In-process testing and precision control method for GPS dual mapping precision machining geometric error

Abstract

The present application relates to precision machining on-machine measurement technology field, disclose a kind of precision machining geometric error on-machine inspection and precision control method of GPS dual mapping.It obtains GPS design specification data, nominal geometric model and on-machine measurement point set data and carries out coordinate unification;Accordingly, GPS specification operation operator chain with specific execution sequence is constructed, and under the constraint of machine tool measurement, the order is kept unchanged to generate test operation operator chain;Based on the on-machine measurement point set data, a multi-scale skin surface model is constructed and a substitute test surface model is generated to obtain geometric error amount;Based on the substitute test surface model, the overall residual of the test side is calculated, combined with the overall residual of the specification side and the operator execution metric to calculate the operator invariant, the operation consistency is determined, and finally the precision control is adjusted.The structure of design specification and on-machine test operation is consistent, so that the existing distortion evaluation and blind compensation problem is fundamentally overcome.

Description

Technical Field

[0001] This invention relates to the field of precision machining in-machine measurement technology, and in particular to a method for in-machine inspection and precision control of precision machining geometric errors using GPS dual mapping. Background Technology

[0002] Geometrical Product Specification (GPS) is an international standard system (such as ISO GPS) used to define and standardize the geometric dimensions, tolerance types, and geometric relationships of mechanical products. It is a fundamental technology for achieving consistency in design, manufacturing, and testing in the field of precision manufacturing.

[0003] In existing technologies, the design phase typically defines tolerances for workpiece geometric features based on GPS specifications, while the on-machine inspection phase completes the geometric error assessment through processing steps such as measurement point acquisition, data filtering, geometric fitting, and error calculation. Specifically, the on-machine inspection system first acquires measurement point data on the workpiece surface, then uses a preset data processing algorithm to extract features and construct the geometry of the measurement points, and finally compares the calculated geometric parameters with the design tolerances for judgment.

[0004] In the process of implementing the relevant technical solutions, the inventors of this application discovered that the above-mentioned prior art has at least the following technical problems:

[0005] 1. Existing in-machine inspection processes often deviate from design semantics, using only design tolerances as the final comparison benchmark, while intermediate processing steps such as feature extraction and filtering are arbitrarily executed based on general algorithms. This approach ignores the inherent geometric constraints and operational execution logic of GPS design specifications, resulting in a lack of rigorous sequential and structural mapping between the geometric evaluation operations on the design side and the actual inspection process. This makes it highly susceptible to detection distortion due to inconsistencies in the evaluation paths.

[0006] 2. Existing vision- or measurement-based on-machine error compensation systems mostly employ an end-to-end black-box optimization strategy of "calculating the absolute value of the deviation and directly solving for compensation parameters." These systems lack a self-checking mechanism for the consistency of the measurement and inspection processes themselves; once point cloud anomalies occur at the processing site or the inspection operation path deviates from the design intent, the system will still blindly issue compensation commands, leading to serious consequences such as overcompensation or even a reverse deterioration of processing accuracy. Summary of the Invention

[0007] To overcome the shortcomings of the prior art, the present invention aims to provide an in-machine inspection and precision control method for precision machining geometric errors based on GPS dual mapping. This method aims to solve the problem that the existing in-machine inspection process deviates from the design semantics and lacks a unified structural mapping, resulting in evaluation distortion. At the same time, it also solves the technical problem that the existing error compensation system lacks an operational consistency self-checking mechanism, which easily leads to blind miscompensation.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A method for on-machine inspection and accuracy control of precision machining geometric errors in GPS dual mapping includes the following steps:

[0010] S1. Obtain the GPS design specification data, nominal geometric model, and on-machine measurement point set data of the workpiece, and perform coordinate unification processing on the on-machine measurement point set data;

[0011] S2. Based on the GPS design specification data, analyze the geometric constraints of the design semantics and construct a chain of GPS specification operation operators with a specific execution order;

[0012] S3. Based on the machine tool measurement constraints, and while keeping the specific execution order unchanged, perform GPS dual mapping on each operator in the GPS standard operation operator chain by operator parameter transformation to generate a verification operation operator chain that is consistent with the design semantic structure.

[0013] S4. Construct a multi-scale skin model based on the on-machine measurement point set data, and process the multi-scale skin model based on the inspection operation operator chain to generate an alternative inspection surface model.

[0014] S5. Calculate the inspection geometric feature value based on the alternative inspection surface model, and at the same time calculate the design geometric feature value of the nominal geometric model according to the GPS specification operation operator chain to obtain the geometric error amount;

[0015] S6. Calculate the overall residual and operator execution metric on the inspection side based on the alternative inspection surface model, and calculate the operator conservation invariant by combining the overall residual and operator execution metric on the specification side; compare the operator conservation invariant with the reference operator conservation invariant to determine the operational consistency status between the inspection operation process and the design specification.

[0016] S7. Based on the geometric error and the operational consistency determination result, perform in-machine precision control adjustment.

[0017] Furthermore, the coordinate unification process described in step S1 includes:

[0018] The coordinate transformation relationship between the machine tool coordinate system and the measurement coordinate system is obtained by minimizing the spatial deviation between the set of measurement reference points and the set of nominal reference points and fitting a rigid body transformation matrix. Translation vector To establish;

[0019] Based on the coordinate transformation relationship, the on-machine measurement point set data is spatially transformed and unified to the workpiece reference coordinate system;

[0020] Perform a benchmark consistency check on the converted benchmark measurement points, assuming the first... The nominal benchmark points are , No. The converted measurement reference points are: Its deviation is defined as ,in Represents the Euclidean norm, when all reference points deviate... All are less than the preset benchmark consistency tolerance threshold. At that time, the coordinates were determined and the unified processing was completed.

