Insulator production process simulation and optimization method based on green low carbon

Abstract

The application provides a green low-carbon-based insulator production process simulation and optimization method, which comprises the following steps: obtaining the three-dimensional structure, symmetric equivalent area, and sintering simulation results and post-sintering detection results of different loading orientations and kiln positions of the insulator, forming double-coordinate simulation residuals in the product coordinate system and the kiln coordinate system; determining the model attribution according to the migration state of the residuals with the product orientation, the kiln position, and the support relationship of the kiln furniture, and locking the material model parameters and the kiln boundary model parameters acting on the contact residual area of the kiln furniture; and constructing and dynamically updating the contact map with mechanical force transmission properties and contact heat conduction properties according to the sintering shrinkage state of the insulator. Through the double-coordinate simulation residual attribution, parameter locking, and dynamic modeling of the contact relationship, the local deviation caused by the change of the kiln furniture contact is no longer compensated by the material model or the kiln boundary model, the problem of model failure after transposition is solved, and the attribution accuracy of the process simulation is improved.

Description

Technical Field

[0001] This invention relates to the field of computer-aided process simulation technology, and in particular to a simulation and optimization method for insulator production process based on green and low-carbon principles. Background Technology

[0002] In insulator firing production, a simulation model is typically established based on the insulator's three-dimensional structure, kiln temperature, and kiln furniture support position. This model is used to calculate the insulator's temperature, shrinkage, and deformation during the heating, sintering, and cooling processes. After firing, the simulation results are compared with the actual detected dimensions, warpage, and crack locations. When discrepancies exist, the thermal conductivity parameters of the billet, sintering shrinkage parameters, kiln heat exchange conditions, or firing temperature trajectory are usually modified to ensure the simulation results match the detection results of the current batch. During this process, the support position and contact state between the insulator and the kiln furniture are typically treated as fixed conditions.

[0003] However, insulators continuously shrink during sintering, causing changes in their dimensions and relative position to the kiln furniture. Areas previously in contact with the kiln furniture may detach, while new contacts may form. These contacts may also change from relative sliding to fixed contact, or vice versa. As the contact relationship changes, the insulator's weight, shrinkage constraint force, and temperature stress are transmitted along the new contact points, and the location where the insulator dissipates heat from the kiln furniture also changes. If the load concentration point coincides with the temperature change concentration point, areas such as the skirt root and connection points are prone to continuous deformation.

[0004] Because current simulations still calculate based on the original support relationships, local deviations caused by contact changes are easily misjudged as inconsistent billet shrinkage or abnormal kiln temperature distribution. While further modifications to material parameters or firing temperatures may reduce the calculation deviation for the current furnace, they do not eliminate the actual contact changes. Instead, they may cause mutual compensation between the material model, kiln boundary model, and kiln furniture contact conditions. When the insulator's loading orientation or the kiln furniture is changed, the original compensation relationship fails, and similar defects will reappear.

[0005] Therefore, the technical problem to be solved by this application is: how to accurately determine the actual source of local deformation deviation during the firing process of insulators, and avoid misattributing deviations caused by changes in kiln furniture contact relationships to the material model or kiln boundary model. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by providing a simulation and optimization method for green and low-carbon insulator production processes, thereby solving the technical problems mentioned in the background section.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] The simulation and optimization method for green and low-carbon insulator manufacturing processes includes the following steps:

[0009] S1. Obtain the three-dimensional structure of the insulator, the symmetrical equivalent region, and the firing simulation results and post-firing test results of different loading orientations and kiln positions. Register the test points and simulation nodes to the product coordinate system and the kiln coordinate system to obtain the dual-coordinate simulation residual.

[0010] S2. Based on the state of the dual-coordinate simulation residual migrating with the product orientation, fixed in the kiln position, or migrating with the kiln furniture support relationship, determine whether it belongs to the material model, kiln boundary model, or kiln furniture contact model, and lock the material model parameters and kiln boundary model parameters acting on the residual area when it belongs to the kiln furniture contact model.

[0011] S3. Based on the sintering shrinkage state of the insulator and the kiln support structure, a dynamic model of the contact relationship is constructed to build a contact diagram with mechanical force transmission properties and contact heat conduction properties. The contact edges and contact states are updated as the sintering shrinks to obtain the migration results of the main load path and the contact heat conduction path.

[0012] S4. Correspond the contact change time with the stress relaxation stage and the structural fixation stage to determine the contact change sensitive stage. Determine whether the load concentration area and the temperature gradient area in the contact change sensitive stage coincide. When they coincide, optimize the generation of kiln furniture support position adjustment, loading orientation adjustment and temperature trajectory adjustment through strategy optimization.

[0013] S5. Based on the adjustment amount, adjust the kiln furniture support, loading orientation, and firing temperature trajectory to obtain the three-dimensional deformation after firing, kiln furniture contact marks, and crack initiation points. Update the contact diagram and dual-coordinate simulation residuals, and update the parameter locking status according to the model's classification.

[0014] S1 specifically includes: establishing a product coordinate system based on the insulator's design symmetry center, common axis, and loading positioning reference; determining symmetrical equivalent regions where the structural topology, nominal thickness, surface normal, and region boundaries all meet the corresponding matching limits after rotation or mirror transformation, forming a geometric reference data package; obtaining the firing simulation node coordinates and post-firing inspection point coordinates corresponding to different loading orientations and kiln positions; establishing point-pair relationships between inspection points whose distance does not exceed the point-pair limit and simulation nodes, and registering them to the product coordinate system and kiln coordinate system respectively, forming simulation and inspection corresponding datasets; calculating the coordinate difference of each effective point pair in the product coordinate system and kiln coordinate system, summarizing them into a region residual vector according to the corresponding grid area, forming a dual-coordinate simulation residual.

[0015] S2 specifically includes: screening dual-coordinate simulation residuals according to product model, billet batch, firing process version, testing equipment calibration version, and simulation mesh version, and making the same loading orientation correspond to different kiln positions and the same kiln position correspond to different loading orientations to form a residual migration comparison set; calculating the discrete amount of product coordinate alignment residuals and kiln coordinate alignment residuals respectively, assigning the residuals that migrate with the product orientation to the material model, assigning the residuals that are fixed to the kiln orientation to the kiln boundary model, and assigning the residuals that are inconsistent in both types of coordinates and migrate with the kiln furniture support relationship to the kiln furniture contact model; when the residuals are assigned to the kiln furniture contact model, locking the material model parameters and kiln boundary model parameters that coincide with the effective node set of the residuals in the area of ​​action, forming a model attribution and parameter locking data package.

