship between the cutting parameters and tool wear, surface quality and residual stress; According to the deformation trend of the mold, a dynamic compensation algorithm is used to adjust the supporting structure parameters of the mold and the optimized cutting parameters to obtain the optimal cutting parameters; A target tool trajectory path is generated according to the support structure parameters and the optimal cutting parameters, so as to process the production line mold according to the target tool trajectory path.

2. The intelligent control method for production line mold processing according to claim 1 is characterized in that: The obtaining of the initial cutting parameters of the tool and the surface pressure data of the mold and determining the pressure distribution state includes: After initializing the tool processing machine, obtaining the initial cutting parameters of the tool; The surface pressure data of the mold is obtained by using multiple pressure sensors covering various areas of the mold surface; The surface pressure data are classified and marked with different colors to generate a mold surface pressure distribution map, wherein the different colors in the mold surface pressure distribution map are used to represent the pressure distribution state of the mold surface.

3. The intelligent control method for production line mold processing according to claim 1 is characterized in that: The method of calculating the stress distribution of each node of the mold by using a preset three-dimensional finite element model according to the pressure distribution state and determining the stress concentration position includes: Performing a stress calculation operation on each node of the pressure distribution state in combination with the three-dimensional finite element model to obtain a stress value of each node; According to the stress values ​​of each node, a drawing operation is performed on the stress values ​​of each node to generate a stress distribution diagram; Performing a coordinate extraction operation on the stress distribution map to obtain the coordinates of a stress region having a stress value higher than a preset stress threshold as a stress concentration location; The calculation formula of the node stress value is as follows: Where, σ j represents the stress value of the jth node, n represents the total number of nodes, K ij The influence coefficient of the unit pressure of the i-th node on the stress of the j-th node is obtained through the three-dimensional finite element model, P i Indicates the pressure value of the i-th node, dA j The area element representing the j-th node is obtained by the three-dimensional finite element model; The process of setting up the three-dimensional finite element model includes: Scan and measure the mold using a scanner to obtain the surface geometric dimensions of the mold; Establishing an actual three-dimensional model of the mold according to the surface geometric dimensions; According to the actual three-dimensional model, a finite element meshing operation is performed on the actual three-dimensional model to obtain a three-dimensional finite element meshing model; wherein the finite element meshing includes tetrahedral element meshing; According to the three-dimensional finite element partition model, a boundary condition setting operation is performed on the three-dimensional finite element partition model to obtain a three-dimensional finite element model; wherein the boundary condition setting includes a fixed support condition and a pressure load condition.

4. The intelligent control method for production line mold processing according to claim 1, characterized in that: The obtaining of the thermal expansion coefficient and the ambient temperature of the mold at the stress concentration position includes: Based on the three-dimensional coordinate information of the stress concentration position, the lattice thermal expansion characteristic parameters of the corresponding material are matched in a preset material database, and the thermal expansion coefficient in the stress gradient direction is calculated in combination with an anisotropic interpolation algorithm; The ambient temperature is collected in real time by a distributed optical fiber temperature sensor array. The sensor array is arranged with a high-density temperature measurement point array in the area where the stress is concentrated, and a low-density temperature measurement point array in the remaining areas.

5. The intelligent control method for production line mold processing according to claim 4 is characterized in that: Inputting the thermal expansion coefficient and the ambient temperature into a pre-established linear regression model to predict the mold deformation trend includes: Inputting the thermal expansion coefficient and the ambient temperature into a preset linear regression model to predict deformation trend data of the mold at the stress concentration location; The deformation trend data is quantified using a preset deformation trend analysis algorithm, and the quantified deformation trend data is classified using a gradient boosting tree algorithm to determine the stability level of the deformation trend and obtain a stability level. According to the stability level, combined with a preset material property database, query the elastic modulus and Poisson's ratio that match the mold material, and determine the deformation constraint conditions of the mold material at different stability levels; According to the deformation constraint conditions, a predetermined Bayesian optimization algorithm is used to perform a probability prediction on the deformation distribution of the mold at the stress concentration position to obtain a deformation distribution probability; Using the deformation trend data and the deformation distribution probability as a mold deformation trend; The linear regression model is as follows: ΔD=α·ΔT·(β1κ+β2)+β3(ΔT) 2 +∈ Among them, ΔD represents the deformation trend data, α represents the thermal expansion coefficient, ΔT represents the ambient temperature, κ represents the stress concentration factor, σ max represents the maximum nodal stress value, σ avg represents the weighted average stress of the node, n represents the total number of nodes, σ j represents the stress value of the jth node, dA j A represents the area element of the jth node. total represents the total surface area of ​​the mold, β1, β2, and β3 represent regression coefficients, and the units of β1, β2, and β3 are m, m, and m / K, respectively. 2 , β1, β2 and β3 are fitted by the Levenberg-Marquardt algorithm, and ∈ represents the preset model residual term.