[0021] Furthermore, the step of constructing a chain of GPS canonical operation operators with a specific execution order in step S2 includes:

[0022] Analyze the tolerance types and datum information in GPS design specification data;

[0023] Based on the tolerance type and reference information, the selected operators are arranged and combined according to the operation sequence specified in the GPS design specifications to construct the GPS standard operation operator chain.

[0024] Further, the GPS dual mapping described in step S3 includes:

[0025] Obtain machine tool measurement path constraint information and measurement reachability information;

[0026] Match each operator in the GPS standard operation operator chain with the machine tool measurement path constraint information;

[0027] The execution parameters and execution methods of each operator are converted based on the matching results;

[0028] The parameter transformation of the filter operator includes: based on the distribution function of the distance between measurement points. Statistical characteristics can be used to reset the filter window size or cutoff frequency.

[0029] The resizing of the filter window size or cutoff frequency includes: using the median of the measurement point spacing. As a statistical characteristic, the cutoff wavelength of the filter operator is set. ,in The preset empirical coefficients are determined in advance based on the accuracy of the machine tool measuring device, the surface characteristics of the workpiece being processed, and the stage of the processing procedure, so that the filtering parameters are adapted to the local density of the measurement points;

[0030] The parameter transformation of the geometric association operator includes: calculating the angle between the measurement direction vector and the nominal reference direction vector based on the measurement accessibility information; if the angle exceeds the preset attitude constraint range determined by the machine tool measurement attitude limit, then adjusting the reference association direction or replanning the measurement attitude.

[0031] Furthermore, the step of constructing the multi-scale skin model described in step S4 includes:

[0032] The on-machine measurement point set data is subjected to scale decomposition processing to obtain geometric deviation data at multiple scale levels;

[0033] The geometric deviation data of each scale layer are combined with the nominal geometric model to generate a multi-scale skin model;

[0034] The scale decomposition process is achieved by constructing an adjacency graph according to spatial neighborhood relationships and performing hierarchical decomposition of deviation data based on neighborhood scale parameters determined by workpiece geometric features or measurement point density.

[0035] The division of each scale layer is determined based on the spatial neighborhood radius or equivalent filtering parameters.

[0036] Furthermore, the step of generating the alternative inspection surface model in step S4 includes:

[0037] According to the execution parameters determined by the test operation operator chain, the multi-scale skin model is sequentially subjected to geometric feature extraction processing, filtering processing, geometric association processing and geometric construction processing;

[0038] Based on the processing results, an alternative verification surface model corresponding to the GPS standard operation operator chain structure is constructed.

[0039] Furthermore, the step of calculating the operator-conserved invariants in step S6 includes:

[0040] The overall geometric measurement of the alternative test surface model is performed to obtain the overall residual measurement value. ,in To determine the total number of measurement points involved in the calculation of the geometric elements, For the first The residual vector from each measurement point to the test geometric element;

[0041] Operator metrics are generated based on the operator execution information recorded in the operator chain of the inspection operation. ,in The total number of operators participating in the generation of alternative test surface models, To pre-calibrate and determine the first operator based on its type and its influence on the final error evaluation result. The weights of each operator According to the first Execution parameters of the operators norm and scale of participating data The operator metric function for computation;

[0042] The overall residual metric and the operator metric are respectively passed through After performing dimensionless normalization on the function and summing the results, we obtain the operator-conserved invariants. .

[0043] Furthermore, the aforementioned The function is a dimensionless normalization processing function, defined as follows: ;

[0044] in, The metric to be normalized and These are the minimum and maximum values ​​of the metric obtained statistically from historical data, respectively; if historical data is insufficient or it is the first time running, the... and Experienced or theoretical values ​​preset according to machining accuracy requirements and machine tool characteristics can be used as substitutes to eliminate dimensional differences between different measurement values, so that the overall residual measurement value and the operator measurement value after processing can be summed under the same dimension.

[0045] Furthermore, the step S6, which involves determining the operational consistency between the inspection process and the design specifications, includes:

[0046] Obtain the operator-conserved invariants obtained from the current test operation. Conserved invariants with reference operator , wherein The conserved invariants of the standard side operators are derived from historical processing data, calibration data, or calculated based on the nominal geometric model and GPS standard operating operator chain.

[0047] Calculate the difference ;

[0048] like Less than the preset consistency tolerance threshold If so, the inspection process is deemed to have acceptable consistency with the design specifications;

[0049] like Greater than or equal to the tolerance threshold If so, it is determined that a structural deviation has occurred in the inspection process.

[0050] Furthermore, the step of performing in-machine precision control adjustment in step S7 includes:

[0051] If the operational consistency determination result of step S6 indicates that the inspection operation process has acceptable consistency with the design specifications, then:

[0052] Based on geometric error Determine the adjustment amount of processing parameters ,in This is the error response matrix established based on the dynamic characteristics of the machine tool;

[0053] according to Calculate the correction factor The processing parameters are adjusted to obtain the correct amount. ;

[0054] Input the corrected processing parameter adjustment into the processing equipment, and re-execute the in-machine measurement and inspection operations of steps S1 to S6 based on the processed workpiece;

[0055] If the operational consistency determination result of step S6 indicates that a structural deviation has occurred in the inspection operation process, then:

[0056] If the processing parameters are not adjusted and the adjustment amount is not input into the processing equipment, the process of re-executing the in-machine measurement and inspection operations of steps S1 to S6 will be triggered or an alarm message will be output.