[0016] S3 specifically includes: establishing an initial contact diagram based on the insulator mesh, the set of effective residual nodes, and the kiln support surface. The initial contact diagram includes product area nodes, kiln support nodes, internal transfer edges, and contact edges, and configuring mechanical force transmission attributes and contact heat conduction attributes for the contact edges; dynamically modeling the contact relationship based on the node temperature, shrinkage strain, node displacement, and normal reaction force at each calculation time during the sintering process; adding contact edges when the normal gap and normal reaction force meet the contact conditions, and deleting contact edges when the release conditions are met; updating the adhesion state or slip state based on the relationship between tangential force and friction constraint; determining the main load path and contact heat conduction path based on the mechanical and heat transmission quantities of each edge, and recording the path migration results caused by the addition or deletion of contact edges or the change of contact state.

[0017] S4 specifically includes: determining the stress relaxation stage and the structural fixation stage based on the high-temperature stress relaxation curve, sintering shrinkage curve, and high-temperature load-bearing curve of the same batch of billets, and defining the time interval between the two as the contact change sensitive stage; defining the grid cells where the contact load density reaches the allowable contact load density at the current temperature as the load concentration region, and defining the grid cells where the temperature gradient reaches the allowable temperature gradient as the temperature gradient region, and determining the overlap state based on whether the two types of regions have common grid cells; when overlap occurs, generating adjustment candidate combinations within the constraints of kiln furniture structure, kiln loading positioning, and kiln control, and filtering the combinations through strategy optimization to make the contact change leave the contact change sensitive stage or eliminate the overlap state, and outputting the process optimization adjustment data package.

[0018] S5 specifically includes: verifying the process optimization adjustment data package and the product model, kiln furniture number, model version number, and firing process version of the furnace to be executed; when the verification is consistent and the adjustment amount meets the control limits of the kiln furniture, kiln loading equipment, and kiln, the kiln furniture support position, loading orientation, and firing temperature trajectory are adjusted to form a firing execution record; after the insulator is fired and cooled to the detection conditions, the surface coordinates and crack images after firing are collected, and the local morphological changes that share the same grid unit with the projection area of ​​the kiln furniture support surface are identified as kiln furniture contact imprints, forming an optimization result detection data package; according to the correspondence between the predicted contact edge and the kiln furniture contact imprint, contact edges are added or deleted and the contact area is corrected, the dual-coordinate simulation residual and model attribution are obtained again, and the parameter lock status is updated accordingly.

[0019] The beneficial effects of this invention are as follows:

[0020] This invention addresses the issues of misjudging the source of deviation and modeling failure after relocation by employing dual-coordinate simulation residual attribution, parameter locking, and dynamic modeling of contact relationships. This prevents local deviations caused by changes in kiln furniture contact from being compensated for by material models or kiln boundary models, thus improving the accuracy of attribution and the stability across kiln cycles in process simulation. By establishing a product coordinate system, symmetrical equivalent regions, and simulation-detection point pairs, and forming dual-coordinate simulation residuals in both the product and kiln coordinate systems, a unified comparison benchmark is established for data under different loading orientations and kiln positions. This reduces registration and regional mismatch errors, improving the consistency of residual calculations.

[0021] This invention utilizes a comparative set of residual migration data from different loading orientations and kiln positions to determine model affiliation and lock the update permissions of incompatible parameters based on residual migration patterns. This prevents error compensation between the material model, kiln boundary model, and kiln furniture contact model, improving the stability and repeatability of model calibration. By constructing a contact diagram that combines mechanical force transmission and contact thermal conductivity properties, and updating the contact edges, contact states, and action transmission paths as sintering shrinks, continuous characterization of support detachment, new contact formation, and stick-slip transition is achieved. This reduces simulation errors caused by fixed contact boundaries and improves the accuracy of deformation prediction.

[0022] This invention maps contact changes to stress relaxation and structural fixation stages, and optimizes the overlap between load concentration areas and temperature gradient areas. This provides clear triggering criteria for adjustments to supports, orientations, and temperature trajectories, reducing the superposition of local loads and temperature rises, and controlling additional heating energy consumption. After optimization, the invention detects three-dimensional deformation, kiln furniture contact imprints, and crack initiation points, updating the contact diagram, dual-coordinate simulation residuals, and parameter locking status accordingly. This ensures continuous feedback of execution results to the next batch of dynamic modeling, suppressing model error accumulation and improving simulation reliability and product consistency in continuous production. Attached Figure Description

[0023] Figure 1 This is a flowchart of the simulation and optimization method for insulator production process based on green and low-carbon technology, as presented in this invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Example: Figure 1 As shown, this embodiment provides a simulation and optimization method for insulator manufacturing processes based on green and low-carbon principles, including the following steps:

[0026] S1. Obtain the three-dimensional structure of the insulator, the symmetrical equivalent region, and the firing simulation results and post-firing test results of different loading orientations and kiln positions. Register the test points and simulation nodes to the product coordinate system and the kiln coordinate system to obtain the dual-coordinate simulation residual.

[0027] S2. Based on the state of the dual-coordinate simulation residual migrating with the product orientation, fixed in the kiln position, or migrating with the kiln furniture support relationship, determine whether it belongs to the material model, kiln boundary model, or kiln furniture contact model, and lock the material model parameters and kiln boundary model parameters acting on the residual area when it belongs to the kiln furniture contact model.

[0028] S3. Based on the sintering shrinkage state of the insulator and the kiln support structure, a dynamic model of the contact relationship is constructed to build a contact diagram with mechanical force transmission properties and contact heat conduction properties. The contact edges and contact states are updated as the sintering shrinks to obtain the migration results of the main load path and the contact heat conduction path.

[0029] S4. Correspond the contact change time with the stress relaxation stage and the structural fixation stage to determine the contact change sensitive stage. Determine whether the load concentration area and the temperature gradient area in the contact change sensitive stage coincide. When they coincide, optimize the generation of kiln furniture support position adjustment, loading orientation adjustment and temperature trajectory adjustment through strategy optimization.

[0030] S5. Based on the adjustment amount, adjust the kiln furniture support, loading orientation and firing temperature trajectory, obtain the three-dimensional deformation after firing, kiln furniture contact imprint and crack initiation point, update the contact diagram and dual-coordinate simulation residual, and update the parameter locking state according to the model's classification.

[0031] S1 specifically includes the following sub-steps:

[0032] S110. Establish the geometric datum and symmetrical equivalent region of the insulator. The product design system reads the product model, product number, three-dimensional structure file, nominal dimensions, dimensional tolerances, loading positioning datum, and pre-firing structural state of the insulator to be analyzed. The three-dimensional structure file is uniformly converted into solid or closed surface data expressed in mm. When the three-dimensional structure is obtained by scanning reconstruction, the zero point and scale of the scanning equipment are corrected using standard dimensional parts that have been calibrated by metrology, and isolated points, duplicate surfaces, normal inversion surfaces, and fragmented surfaces with an area smaller than the grid resolution are deleted.