6. The intelligent control method for production line mold processing according to claim 1, characterized in that: The optimized cutting parameters are obtained based on the initial cutting parameters and in combination with a cutting parameter optimization objective function; wherein the cutting parameter optimization objective function includes the relationship between the cutting parameters and tool wear, surface quality and residual stress, including: The cutting parameter optimization objective function is as follows: Among them, V c Indicates the cutting speed in the cutting parameters, f indicates the feed rate in the cutting parameters, R a Represents the surface roughness of the mold, measured by white light interferometer, σ res represents the residual stress, which is measured by X-ray diffraction, vb represents the tool wear, ω1, ω2, ω3, ω4 and ω5 represent the preset dynamic weight coefficients, V c0 represents the initial cutting speed in the initial cutting parameters, f0 represents the initial feed in the initial cutting parameters, F opt Represents the output value of the objective function, R a,max , σ yickl and VB max Indicates the preset normalized standard value.

7. The intelligent control method for production line mold processing according to claim 1 is characterized in that: The method of adjusting the mold support structure parameters and the optimized cutting parameters using a dynamic compensation algorithm according to the mold deformation trend to obtain the optimal cutting parameters includes: Smoothing and feature extraction are performed on the mold deformation trend to obtain deformation features; Using a principal component analysis algorithm to perform correlation calculation on the deformation characteristics and pre-stored historical data of the support structure, a quantitative mapping relationship model between the mold deformation trend and pre-stored support structure parameters is established; Based on the quantitative mapping relationship model, the support structure parameters are adjusted through the PID algorithm to obtain the preliminary adjusted support structure parameters; Based on the initially adjusted support structure parameters and the optimized cutting parameters, the cutting parameters are updated and calculated using a linear interpolation method, and the performance of the updated cutting parameter set is predicted using a random forest algorithm to obtain a performance prediction result; When the performance prediction result does not meet the preset performance threshold, iteratively updating the cutting parameters that do not meet the conditions by using a gradient descent algorithm; When the performance prediction result meets a preset performance threshold, the most recently updated cutting parameters are used as the optimal cutting parameters.

8. The intelligent control method for production line mold processing according to claim 1, characterized in that: Generating a target tool trajectory path according to the support structure parameters and the optimal cutting parameters, so as to process the production line mold according to the target tool trajectory path, comprises: Performing mesh analysis on the three-dimensional finite element model based on the support point density and stiffness distribution data in the support structure parameters to divide the force-constrained area and the free processing area; Based on the three-dimensional geometric features of the free machining area and the cutting speed, feed rate and cutting depth in the optimal cutting parameters, an initial tool trajectory path is generated using a B-spline curve algorithm; Performing collision detection on the initial tool trajectory path and the force-constrained area, and if a tool support interference area exists, optimizing and reconstructing the trajectory of the tool support interference area through a local path adjustment algorithm; Importing the optimized tool trajectory path into the finite element simulation model, and generating a tool load dynamic curve in combination with the support structure parameters; When the tool load exceeds the preset stress threshold, the trajectory curvature radius and feed angle are iteratively adjusted through a genetic algorithm to generate a smooth path that meets the residual stress constraint. Finally, the path segments within each optimization parameter domain are integrated to generate a target tool trajectory path with continuous tangent vectors, which is then converted into NC code for machining equipment to perform machining.

9. An intelligent control system for production line mold processing, characterized in that: include: The data acquisition module is used to obtain the initial cutting parameters of the tool and the surface pressure data of the mold to determine the pressure distribution state; A stress analysis module is used to calculate the stress distribution of each node of the mold through a preset three-dimensional finite element model according to the pressure distribution state and determine the stress concentration location; A data matching module is used to obtain the thermal expansion coefficient and ambient temperature of the mold at the stress concentration position; a deformation trend analysis module, configured to input the thermal expansion coefficient and the ambient temperature into a pre-established linear regression model to predict the mold deformation trend; a cutting parameter optimization module, configured to obtain optimized cutting parameters based on the initial cutting parameters and a cutting parameter optimization objective function; wherein the cutting parameter optimization objective function includes the relationship between the cutting parameters and tool wear, surface quality, and residual stress; An optimal parameter analysis module, configured to adjust the mold support structure parameters and the optimized cutting parameters using a dynamic compensation algorithm according to the mold deformation trend to obtain optimal cutting parameters; The processing path generation module is used to generate a target tool trajectory path according to the support structure parameters and the optimal cutting parameters, so as to process the production line mold according to the target tool trajectory path.

10. An electronic device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the method implements the intelligent control method for production line mold processing according to any one of claims 1 to 8.

Images