[0057] The present invention has the following beneficial effects:

[0058] 1. This invention constructs a chain of GPS specification operation operators with a specific execution order by analyzing the design semantics and geometric constraints in GPS design specification data. While maintaining this specific execution order, it performs a GPS dual mapping of operator parameters based on machine tool measurement constraints, generating a chain of inspection operation operators consistent with the design semantic structure. Compared to existing technologies that only use design tolerances as the final numerical comparison benchmark while intermediate processing steps are freely executed based on general algorithms, this invention's dual mapping mechanism establishes a structural correspondence between the specification side and the inspection side at the operator level. This ensures that the on-machine inspection process inherits the evaluation logic of the design specification in both execution order and parameter configuration, thereby avoiding geometric error evaluation distortion caused by differences in operator execution order or parameter mismatch.

[0059] 2. This invention constructs a multi-scale surface model by performing scale decomposition on the on-machine measurement point set data and combining it with a nominal geometric model. This model is controlled by a chain of verification operation operators transformed through dual mapping. Geometric feature extraction, filtering, geometric correlation, and geometric construction are performed sequentially according to the determined execution parameters, generating an alternative verification surface model corresponding to the GPS specification operation operator chain structure. Since the execution parameters of each processing step are uniformly constrained by the dual-mapped verification operation operator chain, the on-machine measurement point set data can complete scale decomposition and geometric reconstruction within a unified geometric framework strictly corresponding to GPS design specifications, effectively suppressing the interference of environmental noise at the processing site on the geometric feature evaluation results.

[0060] 3. This invention introduces an overall residual metric and operator execution metric based on a substitute inspection surface model, and combines this with the overall residual on the specification side to calculate a normalized dimensionless operator-conserved invariant. This enables an objective quantitative assessment of the consistency between the inspection operation process and the design specifications. During the precision control adjustment phase, this operator-conserved invariant further functions as a constraint for processing parameter correction: when the difference in the operator-conserved invariant is less than a preset consistency tolerance threshold, the system calculates a correction coefficient to fine-tune the processing parameter adjustment; when the difference in the operator-conserved invariant reaches or exceeds the tolerance threshold, it indicates a structural deviation or measurement failure during the on-machine inspection operation. The system proactively intercepts the current processing parameter adjustment command and triggers a re-inspection process, thereby avoiding invalid or harmful processing error compensation based on structurally inconsistent erroneous measurement data, and improving the fault tolerance and robustness of the precision control system in industrial environments. Attached Figure Description

[0061] Figure 1 This is a schematic diagram of the overall process for the in-machine inspection and precision control method of GPS dual mapping for precision machining geometric errors provided in an embodiment of the present invention.

[0062] Figure 2 A schematic diagram of the sub-process for constructing a coordinate unification processing and standardized operation operator chain provided in an embodiment of the present invention;

[0063] Figure 3 This is a schematic diagram illustrating the dual mapping relationship between the GPS standard operation operator chain and the inspection operation operator chain under machine tool measurement constraints provided in an embodiment of the present invention.

[0064] Figure 4 This is a schematic diagram illustrating the process of constructing a multi-scale skin surface model and generating an alternative test surface model based on a chain of test operation operators, as provided in an embodiment of the present invention. Detailed Implementation

[0065] The technical solutions in 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.

[0066] Please refer to Figure 1 This embodiment provides an in-machine inspection and accuracy control method for precision machining geometric errors based on GPS dual mapping, including the following steps:

[0067] S1. Obtain the GPS design specification data, nominal geometric model, and on-machine measurement point set data of the workpiece, and perform coordinate unification processing on the on-machine measurement point set data;

[0068] S2. Based on the GPS design specification data, analyze the geometric constraints of the design semantics and construct a chain of GPS specification operation operators with a specific execution order;

[0069] S3. Based on the machine tool measurement constraints, and while keeping the specific execution order unchanged, perform GPS dual mapping on each operator in the GPS standard operation operator chain by operator parameter transformation to generate a verification operation operator chain that is consistent with the design semantic structure.

[0070] S4. Construct a multi-scale skin model based on the on-machine measurement point set data, and process the multi-scale skin model based on the inspection operation operator chain to generate an alternative inspection surface model.

[0071] S5. Calculate the inspection geometric feature value based on the alternative inspection surface model, and at the same time calculate the design geometric feature value of the nominal geometric model according to the GPS specification operation operator chain to obtain the geometric error amount;

[0072] S6. Calculate the overall residual and operator execution metric on the inspection side based on the alternative inspection surface model, and calculate the operator conservation invariant by combining the overall residual and operator execution metric on the specification side; compare the operator conservation invariant with the reference operator conservation invariant to determine the operational consistency status between the inspection operation process and the design specification.

[0073] S7. Based on the geometric error and the operational consistency determination result, perform in-machine precision control adjustment.

[0074] Specifically, the GPS dual mapping refers to, while maintaining the specific execution order of the GPS standard operation operator chain, transforming the parameters of each operator in the operator chain to adapt it to the machine tool measurement constraints, thereby generating a verification operation operator chain that corresponds to the design semantics in structure and matches the machine tool measurement environment in parameters.

[0075] In step S1, GPS design specification data, nominal geometric model, and on-machine measurement point set data are first acquired. The GPS design specification data includes information such as tolerance types, datum system, and evaluation rules; the nominal geometric model represents the theoretical geometric structure under design conditions; and the on-machine measurement point set data is collected by the machine tool measuring device. Subsequently, the on-machine measurement point set data undergoes coordinate unification processing to ensure that the measurement data and the nominal geometric model are in the same workpiece datum coordinate system, establishing a unified spatial reference for subsequent geometric processing.