[0033] The product coordinate system originates from the insulator design symmetry center, and uses the common axis of the head connection area, the shed center, and the bottom connection area as the vertical axis. The other two axes are determined by the loading positioning reference. If the design documents do not specify the common axis, at least two mutually spaced coaxial section centers are fitted. If the maximum distance between the fitted axis and each section center does not exceed the geometric reference tolerance, the fitted axis is used and deemed valid. The geometric reference tolerance is calculated using the root mean square formula, which is equal to the arithmetic square root of the sum of the squares of the dimensional tolerance and the scanning calibration error. If the tolerance is exceeded, subsequent processing of the current product is stopped, and the process returns to the design model for verification.

[0034] The symmetrical equivalent region refers to the insulator surface region whose structural topology, nominal thickness, surface normal, and region boundary can correspond to each other after rotation around the common axis or mirroring about the design symmetry plane. A complete shed unit, connection unit, or other repeating structural unit is selected as the reference region. The corresponding region is generated according to the rotation index or mirroring relationship in the design document. The distance, normal angle, and thickness difference between the corresponding points after transformation are compared with the matching limit determined by the product size tolerance and the scanning equipment calibration result. When all are less than or equal to the corresponding matching limit, they are classified into the same symmetrical equivalent region. When any one exceeds the matching limit, it is marked as a non-equivalent region.

[0035] Each symmetric equivalent region is assigned a unique number, and the region nodes, region boundaries, region center, surface normal, belonging structural element, and coordinate transformation matrix between the region and the reference region are recorded.

[0036] Output geometric reference data package. The data fields include product model, product number, product coordinate system, loading positioning reference, common axis, symmetric equivalent region number, region node set, non-equivalent region mark and coordinate transformation matrix. It serves as a unified geometric reference for S120 to register simulation results with post-burning test results, S130 to calculate dual-coordinate simulation residuals and S310 to perform dynamic modeling of contact relationships.

[0037] S120. Generate simulation-detection datasets for different loading orientations and kiln positions. The production execution system reads the furnace number, product model, product number, billet batch, loading orientation, kiln car number, kiln position number, and firing process version. The kiln control system reads the temperature setting trajectory and actual temperature records for each temperature zone. The firing simulation platform reads the simulation output corresponding to the same product number, billet batch, and firing process version.

[0038] As an executable implementation method, the firing simulation platform adopts a transient thermo-mechanical coupling model, discretizing the three-dimensional structure in S110 into a mesh with unique node and element numbers. Based on the kiln temperature trajectory, the thermophysical parameters of the billet, and the sintering shrinkage curve of the same batch of samples, the nodal temperature, shrinkage strain, nodal displacement, and predicted coordinates under the post-firing cooling state at each calculation time are obtained sequentially. When using other discretization methods, the input and output meanings of nodal temperature, nodal displacement, and regional shrinkage should be kept consistent.

[0039] After the insulators are removed from the kiln and cooled to the allowable temperature of the testing equipment, with temperature changes not exceeding the stable range specified in the testing specifications, the surface coordinates after firing are collected using a 3D scanning device or coordinate measuring device calibrated with standard-sized parts. Crack locations are collected using a surface image detection device. 3D coordinates and displacements are uniformly expressed in mm, temperature in °C, and loading orientation is represented by a rotation matrix of the product coordinate system relative to the kiln coordinate system. Isolated points, duplicate points, and points with equipment confidence levels lower than the valid standard are deleted from the scan data. Missing markers are retained for obscured areas, and interpolated points are not used to replace actual test points.

[0040] The post-firing inspection model is registered to the product coordinate system based on the loading positioning reference in S110, and then transformed to the kiln coordinate system according to the loading orientation and kiln position. Point-to-point relationships are established when the nearest projected distance from the inspection point to the simulation surface is less than or equal to the point-to-point limit. The point-to-point limit is calculated using the root mean square formula, which is equal to the square root of the sum of the squares of the equipment calibration error, grid size error, and registration repeatability error. Points where the nearest projected distance from the inspection point to the simulation surface exceeds the point-to-point limit are not included in the residual calculation. If the effective inspection coverage of a certain symmetrical equivalent region does not meet the product inspection specification requirements, it is marked as a region with insufficient data.

[0041] Output the simulation-detection dataset. Each record includes furnace number, product model, product number, billet batch, firing process version, loading orientation, kiln position number, symmetrical equivalent region number, simulation node coordinates, detection point coordinates, crack detection coordinates, point pair validity markers, and detection confidence information.

[0042] S130. Calculate the dual-coordinate simulation residual. Read the simulation-detection dataset output by S120, and only process records where the point pairs are validly marked and their regions are not in areas with insufficient data. The simulation residual refers to the difference between the coordinates of the post-burn-in detection point and the coordinates of the corresponding simulation node after geometric registration, excluding the internal residuals used by the simulation solver to determine numerical convergence.

[0043] For the j-th valid point pair in the i-th symmetric equivalent region, the product coordinate residual vector is calculated in the product coordinate system according to the following formula:

[0044]

[0045] in, This is the product coordinate residual vector, in mm; The coordinates of the post-burning inspection point in the product coordinate system are in mm. The coordinates of the corresponding simulation nodes in the product coordinate system are in mm; all data are from S120.

[0046] Then, using the loading orientation and kiln position transformation matrix, the same detection point and simulation node are transformed to the kiln coordinate system, and the kiln coordinate residual vector is obtained using the same difference rule. For each symmetrical equivalent region, the residual vector of each effective point pair is multiplied by the area of ​​its corresponding grid cell, summed, and then divided by the total corresponding area of ​​the effective point pairs in the region to form the region product coordinate residual vector and the region kiln coordinate residual vector; when the sum of the effective areas of the region is 0, the region is marked as having no effective residual, and no division operation is performed.

[0047] Crack detection results are saved as the symmetrical equivalent region where the crack is located and the coordinates within that region. When the simulation fails to predict a crack or no crack is actually detected, the corresponding status marker is retained, and the missing position is not replaced by a zero coordinate. Records with residual amplitudes less than or equal to the measurement uncertainty are marked as not exceeding the measurement uncertainty. The measurement uncertainty is calculated using the root mean square formula, which is equal to the square root of the sum of the squares of the calibration error of the detection equipment, the registration error, and the grid mapping error caused by the spatial interpolation of the simulation system. However, the original residual vector is still retained for comparison across furnace batches.

[0048] Output a dual-coordinate simulation residual dataset with fields including furnace number, product number, billet batch, firing process version, loading orientation, kiln position number, symmetrical equivalent region number, product coordinate residual vector, kiln coordinate residual vector, crack region and coordinates, effective coverage, measurement uncertainty status, and data validity marker, which serve as the sole residual input for S210.

[0049] S2 specifically includes the following sub-steps:

[0050] S210. Construct a residual migration comparison set that can separate the effects of product orientation and kiln orientation. Read the dual-coordinate simulation residual dataset output by S130, and filter comparable records according to product model, billet batch, firing process version, testing equipment calibration version, and simulation mesh version; records with different versions should not be directly placed in the same comparison group.

[0051] For the same symmetrical equivalent region, establish a cross comparison group so that the same loading orientation corresponds to at least 2 different kiln positions, and the same kiln position corresponds to at least 2 different loading orientations, so as to avoid the loading orientation and kiln position being always bound together and unable to distinguish the influence of product orientation and kiln orientation; if the cross condition is not met, the comparison group is marked as a supplementary group and is not included in the model attribution judgment.

[0052] For records generated from repeated testing of the same product, if the difference between the retests does not exceed the repeatability error of the testing equipment, the median value of the coordinate components is taken; if it equals the repeatability error, it is still considered a valid retest; if it exceeds the error, it is marked as a test conflict record and excluded.

[0053] Using the coordinate transformation matrix in S110, the same symmetrical equivalent regions of each furnace are aligned to the unified product coordinate system and the unified kiln coordinate system, respectively, keeping the residual unit in mm and not performing dimensionless normalization on the residual amplitude. For node loss caused by local occlusion, only the valid region nodes commonly contained in all comparison records are used; when the common valid area does not meet the requirements of the detection specification, the comparison group is converted into a supplementary group.

[0054] Output residual migration comparison sets. Each comparison set includes the symmetric equivalent region number, product coordinate alignment residual sequence, kiln coordinate alignment residual sequence, loading orientation sequence, kiln position sequence, common valid node set, effective area and comparison status, which are used by S220 for residual migration dynamic modeling and model attribution determination.

[0055] S220. Determine the model affiliation based on the dual-coordinate residual migration law. For each valid comparison group in S210, calculate the dispersion of the product coordinate alignment residual vector relative to the product coordinate residual center and the dispersion of the kiln coordinate alignment residual vector relative to the kiln coordinate residual center. Both residual centers are formed by the median values ​​of the residual coordinate components at common valid nodes. The dispersion is represented by the square root of the average of the squared differences of the residuals at each valid node, in mm.

[0056] The product coordinate consistency limit and the kiln coordinate consistency limit are determined by synthesizing the repeatability error of the testing equipment, the coordinate registration error, the mesh mapping error, and the numerical error obtained from repeated simulations of the same qualified product using the root mean square formula. That is, they are equal to the arithmetic square root of the sum of the squares of the above four errors. Recalibration is required when the testing equipment, product model, or mesh version changes. When the product coordinate dispersion is less than or equal to the product coordinate consistency limit, the judgment residual migrates with the product orientation; when the kiln coordinate dispersion is less than or equal to the kiln coordinate consistency limit, the judgment residual is fixed at the kiln orientation.

[0057] When the residuals migrate with the product orientation and are not fixed at the kiln location, the model is classified as a material model; when the residuals are fixed at the kiln location and do not migrate with the product orientation, the model is classified as a kiln boundary model; when both types of discrete quantities exceed the corresponding consistency limit, and the residual region migrates with the relative position of the kiln support region and recurs in the vicinity of the kiln support region, the model is classified as a kiln contact model.

[0058] When both types of discrete quantities are less than or equal to the uniform limit and the residual amplitude does not exceed the measurement uncertainty, it is marked as no correction is needed; when both types of discrete quantities are less than or equal to the uniform limit but the residual amplitude exceeds the measurement uncertainty, it is marked as a conflict of assignment and returned to S210 to supplement the loading orientation or kiln position; when both types of discrete quantities exceed the limit and the residual does not migrate with the kiln furniture support relationship, it is marked as insufficient data and no model assignment is performed.

[0059] Output the dynamic modeling results of residual migration, with fields including symmetric equivalent region number, product coordinate discreteness, kiln coordinate discreteness, corresponding consistency limit, support relationship migration status, model attribution, attribution basis, and validity status.

[0060] S230: Form a model attribution and parameter locking data package. Read the residual migration dynamic modeling results output by S220, and the firing simulation platform reads the current model configuration table. The model configuration table includes a material model parameter table and a kiln boundary model parameter table; the material model parameter table records at least the material thermal conductivity parameters, thermal expansion parameters, sintering shrinkage parameters, parameter units, effective area, current value, allowable value range, and version number; the kiln boundary model parameter table records at least the temperature boundaries of each temperature zone, surface convection heat transfer boundaries, radiation heat transfer boundaries, effective kiln location, current value, allowable value range, and version number.

[0061] For residual regions belonging to the kiln furniture contact model, based on the spatial mapping between the effective regions of each parameter and the set of effective residual nodes, the material model parameters and kiln boundary model parameters that can change the calculation results of the residual region are locked. The locking means that the strategy optimization program is prohibited from changing the corresponding parameter values ​​in the current residual correction task. The current value of the parameter is not changed, nor is it affected that the same parameter is corrected according to the independent task version when processing other tasks that have been assigned to the material model or kiln boundary model.

[0062] The locking order is as follows: first, confirm the model's ownership; then, record the current parameter version; subsequently, freeze the parameter update permissions for the corresponding task, prohibiting modification of parameters followed by residual ownership determination. Regions where the model belongs to a material model or kiln boundary model will not generate kiln furniture contact processing tasks; regions with ownership conflicts or insufficient data will not lock any parameters. After locking, the model configuration version is verified again. If the version used in S220 for ownership determination is inconsistent with the current version, the locking is revoked and the process returns to S210 for recalculation.

[0063] Output model attribution and parameter locking data package, including symmetric equivalent region number, model attribution, set of valid residual nodes, locked parameter identifier, locking scope, current parameter value, parameter version number, locking reason and validity status. Records where the model is attributed to the kiln furniture contact model and the locking is valid are used as input S310.

[0064] S3 specifically includes the following sub-steps:

[0065] S310. Establish an initial contact diagram that simultaneously characterizes mechanical force transmission and contact heat conduction. Read the kiln furniture contact model record and residual effective node set from S230, read the product coordinate system from S110, read the insulator mesh from the firing simulation platform, read the kiln furniture support position, support surface profile, support surface normal, kiln furniture material and allowable assembly position from the kiln furniture design system, and read the actual loading orientation and kiln furniture number from the kiln loading record.

[0066] Before loading the kiln, use calibrated 3D scanning equipment or coordinate measuring equipment to collect the actual support surface coordinates of the kiln furniture; when the difference between the measured size and the design size is less than or equal to the combined value of the kiln furniture manufacturing tolerance and measurement error, the measured size shall be used; if it exceeds, stop using the current kiln furniture and output the kiln furniture verification mark.