[0076] In step S2, tolerance types and benchmark system information are analyzed based on GPS design specification data. Geometric feature extraction operators, filtering operators, geometric correlation operators, geometric construction operators, and error evaluation operators are selected from a preset operator template library and arranged in the processing order specified in the specification to form a GPS specification operation operator chain. This operator chain is used to express the geometric evaluation process on the design side.

[0077] In step S3, machine tool measurement constraints are acquired, including measurement path information, measurement accessibility information, and measurement strategy information. A dual mapping process is then performed on the GPS standard operation operator chain. By transforming the operator execution parameters and execution methods to match the machine tool measurement conditions, a verification operation operator chain is generated. This verification operation operator chain is structurally consistent with the GPS standard operation operator chain and conforms to the on-machine measurement environment in its execution form.

[0078] In step S4, a multi-scale surface model is constructed based on the on-machine measurement point set data under a unified coordinate system. This model, formed by scale decomposition of the measurement point data and combination with the nominal geometric model, can represent the multi-level geometric deviations of the actual processed surface. Subsequently, following the execution order of the inspection operation operator chain, the multi-scale surface model undergoes geometric feature extraction, filtering, geometric correlation, and geometric construction processing in sequence to generate an alternative inspection surface model. This model serves as the unified geometric carrier for subsequent error calculations.

[0079] In step S5, the inspection geometric feature values ​​are calculated based on the alternative inspection surface model, and the design geometric feature values ​​are calculated on the nominal geometric model according to the GPS specification operation operator chain. The two are then calculated under the same feature expression to obtain the geometric error. This geometric error is used to characterize the deviation between the actual processing state and the design specifications.

[0080] In step S6, the overall geometric metric calculation is performed on the alternative inspection surface model to obtain operator-conserved invariants, and the consistency of the inspection operation operator chain execution process is determined based on these operator-conserved invariants. This determination process determines the consistency state of the inspection operation by comparing the operator-conserved invariants obtained under different inspection paths.

[0081] In step S7, the machining parameter adjustment amount is determined based on the geometric error amount, and the machining parameter adjustment amount is corrected in combination with the operation consistency judgment result. The corrected machining parameter adjustment amount is then input into the machining equipment. After machining is completed, the in-machine measurement point set data is reacquired, and steps S1 to S6 are repeated to form an in-machine closed-loop control process.

[0082] Through the above steps, a unified processing framework is achieved between GPS design specifications and in-machine verification procedures, enabling in-machine verification and accuracy control of geometric errors under a unified data structure and unified operator structure.

[0083] like Figure 2 As shown, the specific process of coordinate unification in step S1 is as follows:

[0084] Obtain the coordinate transformation relationship between the machine tool coordinate system and the measurement coordinate system;

[0085] Spatial transformation is performed on the on-machine measurement point set data based on coordinate transformation relationships;

[0086] The transformed on-machine measurement point set data is unified under the workpiece reference coordinate system.

[0087] Specifically, during the machine measurement process, the spatial coordinate data output by the measuring device is expressed in the machine tool measurement coordinate system, while the nominal geometric model and GPS design specification data are defined in the workpiece reference coordinate system. Therefore, it is necessary to establish a unified correspondence between different coordinate systems.

[0088] First, the on-machine measurement point set data is acquired. This data is obtained by sampling the workpiece surface using the measuring device during machining. Each measurement point contains three-dimensional spatial position information, and all measurement points constitute a discrete spatial point set. This point set records the spatial distribution of the actual machined surface. Next, the datum correspondence between the machine tool coordinate system and the workpiece datum coordinate system is determined. This process is completed by reading the datum elements defined in the GPS design specification data. Datum elements include datum surfaces, datum axes, or datum points, and the set of actual measurement datum points corresponding to these datum elements is acquired through on-machine measurement.

[0089] After obtaining the set of measurement reference points and the set of nominal reference points, the spatial correspondence between the two sets of reference points is established, and the coordinate transformation parameters are calculated. The coordinate transformation relationship is expressed using a rigid body spatial transformation model, and its spatial transformation process is represented as follows:

[0090] ;

[0091] in: Represents the first in the machine tool measurement coordinate system Coordinates of the measurement points;

[0092] This indicates the first [number] after transformation to the workpiece reference coordinate system. Coordinates of the measurement points;

[0093] R represents the spatial rotation matrix, which describes the rotational relationship between the coordinate axes.

[0094] t represents the spatial translation vector, used to describe the positional offset between the origins of two coordinate systems.

[0095] The rotation matrix is ​​determined by the attitude relationship between reference directions, and its calculation is completed through alignment of reference feature directions. The translation vector is calculated from the position difference of reference points, thus forming a complete spatial transformation relationship. After obtaining the coordinate transformation relationship, spatial transformation processing is sequentially performed on all on-machine measurement points, mapping all measurement points from the machine tool measurement coordinate system to the workpiece reference coordinate system. The topological relationship between points remains unchanged during the spatial transformation process, thereby ensuring the geometric consistency of the measurement data.

[0096] To ensure the stability of the coordinate unification process, a datum consistency check is performed on the transformation results after spatial transformation. The check is achieved by calculating the spatial deviation between the transformed datum measurement points and the nominal datum points, and the deviation calculation is expressed as follows:

[0097] ;

[0098] in: This represents the spatial deviation of the k-th reference point;

[0099] Represents the coordinates of the nominal reference point;

[0100] This represents the converted coordinates of the measurement reference point;

[0101] This represents the Euclidean norm, which is the Euclidean norm of the difference vector between the coordinates of two points. Its value is equal to the Euclidean distance between the two points.