[0067] The region containing the effective residual nodes and the insulator units adjacent to them are designated as product region nodes. Kiln furniture support surface units are designated as kiln furniture support nodes. Shared surfaces between adjacent insulator units are designated as internal transfer edges. Potential contact relationships between product region nodes and kiln furniture support nodes are designated as candidate contact edges. The contact diagram consists of product region nodes, kiln furniture support nodes, internal transfer edges, and contact edges. Internal transfer edges record the mechanical interactions and heat transfer relationships between units, while contact edges record the normal gap, contact normal, contact area, normal reaction force, friction state, and contact heat conduction state.

[0068] Calculate the normal gap from the product surface node along the normal direction of the support surface to the kiln furniture support surface; when the normal gap is less than or equal to the contact judgment tolerance, an initial contact edge is established. The contact judgment tolerance is calculated using the root mean square formula, that is, the contact judgment tolerance is equal to the arithmetic square root of the sum of the squares of the grid size error, the kiln furniture measurement error and the kiln loading positioning error; when the normal gap exceeds the contact judgment tolerance, it is retained as a candidate contact edge.

[0069] Friction parameters are obtained through high-temperature friction calibration using the same insulator material and the same kiln furniture material. The specific process involves placing the insulator and kiln furniture samples in a high-temperature mechanical testing machine with an independent heating furnace. Within the temperature and normal load range corresponding to the firing process, shear force is applied via a push-pull mechanism, and the ratio of tangential force to normal force is measured. Contact thermal conductivity parameters are obtained through pressure-bearing contact samples of the same material combination. The specific process involves clamping the insulator and kiln furniture samples in a high-temperature thermal conductivity meter and applying a specified normal pressure. Contact heat flow is measured using a heat flow meter, and thermocouples are placed on both sides of the contact surface to measure the temperature difference. Parameters are only used within the calibrated temperature, load, and contact state range. Parameters outside this range are marked as insufficient and not extrapolated. An initial contact diagram and its model version are output for dynamic contact relationship modeling by the S320.

[0070] S320: Dynamically update the contact edges and contact states as sintering shrinkage occurs. Read the initial contact diagram output by S310 and obtain the nodal temperature, local shrinkage strain, nodal displacement, element stress, and kiln furniture nodal positions from the firing simulation platform, using the same time reference as the firing temperature trajectory. The local shrinkage strain of the same batch of billets is determined by the temperature-time-shrinkage curve obtained from thermal expansion or sintering shrinkage tests; the test curve is interpolated according to the simulation time and nodal temperature, and is only used within the measured temperature and time range. If the range is exceeded, the calculation of the corresponding segment is stopped and a material data insufficiency mark is output.

[0071] At each calculation moment, the shrinkage strain and product profile are updated based on the node temperature. Then, the contact reaction force and contact heat flow between the insulator and the kiln furniture are solved. Subsequently, the updated contact edge is used to correct the mechanical boundary and the contact thermal boundary, and the solution is repeated until the changes in node displacement and contact heat flow in two consecutive iterations do not exceed the convergence limit set by the solver based on the mesh size and energy balance error.

[0072] If there was no contact edge in the previous moment, and the current normal gap is less than or equal to the contact determination tolerance and the calculated normal reaction force is in the compression direction, a new contact edge is added. If there was a contact edge in the previous moment, and the current normal gap exceeds the contact release tolerance or the normal reaction force is no longer in the compression direction, the contact edge is deleted. The contact release tolerance is equal to the sum of the contact determination tolerance and the contact iteration position error. When the current gap is equal to the contact release tolerance, the state of the previous moment is maintained.

[0073] For persistent contact edges, the state is determined to be adhesive when the tangential force is less than or equal to the product of the friction parameter and the normal reaction force; when it exceeds this value, the state is determined to be slipping; and when it equals the critical value, it is classified as adhesive. If a single time step fails to converge, the time step size is shortened to half of the original time step size and recalculated. If convergence is still not achieved after reaching the minimum time step size allowed by the simulation platform, the result of the previous valid time step is retained, and that segment is marked as a solution anomaly and is not included in the S330 path migration judgment.

[0074] Output dynamic contact diagram sequence and state transition records for contact edge addition, deletion, adhesion to slip, and slip to adhesion.

[0075] S330. Determine the migration results of the main load path and the contact heat conduction path. Read the dynamic contact diagram sequence output by S320. At each effective calculation time, write the resultant force on the shared surface of adjacent insulator units into the mechanical properties of the internal transmission edge, and write the contact reaction force into the mechanical properties of the contact edge; write the heat conduction between adjacent units into the thermal properties of the internal transmission edge, and write the contact heat flow between the insulator and the kiln furniture into the thermal properties of the contact edge. Force is uniformly represented by N, and heat flow is uniformly represented by W.

[0076] The product area nodes subjected to gravity, shrinkage constraint force, or temperature stress are taken as load source nodes, and the kiln furniture support nodes are taken as load endpoint nodes. For each continuous path from the load source node to the kiln furniture support node, the edge with the smallest mechanical transmission amount in the path is taken as the force transmission capacity of the path. The path with the largest force transmission capacity is determined as the main load path. If there are multiple paths with the same maximum value, all paths are retained.

[0077] The product area nodes with nodal temperature differences and heat transfer to the kiln furniture are taken as heat source nodes. For each continuous path from the heat source node to the kiln furniture support node, the edge with the smallest heat transfer in the path is taken as the thermal conductivity of the path. The path with the largest thermal conductivity is determined as the contact heat conduction path. All paths with the same maximum value are retained.

[0078] When the load master path or contact heat conduction path changes at any of the contact edge, termination support node, or covered symmetrical equivalent region at adjacent calculation times, it is recorded as a path migration, and associated with the addition, deletion, or contact state transition of the contact edge that caused the migration. Only migration records where the path coverage area shares a common node or mesh adjacency with the effective node set of the S230 residual are retained; other migrations are not included in this correction task.

[0079] The output is the dynamic modeling result of the contact diagram. The fields include the calculation time, contact change type, main load path, contact heat conduction path, node set before and after migration, action area, corresponding residual area and effectiveness status, which serve as the sole input to S410.

[0080] S4 specifically includes the following sub-steps:

[0081] S410. Determine the billet process stage corresponding to the moment of contact change. Read the dynamic modeling results of the contact diagram in S330, and read the nodal temperature, shrinkage rate, equivalent stress, and viscoplastic strain increment of the same insulator region at the same calculation time from the firing simulation platform; read the high-temperature stress relaxation curve, sintering shrinkage curve, and high-temperature load-bearing capacity curve corresponding to the current billet batch from the material property database.

[0082] The high-temperature stress relaxation curve is obtained by applying a constant strain corresponding to the simulated support strain to a specimen with the same formula and the same forming state at multiple process temperatures and recording the stress decay over time. When the residual stress of the specimen drops to within the limit determined by the allowable residual stress specified in the product design specification, the corresponding temperature-time interval is defined as the stress relaxation stage.

[0083] The sintering shrinkage curve is obtained by thermal expansion or sintering shrinkage test of the same batch of samples; when the shrinkage rate does not exceed the shrinkage rate corresponding to the resolution of the test equipment within a time not shorter than the minimum heat preservation control cycle of the kiln control system, and the high temperature bearing curve shows that the allowable bearing stress of the material is not less than the product of the support stress of the area corresponding to S330 and the product design safety factor, the structural fixing stage is determined to begin.

[0084] The period after the end of the stress relaxation stage and before the start of the structural fixation stage is defined as the contact change sensitive stage. When the contact change time equals the end of the stress relaxation stage, it is classified as the contact change sensitive stage; when it equals the start of the structural fixation stage, it is classified as the structural fixation stage. If the material curve is inconsistent with the current billet formulation, forming method, or firing atmosphere, the original stage boundary is not invoked, and the corresponding record is marked as insufficient stage data.

[0085] Output a table corresponding to the contact change stages, including the calculation time, contact change type, insulator region, stress relaxation stage boundary, contact change sensitive stage boundary, structural fixation stage boundary, process stage, and effectiveness status, for S420 to filter records within the contact change sensitive stage.

[0086] S420: Determine the spatial overlap between the load concentration region and the temperature gradient region. Read the records in S410 that belong to the contact change sensitive stage of the process and whose effectiveness status is valid. Read the load main path, contact heat conduction path, contact edge normal reaction force and action area at the same calculation time from S330. Read the node coordinates, node temperature and element stress on the same mesh from the firing simulation platform.

[0087] The local contact load density, expressed in MPa, is obtained by dividing the normal reaction force of the contact edge by the actual contact area corresponding to that edge. The actual contact area is the sum of the grid areas of the product surface under compressive contact conditions. If the area is 0, no division is performed, and the record is marked as invalid. The allowable contact load density at the current temperature is determined by dividing the high-temperature compressive strength of the same formulation sample by the safety factor used in the product design specification. Grid cells with a local contact load density greater than or equal to the allowable contact load density are classified into load concentration regions.

[0088] The temperature gradient is calculated based on the temperature difference between adjacent mesh nodes and the distance between nodes, in °C / mm. The allowable temperature gradient is obtained by progressively increasing the local temperature gradient in a thermo-mechanical model with locked material parameters, determining the gradient at which the predicted maximum tensile stress reaches the allowable tensile stress at the current temperature. The allowable tensile stress is determined by the high-temperature tensile or bending strength of the same formulation specimen and the design safety factor. Mesh cells with an actual temperature gradient greater than or equal to the allowable temperature gradient are classified into the temperature gradient region.

[0089] The two types of regions are mapped to the unified product mesh output by the sintering simulation platform. If there is at least one complete mesh cell that belongs to both the load concentration region and the temperature gradient region, it is considered to be spatially coincident. If they only meet at the boundary and have no common mesh cells, they are considered to be non-coincident.

[0090] Output contact change coincidence determination set, including calculation time, contact change type, load concentration area, temperature gradient area, common mesh element, coincidence determination result, main load path, contact heat conduction path and effectiveness status, as input for S430 to perform strategy optimization.

[0091] S430: Generate an executable process optimization adjustment data package. Read the records in S420 where the overlap judgment result is overlapped and the validity status is valid. Read the allowable adjustment range of the support position, the adjustment resolution of the support structure, the height range of the support surface, and the minimum structural spacing between adjacent support parts from the kiln furniture design system. Read the achievable loading orientation and positioning resolution from the kiln loading control system. Read the temperature setting trajectory, temperature adjustment resolution, maximum temperature, minimum temperature, heating rate, and heating power limit of the corresponding firing temperature zone from the kiln control system.

[0092] Candidate quantities for support position, loading orientation, and temperature trajectory are generated according to the equipment adjustment resolution. Candidate quantities exceeding any structural or control limit are directly deleted. For each candidate combination, while keeping the material model parameters locked in S230 and the kiln boundary model correction parameters unchanged, S310 to S420 are re-executed. The temperature trajectory candidate quantities are process control quantities to be executed and are not used as model residual fitting parameters.

[0093] Candidate combinations are determined to be executable if they cause the original contact change moment to enter the stress relaxation stage or the structural fixation stage, or if the load concentration area and temperature gradient area in the contact change sensitive stage no longer coincide, and no new coincidence is formed in other symmetrical equivalent regions.

[0094] The executable combinations are prioritized using either a grid traversal method or a multi-objective lexicographic optimization algorithm: First, the combination with the minimum predicted additional heating energy is selected. The additional heating energy is obtained by integrating the difference between the heating power corresponding to the candidate temperature trajectory and the heating power corresponding to the original temperature trajectory over the firing time, with the unit being kWh. When the additional heating energy is the same, the combination with the minimum total adjustment distance of the support position, the minimum adjustment angle of the loading orientation, and the maximum time interval between the contact change moment and the boundary of the contact change sensitive stage is selected in sequence.

[0095] If no combination completely eliminates overlap, the output is limited optimization result only if the candidate combination can shorten the overlap duration and the maximum contact load density and maximum temperature gradient do not exceed the original scheme; otherwise, the output shows no executable adjustment and the current optimization cycle is terminated.

[0096] The output process optimization adjustment data package includes kiln furniture support position adjustment amount, loading orientation adjustment amount, temperature trajectory adjustment amount of the corresponding firing temperature zone, target contact change time, target load main path, target contact heat conduction path, predicted additional heating energy, optimization status and model version number, which are used as inputs to S510.

[0097] S5 specifically includes the following sub-steps:

[0098] S510: Perform process optimization adjustments and generate a firing execution record. Read the process optimization adjustment data package output from S430 and verify whether the product model, kiln furniture number, model version number, and firing process version are consistent with the furnace to be executed; if any item is inconsistent, do not issue an adjustment command and return to S430 to regenerate. For adjustable kiln furniture, move the support parts according to the adjustment amount of the kiln furniture support position through the support adjustment mechanism; for kiln furniture composed of replaceable support parts, install the support parts corresponding to the adjustment amount before loading the kiln.

[0099] After adjustment, use coordinate measuring equipment to measure the actual support position. The position is considered valid if the difference between the actual and target positions is less than or equal to the combined value of the kiln furniture manufacturing tolerance, the positioning error of the adjusting mechanism, and the measurement error. If it exceeds this value, stop loading the kiln and readjust. The insulator is rotated by the kiln loading equipment according to the loading orientation adjustment amount. The position is considered valid if the actual orientation deviation does not exceed the positioning error of the kiln loading equipment.

[0100] After confirming the support position and loading orientation, the temperature trajectory adjustment amount is written to the corresponding firing temperature zone controller. When the adjusted temperature trajectory reaches the equipment temperature or heating rate limit, it is executed according to the limit value. If the limit is exceeded and the trajectory cannot be kept continuous after limiting the amplitude, the temperature trajectory is not executed and the process returns to S430.