[0102] When all reference point deviations meet the preset reference consistency conditions, the coordinate unification process is complete, and an on-machine measurement point set data in a unified coordinate system is generated. The on-machine measurement point set data after coordinate unification is expressed using the same spatial reference as the nominal geometric model. This ensures that subsequent GPS standard operation operator chain construction, multi-scale surface model generation, and verification operation operator chain execution are all based on a unified geometric reference, avoiding additional spatial errors introduced due to coordinate system differences, thus guaranteeing the reference consistency of subsequent geometric feature calculations.

[0103] The specific process of constructing the canonical operation operator chain in step S2 is as follows:

[0104] Analyze the tolerance types and datum information in GPS design specification data;

[0105] Based on the analysis results, geometric feature extraction operators, filtering operators, geometric correlation operators, geometric construction operators, and error evaluation operators are selected from the preset operator template library;

[0106] The selected operators are arranged and combined according to the operating sequence specified in the GPS design specifications to form a GPS standard operating operator chain.

[0107] Specifically, GPS design specification data is stored in a structured format, including tolerance annotation information, controlled geometric feature identification information, benchmark system information, and evaluation rule information. To ensure the determinism of the operator chain construction process, semantic parsing is first performed on the GPS design specification data. Let the GPS design specification data be represented as a dataset G, then: ;in: L represents the l-th specification description item; L represents the number of specification description items.

[0108] Each specification description item It must include at least a tolerance type identifier, a datum sequence identifier, and a target geometric feature identifier. During the parsing process, a field splitting operation is performed on each specification description item to obtain the following information:

[0109] Tolerance type identifier ; Baseline sequence identifier Target geometric feature identification Tolerance type identifier Used to determine the error assessment category, such as shape tolerance, orientation tolerance, or position tolerance; reference sequence identifier. Used to determine the geometric association order; target geometric feature identifier Used to determine the controlled geometric entity.

[0110] After completing the specification analysis, identify the tolerance type. Match the corresponding operator combination structure from the preset operator template library. The operator template library is stored as an operator set B, represented as:

[0111] ;

[0112] in: This represents the template for the geometric feature extraction operator of the k-th class;

[0113] This represents the template for the k-th type of filter operator;

[0114] Represents the template for the k-th type of geometric correlation operator;

[0115] Represents the template for the k-th type of geometric construction operator;

[0116] This represents the template for the error evaluation operator of type k.

[0117] The operator template contains operator execution rules, input data types, output data types, and parameter constraint information. This is based on the tolerance type identifier. Determine the required operator category. For example, when the tolerance type is identified as a positional tolerance, the predefined combination structure in the operator template library includes geometric feature extraction operators, reference geometry association operators, controlled feature geometry construction operators, and error evaluation operators; when the tolerance type is identified as a shape tolerance, then the corresponding geometric feature extraction operators, filtering operators, and error evaluation operators are selected.

[0118] After selecting the operator category, identify it based on the target geometric features. Determine the parameter set for the geometric feature extraction operator. Let the target geometric feature region be the set of regions. Then the input region of the geometric feature extraction operator is represented as: ;in: Represents the nominal geometric model; This indicates the region where the controlled geometric feature is located.

[0119] The parameters of the filtering operators are determined by the filtering requirements defined in the specification, including the filtering type and filtering cutoff conditions. The parameters of the geometric correlation operators are identified based on the reference sequence. The reference elements in the reference sequence are established with sequential relationships. The geometric construction operator selects the corresponding geometric model representation based on the type of controlled geometric feature, such as a plane, cylindrical surface, or space curve model. The error evaluation operator determines the error calculation rules based on the tolerance type identifier. After selecting and binding parameters for all types of operators, the operators are arranged according to the processing logic specified in the GPS design specifications. Let the constructed GPS specification operation operator chain be represented as:

[0120] ;

[0121] Where: E represents the geometric feature extraction operator;

[0122] H represents the filter operator;

[0123] A represents the geometric correlation operator;

[0124] C represents the geometric construction operator;

[0125] M represents the error evaluation operator;

[0126] This indicates that the operators are executed sequentially.

[0127] The operator order is determined by the tolerance type and the reference sequence, with different tolerance categories corresponding to different arrangement rules. After arrangement, each operator instance and its parameter set are encapsulated into a structured operator chain data structure for subsequent dual mapping processing.

[0128] Through the above steps, a chain of GPS specification operation operators consistent with the GPS design specification data is formed, enabling the geometric evaluation process on the design side to be expressed in the form of executable operators. This provides a unified structural basis for the subsequent generation of GPS dual mapping and verification operation operator chains, thereby ensuring the consistent expression of design semantics in the subsequent verification process.

[0129] like Figure 3 As shown, the GPS standard operation operator chain on the design side, through constraints and adjustments of machine tool measurement conditions, generates the inspection operation operator chain on the inspection side via dual mapping D. Specifically, in step S3, the step of performing GPS dual mapping on the GPS standard operation operator chain based on machine tool measurement constraints includes:

[0130] Obtain machine tool measurement path constraint information and measurement reachability information;

[0131] Match each operator in the GPS standard operation operator chain with the machine tool measurement path constraint information;

[0132] The execution parameters and execution methods of each operator are converted based on the matching results;

[0133] Generate a chain of verification operation operators according to the transformed operator order.

[0134] Specifically, the machine tool measurement constraints are derived from the operating status data of the machine tool control system and the measurement system, including measurement path constraint information, measurement accessibility information, and measurement posture restriction information.

[0135] Measurement path constraint information is used to limit the range of movable trajectories of the measuring device in the machine tool coordinate system; measurement accessibility information is used to limit the approach direction and interference conditions between the measuring device and the workpiece; measurement attitude constraint information is used to limit the range of spatial attitude angles of the measuring device. Let the GPS standard operation operator chain be represented as:

[0136] ;

[0137] in: This represents the geometric feature extraction operator on the design side;

[0138] Indicates the design of the side-filtering operator;

[0139] Indicates the geometric correlation operator on the design side;

[0140] Indicates the geometric construction operator on the design side;

[0141] This represents the design-side error evaluation operator.

[0142] To establish the correspondence between the design-side operators and the verification-side operators, a dual mapping operator D is defined to implement operator-level mapping, and its expression is as follows: ;in: This represents a chain of check operation operators; Represents the set of machine tool measurement constraints; This represents the dual mapping function of the operator. The set of machine tool measurement constraints is represented as: ;in: Represents the set of measurement path constraints; Represents the set of reachability constraints for measurement; This represents the set of measurement attitude constraints.

[0143] During the mapping process, matching and transformation operations are performed on each operator in the GPS canonical operation operator chain one by one. For the geometric feature extraction operator... First, based on the measurement path constraint set Determine whether the target geometric region is within the measurable range; if the target region is outside the coverage of the current measurement path, adjust the region division strategy to ensure that the measurement path covers all target regions.

[0144] For filter operators The adjustment rules for the filtering parameters are determined based on the density of measurement points and the distribution of the sampling path. The adjustment of the filtering parameters is based on the measurement point spacing distribution function. Let the measurement point spacing function be expressed as: ;in: This represents the i-th measurement point in a unified coordinate system;

[0145] This represents the distance to the nearest neighbor of this point; This represents the Euclidean norm.

[0146] Based on the statistical results of the distance distribution between measurement points, the sampling parameters of the filtering operator are reset to ensure that the filtering process is consistent with the distribution of measurement points.

[0147] For geometric correlation operators Based on the set of reachability constraints The reference direction is corrected. If the reference direction is inconsistent with the measurement direction, it is adjusted by calculating the angle between the measurement direction vector and the nominal reference direction vector. The angle calculation is expressed as follows: ;in: Represents the measurement direction vector; Represents the nominal reference direction vector; This represents the angle between two vectors. When the angle meets the preset attitude constraints, the association operation is performed; otherwise, the measurement attitude parameters are replanned.

[0148] For geometric construction operators Based on the set of measurement attitude constraints The input data for model construction is filtered to remove measurement points that do not meet the attitude conditions, thus ensuring that the data used in the construction process meets the actual measurement attitude requirements. For the error evaluation operator... To maintain consistency between its error calculation rules and the design side, only the input data source is replaced, enabling it to perform error calculations based on the geometric model of the inspection side. After completing the parameter adjustments and execution mode conversions of the above operators, the inspection operation operator chain is constructed according to the original operator order, expressed as follows:

[0149] ;

[0150] in: This is the geometric feature extraction operator after path constraint matching;

[0151] This is the filter operator adjusted after the distribution of measurement points;

[0152] This refers to the geometric correlation operator after correction to the reference direction;

[0153] The geometric construction operator is selected after attitude screening;

[0154] This is an error evaluation operator performed based on test data.

[0155] The inspection operation operator chain is structurally consistent with the GPS standard operation operator chain, and its execution parameters are consistent with the machine tool measurement constraints.

[0156] Through the above dual mapping process, the geometric evaluation logic on the design side is transformed into an executable inspection operation process in the machine environment. While maintaining the consistency of the operator order, the physical matching of the execution parameters is achieved, thereby ensuring the structural consistency between the design specifications and the machine inspection.

[0157] It is important to emphasize here that "maintaining a specific execution order" means that at the macroscopic structural level of the operator chain, the execution order of various types of operators (such as filtering operators, geometric correlation operators, and geometric construction operators) is strictly consistent with the design side. "Operator-by-operator parameter transformation," on the other hand, involves physically adapting the execution parameters (such as filter cutoff wavelength and reference correlation direction) within the same operator based on machine tool measurement constraints. This adjustment of internal parameters does not change the macroscopic structural order of the operator chain, thus not violating the core principle of dual mapping, but rather ensuring the maximum preservation of design semantics within the premise of physical realizability.

[0158] Please refer to the reference. Figure 4 Step S4 includes two processes: constructing a multi-scale skin surface model and generating an alternative test surface model based on a chain of test operation operators. Specifically, the step of constructing the multi-scale skin surface model based on the on-machine measurement point set data includes:

[0159] The on-machine measurement point set data is subjected to scale decomposition processing;

[0160] Obtain geometric deviation data at multiple scale levels;

[0161] Combine the geometric deviation data of each scale layer with the nominal geometric model;

[0162] Generate multi-scale skin surface models.

[0163] Furthermore, the step of generating the alternative test surface model includes:

[0164] Controlled by the chain of inspection operation operators generated in step S3, geometric feature extraction, filtering, geometric association and geometric construction are sequentially performed on the multi-scale skin model according to the execution parameters determined by each operator in the chain of inspection operation operators after the GPS dual mapping transformation;

[0165] Based on the processing results, an alternative verification surface model corresponding to the GPS standard operation operator chain structure is constructed.

[0166] Specifically, after completing the coordinate unification process in step S1, the on-machine measurement point set data in a unified coordinate system is obtained, denoted as the point set. The point set reflects the discrete spatial distribution of the actual machined surface. To distinguish geometric deviations at different scales, it is necessary to perform scale decomposition on the on-machine measurement point set data. First, an expression for the deviation of the measurement points relative to the nominal geometric model is established. Let the nominal geometric model be... For each measurement point Determine the corresponding nearest point on the nominal geometric model. The deviation vector is then expressed as: ;in: This represents the i-th measurement point in a unified coordinate system; Representing the nominal geometric model and The corresponding nearest point; This represents the spatial deviation vector of the i-th measurement point relative to the nominal geometric model.