[0101] During the firing process, the set temperature, actual temperature, heating power, conveying status and control alarm are recorded according to the original sampling cycle of the kiln control system. When the actual temperature deviation continuously reaches the confirmation cycle specified by the equipment fault judgment and exceeds the sum of the temperature sensor calibration error and the control allowable deviation, the current furnace is marked as an execution abnormality. The execution abnormality data is not used for the S530 model affiliation update.

[0102] Output firing execution records, including furnace number, product number, actual support position, actual loading orientation, actual temperature trajectory, actual heating power, command execution deviation, alarm status, model version number, and execution validity status, as the basis for S520 detection tasks.

[0103] S520: Obtain the optimized three-dimensional deformation, kiln furniture contact imprint, and crack initiation point. After the insulator has been fired and cooled to the allowable temperature of the testing equipment, and the surface temperature change does not exceed the thermal drift compensation range of the equipment within the stable time specified in the testing specification, a testing task is established based on the furnace number, product number, loading orientation, actual kiln position, and kiln furniture number in S510.

[0104] The coordinates of the entire surface of the insulator and the local morphology of the area adjacent to the kiln support were collected using a 3D scanning device calibrated with standard-sized parts. The coordinates and deformation were uniformly expressed in mm. Isolated points, duplicate points and points with equipment confidence levels lower than the effective standard were deleted, and missing marks were retained in occluded areas.

[0105] Based on the loading positioning reference in S110, the post-firing scanning model is registered to the product coordinate system, and converted to the kiln coordinate system according to the actual loading orientation and actual kiln position recorded in S510. The three-dimensional deformation vector of each symmetrical equivalent region relative to the design structure is calculated. The kiln furniture contact imprint refers to the local flattening, friction groove, local contour change, or identifiable contact boundary that appears on the surface of the insulator after firing, and the local morphological region and the corresponding kiln furniture support surface projection area on the product surface share common grid units; only color changes without spatial correspondence are not considered valid contact imprints.

[0106] Surface image inspection equipment is used to collect data on crack skeleton, endpoints, bifurcation locations, and opening width variations. Along the crack skeleton, from the endpoints to the bifurcation areas, the opening width and number of branches are examined. The end where the opening width continuously increases and the number of branches increases is recorded as the crack initiation point. If a unique direction cannot be formed, it is marked as an undetermined crack initiation point and is not replaced by the crack midpoint. The 3D deformation, kiln furniture contact imprints, and crack initiation points are mapped to the symmetrical equivalent region of S110 and the product area node of S310.

[0107] The predicted contact area corresponding to the predicted contact edge and the actual contact area corresponding to the kiln furniture contact imprint are mapped to the same product grid. If the two areas share at least one common grid cell, the predicted contact edge is determined to correspond to the kiln furniture contact imprint. If the two areas do not share a common grid cell, but the nearest boundary distance does not exceed the combined value of the contact imprint measurement error and the grid mapping error, adjacent grid cells within the distance range are merged into the common area before a corresponding determination is made. If the nearest boundary distance exceeds the combined value, or if the two areas only connect at the grid boundary without sharing a common grid cell, they are determined to be non-corresponding. The contact coverage ratio is obtained by dividing the area of ​​the common area by the area of ​​the predicted contact area. If the predicted contact area area is zero, the corresponding record is marked as an invalid contact area, and no division operation is performed.

[0108] The output optimization result detection data package includes product number, region number, three-dimensional deformation vector, contact imprint node set, predicted contact area, actual contact area, corresponding judgment result, contact coverage ratio, crack initiation coordinates, detection confidence information, missing status and validity status, which are used as inputs to S530.

[0109] S530: Update the contact diagram and re-verify the model attribution of the simulation residuals. Read the optimization result detection data package from S520, the burn-in execution record from S510, the contact diagram dynamic modeling results from S330, and the parameter lock status from S230. Only process records where both the execution validity status and the detection validity status are valid.

[0110] While keeping the material model parameters and kiln boundary model parameters locked in S230 unchanged, the predicted contact area, actual contact area, corresponding judgment result, and contact coverage ratio output by S520 are read. Based on the corresponding judgment result of common mesh cells and nearest boundary distance, the predicted contact edge is matched with the actual kiln furniture contact imprint one by one: when a contact edge is predicted to exist and the corresponding judgment result is corresponding, the contact edge is retained, the contact area is corrected by the sum of the areas of the mesh cells contained in the actual contact area, and the contact coverage ratio is written into the contact edge correction record; when a contact edge is predicted to exist but the corresponding judgment result is not corresponding in the area where the detection coverage is effective, the contact edge is marked as an edge to be deleted; when no contact edge is predicted but an effective kiln furniture contact imprint is detected, a feedback contact edge is added between the product area node and the kiln furniture support node corresponding to the actual contact area.

[0111] When the crack initiation point and the load main path termination area or path migration area share the same mesh element, the corresponding contact state transition rule is retained; when the crack initiation point and the predicted action area do not coincide, and the angle between the three-dimensional deformation direction and the predicted load main path direction exceeds the combined range of the detection direction error and the mesh direction error, the corresponding friction parameter or contact thermal conductivity parameter is marked as needing recalibration, and the parameter value is not directly modified when calibration data is missing.

[0112] Re-execute S310 to S330 using the updated contact edge, contact area, and effective calibration parameters to obtain updated simulation results; recalculate the dual-coordinate simulation residuals using the same data structure as S130, and then execute S210 and S220 in sequence.

[0113] When new residuals migrate with the product orientation, the model is updated to the material model; when fixed at the kiln orientation, it is updated to the kiln boundary model. When both types of coordinates are inconsistent and still migrate with the kiln furniture support relationship, they continue to be classified as kiln furniture contact models. Data coverage is insufficient or there are classification conflicts, and the data is marked as needing to be supplemented. When the model classification changes, the parameter locks that are inconsistent with the original classification are revoked, and the parameter lock states are regenerated according to S230. When the model classification remains as the kiln furniture contact model, the updated contact diagram is used as the initial model for the next batch of dynamic modeling and strategy optimization.

[0114] The final output optimization feedback data package includes the updated contact diagram, contact edge correction record, updated dual-coordinate simulation residual, re-verified model attribution, parameter lock status, parameters to be recalibrated, and model version number.

[0115] In this embodiment, all physical quantities participate in data input, result comparison, and output according to the units described in the aforementioned steps. When the firing simulation platform performs scale transformation or dimensionless processing within the solver to improve the stability of the numerical solution, the corresponding physical units should be restored before outputting nodal temperature, nodal displacement, contact reaction force, contact heat flux, temperature gradient, and additional heating energy. Point-to-point limits, consistency limits, contact judgment tolerance, contact release tolerance, allowable contact load density, and allowable temperature gradient are determined based on equipment calibration error, mesh mapping error, material test results, product design specifications, and equipment control capabilities, and are used according to the synthesis or comparison methods described in the aforementioned steps.