[0167] Construct a deviation set from all deviation vectors To achieve scale decomposition, the bias set is layered in the frequency or spatial domain. Specifically, an adjacency graph is constructed based on spatial neighborhood relationships, and the bias data is then layered according to neighborhood scale parameters. Let the number of scale layers be K, then the bias vector can be represented as the sum of multiple scale components: ;in: This represents the bias component of the k-th scale layer; K represents the number of scale layers. The division of each scale layer is determined based on the spatial neighborhood radius or equivalent filtering parameters. During the scale decomposition process, larger scale layers correspond to overall shape biases, while smaller scale layers correspond to local surface biases. After scale decomposition, a set of geometric bias data for multiple scale layers is obtained. Subsequently, the geometric deviation data of each scale layer are combined with the nominal geometric model. This combination process is achieved by superimposing the deviation components onto the nominal geometric model, forming a multi-scale skin model. It is expressed as: Where: S represents a multi-scale skin model; D represents the nominal geometric model; D(k) represents the bias data set of the k-th scale layer.

[0168] In terms of data structure, the multi-scale skin model includes a nominal geometric topology, scale layer index information, and a corresponding set of bias vectors, allowing specific scale layers to be selected for computation as needed during subsequent operator execution. After the multi-scale skin model is constructed, processing is performed sequentially according to the chain of verification operation operators generated in step S3. First, geometric feature extraction is performed. The target geometric region is determined based on the geometric feature extraction operators in the verification operation operator chain, and the data of the corresponding region in the multi-scale skin model is filtered to form a subset of target feature data. Then, filtering is performed. Based on the filtering operator parameters in the verification operation operator chain, the selected scale layer or multiple scale layers are reconstructed or smoothed to form a data set for geometric correlation.

[0169] Then, geometric correlation processing is performed. Based on the geometric correlation operators in the verification operation operator chain, the target feature data subset is aligned to a benchmark. This process is achieved by minimizing the sum of squared residuals between the data points in the target feature data subset and the constructed geometric model; its objective function is expressed as:

[0170] ;

[0171] in: Indicates the value of the correlation error function;

[0172] This indicates the number of data points in the subset of target feature data;

[0173] This represents the residual vector from the i-th data point to the constructed geometric model;

[0174] This represents the Euclidean norm.

[0175] By minimizing the objective function value of the correlation error, an optimized correlation is achieved between the target feature data subset and the constructed geometric model.

[0176] After completing the geometric association processing, geometric construction processing is performed. Based on the geometric construction operators in the verification operation operator chain, the associated dataset is geometrically reconstructed to generate verification geometric elements, such as planar models, cylindrical surface models, or space curve models. Finally, the constructed verification geometric elements and the data index information involved in the construction are uniformly encapsulated to generate an alternative verification surface model. The alternative verification surface model includes the following: constructed geometric model parameters, the set of measurement points involved in the construction, scale layer information, and associated residual data.

[0177] Through the above-mentioned multi-scale surface model construction and verification operation operator chain execution process, the on-machine measurement point set data completes scale-level expression and geometric reconstruction processing under a unified geometric framework, forming a structured alternative verification surface model, providing a consistent data foundation for subsequent verification geometric feature value calculation and operator conservation invariant calculation.

[0178] Furthermore, in step S5, the step of calculating the test geometric feature values ​​based on the alternative test surface model includes:

[0179] Determine the inspection geometry elements based on the alternative inspection surface model;

[0180] The inspection geometric elements are calculated based on the error evaluation operators in the inspection operation operator chain;

[0181] Obtain the geometric characteristic values ​​for testing.

[0182] Specifically, an alternative inspection surface model has been generated in step S4. The alternative inspection surface model includes the constructed inspection geometric element parameters, the set of measurement points involved in the construction, and the associated residual data. The inspection geometric elements can be geometric entities such as planes, lines, cylindrical surfaces, conical surfaces, or space curves, depending on the tolerance type.

[0183] First, the verification geometric elements are determined based on the outputs of the geometric construction operators in the verification operation operator chain. Let the constructed verification geometric elements be represented as a set. ,but:

[0184] ;

[0185] in: This represents the j-th geometric element being tested;

[0186] r represents the number of geometric elements to be inspected.

[0187] When the geometric element being tested is a plane, its parameters can be expressed as a combination of the plane normal vector and the plane constant term; when the geometric element being tested is a cylindrical surface, its parameters can be expressed as the axis direction vector, the axis position parameter, and the radius parameter. All of these parameters are derived from the calculation results of the geometric construction operator in step S4.

[0188] Subsequently, the error is calculated for the inspected geometric elements according to the error evaluation operator in the inspection operation operator chain. The error evaluation operator determines the error calculation method based on the tolerance type. Taking distance-type errors as an example, let the set of measurement points involved in the error calculation be... The corresponding test geometric elements are Then the distance from each measurement point to the geometric element being tested is expressed as: ;in: This represents the distance from the i-th measurement point to the geometric element being inspected; dist() represents the shortest distance function from the point to the geometric element. Indicates the measurement points involved in error assessment; This represents the geometric element to be inspected. In positional tolerances, the error assessment operator establishes the relative positional relationship between the controlled feature and the reference feature based on the reference system. Let the reference geometric element be... The controlled geometric elements are Then its relative offset is expressed as:

[0189] ;

[0190] in:

[0191] Represents the feature center coordinate vector of the controlled geometric element;

[0192] Represents the coordinate vector of the feature center of the reference geometric element;

[0193] This indicates the spatial offset between the two.

[0194] This represents the Euclidean norm.