[0116] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0117] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0118] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A simulation and optimization method for green and low-carbon insulator manufacturing processes, characterized in that, Includes the following steps: S1. Obtain the three-dimensional structure of the insulator, the symmetrical equivalent region, and the firing simulation results and post-firing test results of different loading orientations and kiln positions. Register the test points and simulation nodes to the product coordinate system and the kiln coordinate system to obtain the dual-coordinate simulation residual. S2. Based on the state of the dual-coordinate simulation residual migrating with the product orientation, fixed in the kiln position, or migrating with the kiln furniture support relationship, determine whether it belongs to the material model, kiln boundary model, or kiln furniture contact model, and lock the material model parameters and kiln boundary model parameters acting on the residual area when it belongs to the kiln furniture contact model. S3. Based on the sintering shrinkage state of the insulator and the kiln support structure, a dynamic model of the contact relationship is constructed to build a contact diagram with mechanical force transmission properties and contact heat conduction properties. The contact edges and contact states are updated as the sintering shrinks to obtain the migration results of the main load path and the contact heat conduction path. S4. Correspond the contact change time with the stress relaxation stage and the structural fixation stage to determine the contact change sensitive stage. Determine whether the load concentration area and the temperature gradient area in the contact change sensitive stage coincide. When they coincide, optimize the generation of kiln furniture support position adjustment, loading orientation adjustment and temperature trajectory adjustment through strategy optimization.

2. The simulation and optimization method for insulator production process based on green and low-carbon technology according to claim 1, characterized in that, Also includes: S5. Based on the adjustment amount, adjust the kiln furniture support, loading orientation, and firing temperature trajectory to obtain the three-dimensional deformation after firing, kiln furniture contact marks, and crack initiation points. Update the contact diagram and dual-coordinate simulation residuals, and update the parameter locking status according to the model's classification.

3. The simulation and optimization method for insulator production process based on green and low-carbon technology according to claim 1, characterized in that, S1 specifically includes: Based on the design symmetry center, common axis and loading positioning reference of the insulator, a product coordinate system is established. Regions whose structural topology, nominal thickness, surface normal and region boundary all meet the corresponding matching limits after rotation or mirror transformation are determined as symmetrical equivalent regions, forming a geometric reference data package. Obtain the coordinates of the firing simulation nodes and the coordinates of the post-firing detection points corresponding to different loading orientations and kiln positions. Establish point-to-point relationships between the detection points whose distance does not exceed the point-to-point limit and the simulation nodes, and register them to the product coordinate system and the kiln coordinate system respectively to form simulation and detection corresponding datasets.

4. The simulation and optimization method for insulator production process based on green and low-carbon technology according to claim 3, characterized in that, Also includes: Calculate the coordinate difference between each effective point pair in the product coordinate system and the kiln coordinate system, summarize it into a region residual vector according to the corresponding grid area, and form a dual-coordinate simulation residual.

5. The simulation and optimization method for insulator production process based on green and low-carbon technology according to claim 1, characterized in that, S2 specifically includes: The dual-coordinate simulation residuals are selected according to product model, billet batch, firing process version, testing equipment calibration version and simulation grid version, and the same loading orientation corresponds to different kiln positions and the same kiln position corresponds to different loading orientations to form a residual migration comparison set. The discrete quantities of product coordinate alignment residual and kiln coordinate alignment residual are calculated separately. The residual that migrates with the product orientation is assigned to the material model, the residual that is fixed in the kiln orientation is assigned to the kiln boundary model, and the residual that is inconsistent in both types of coordinates and migrates with the kiln furniture support relationship is assigned to the kiln furniture contact model. When the residual is assigned to the kiln furniture contact model, the material model parameters and kiln boundary model parameters that coincide with the effective node set of the residual in the locked area are used to form a model attribution and parameter locking data package.

6. The simulation and optimization method for insulator production process based on green and low-carbon technology according to claim 1, characterized in that, S3 specifically includes: An initial contact diagram is established based on the insulator mesh, the set of effective residual nodes, and the kiln furniture support surface. The initial contact diagram includes product area nodes, kiln furniture support nodes, internal transmission edges, and contact edges. Mechanical force transmission properties and contact heat conduction properties are configured for the contact edges. Dynamic modeling of contact relationships is performed based on the nodal temperature, shrinkage strain, nodal displacement, and normal reaction force at each calculation moment during the sintering process. When the normal gap and normal reaction force meet the contact conditions, a new contact edge is added, and when the release conditions are met, the contact edge is deleted. The adhesion state or slip state is updated according to the relationship between tangential force and friction constraint.

7. The simulation and optimization method for insulator production process based on green and low-carbon technology according to claim 6, characterized in that, Also includes: Determine the main load path and contact heat conduction path based on the mechanical and thermal transfer quantities of each side, and record the path migration results caused by the addition or deletion of contact sides or the change of contact state.

8. The simulation and optimization method for insulator production process based on green and low-carbon technology according to claim 1, characterized in that, S4 specifically includes: Based on the high-temperature stress relaxation curve, sintering shrinkage curve and high-temperature bearing capacity curve of the same batch of billets, the stress relaxation stage and the structural fixation stage are determined, and the time interval between the two is determined as the contact change sensitive stage. Mesh cells with contact load density reaching the allowable contact load density at the current temperature are defined as load concentration regions, and mesh cells with temperature gradient reaching the allowable temperature gradient are defined as temperature gradient regions. The overlap state is determined by whether the two types of regions have common mesh cells. When overlap occurs, adjustment candidate combinations are generated within the constraints of kiln furniture structure, kiln loading and positioning, and kiln control. Through strategy optimization screening, combinations that allow contact changes to leave the contact change sensitive stage or eliminate the overlap state are selected, and process optimization adjustment data packages are output.

9. The simulation and optimization method for insulator production process based on green and low-carbon technology according to claim 2, characterized in that, S5 specifically includes: Verify the process optimization and adjustment data package with the product model, kiln furniture number, model version number and firing process version of the furnace to be executed. When the verification is consistent and the adjustment amount meets the control limits of kiln furniture, kiln loading equipment and kiln, execute the adjustment of kiln furniture support position, loading orientation and firing temperature zone temperature trajectory, and form a firing execution record. After the insulator is fired and cooled to the testing conditions, the surface coordinates and crack images after firing are collected. The local morphological changes that share the same grid unit with the projection area of ​​the kiln furniture support surface are identified as kiln furniture contact imprints, forming an optimized result detection data package.

10. The simulation and optimization method for insulator production process based on green and low-carbon technology according to claim 9, characterized in that, Also includes: Based on the correspondence between the predicted contact edge and the kiln furniture contact imprint, contact edges are added or deleted and the contact area is corrected. The dual-coordinate simulation residual and model attribution are obtained again, and the parameter locking status is updated accordingly.

Images