[0195] In shape tolerances, the error assessment operator calculates the maximum deviation value based on the distance distribution from all measurement points to the inspection geometry. Let the distance set be... The shape error value is then expressed as: ;in: Indicates the testing of geometric eigenvalues; This represents the maximum value in the distance set; This represents the minimum value in the distance set. The error assessment operator selects the appropriate error expression method according to different tolerance categories and outputs the corresponding test geometric feature values.

[0196] Simultaneously, based on the GPS standard operation operator chain constructed in step S2, the same error assessment process is performed on the nominal geometric model to obtain the design geometric eigenvalues. The design geometric eigenvalues ​​are expressed as: ;in: This represents the design geometric characteristic value; its calculation method is consistent with the evaluation structure for verifying geometric characteristic values, except that the input data source is the nominal geometric model.

[0197] Finally, the geometric eigenvalues ​​will be tested. With design geometric eigenvalues By performing a corresponding comparison, the geometric error is obtained:

[0198] ;

[0199] in: Indicates the amount of geometric error;

[0200] Indicates the testing of geometric eigenvalues;

[0201] This represents the design geometric characteristic value.

[0202] Through the above processing, the determination of inspection geometric elements, the execution of error evaluation operators, and the acquisition of inspection geometric feature values ​​are realized. This enables the geometric features under the actual processing state to be quantitatively expressed under an evaluation structure consistent with the design specifications, thereby providing a unified error input for subsequent operator conservation invariant calculations and in-machine precision control and adjustment.

[0203] Furthermore, in step S6, the step of calculating the operator-conserved invariants based on the alternative test surface model includes:

[0204] Perform overall geometric measurement calculations on the alternative test surface model;

[0205] Operator metric data is generated based on the operator execution information recorded in the operator chain of the inspection operation;

[0206] Calculate the operator conserved invariants based on the operator metric data.

[0207] In step S6, the steps for determining operational consistency based on operator conservation invariants include:

[0208] Obtain the operator-conserved invariants obtained during the execution of different test operation operator chains;

[0209] Comparative analysis of operator-conserved invariants;

[0210] The consistency status of the chain of test operation operators is determined based on the comparison results.

[0211] Specifically, after generating the alternative test surface model in step S4, the alternative test surface model includes the constructed geometric element parameters, the set of measurement points involved in the calculation, residual data, and scale layer information. The operator conservation invariants are obtained by calculating the overall geometric metrics based on the above data.

[0212] First, a global geometric measurement calculation is performed on the alternative test surface model. Let the set of measurement points involved in the calculation be... The corresponding test geometric elements are Then the residual vector from each measurement point to the test geometric element is represented as: ;in: Represents the residual vector of the i-th measurement point; This represents the i-th measurement point; This indicates the projection point of the measurement point onto the inspected geometric element.

[0213] In the calculation of the overall geometric metric, all residual vectors are comprehensively measured to form the overall residual metric value. The overall residual metric value is expressed as: ;in: This represents the overall residual measure. Indicates the number of measurement points involved in the calculation; This represents the Euclidean length of the residual vector. Subsequently, operator metric data is generated based on the operator execution information recorded in the test operation operator chain. The operator execution information includes the execution order of each operator, execution parameters, the data indexes involved in the calculation, and the scale layer identifier. Let the set of operator execution information be represented as: ;in: Indicates the first The execution record of each operator; s represents the number of operators. Includes execution parameter vector and the parameters of the data scale involved Based on the operator execution information, operator metric data is generated. The operator metric values ​​are obtained by weighting the execution parameters of each operator and their corresponding residual contributions.

[0214] ;

[0215] in: Represents the operator metric;

[0216] This represents the weight parameters of the j-th operator;

[0217] This represents the value of the operator metric function calculated based on the execution parameters and the scale of the data involved.

[0218] This represents the execution parameter vector of the j-th operator;

[0219] This represents the number of data points involved in the calculation of the j-th operator.

[0220] As a preferred embodiment, the operator metric function Specifically, it can be defined as the product of the norm of the execution parameter and the size of the data involved, that is... This definition concisely quantifies the overall cost during operator execution, facilitating the subsequent calculation of conserved invariants.

[0221] In obtaining the overall residual metric value With operator metric Then, the two are combined to obtain the operator-conserved invariants. The operator-conserved invariants are expressed as: Where: I represents the operator-conserved invariant; This represents the overall residual measure. This represents the operator metric.

[0222] It should be noted that the above operators are conserved invariants. The calculation formula is ,in This is the dimensionless normalization function defined in claim 8. This process is used to eliminate the overall residual metric. With operator metric The dimensional differences between them ensure that they can be summed under the same dimension, thus objectively reflecting the structural consistency of the verification operation process. The 'conservation' mentioned here means that this invariant remains constant if and only if the actual execution structure of the verification operation operator chain maintains strict duality with the design semantics of the GPS specification operation operator chain. It tends to a stable minimum value in multiple measurements and statistics; this invariant... The degree of deviation from the stable minimum value objectively and quantitatively reflects the degree of inconsistency between the current on-machine inspection operation and the design specifications.

[0223] During the execution of different test operation operator chains, the corresponding operator-conserved invariants are calculated respectively. Let the invariants corresponding to the two test operation operator chains be respectively... and In the process of determining operational consistency, the operator-conserved invariants are compared and analyzed. The comparison values ​​are expressed as:

[0224] ;

[0225] in: This represents the difference between the conserved invariants of two operators. When the difference meets the preset consistency criteria, it is determined that the two test operation operator chains are consistent in execution structure; when the difference does not meet the consistency criteria, the source of the difference is further located based on the operator execution record L to determine the operator node that caused the inconsistency.

[0226] Through the above-mentioned operator conservation invariant calculation and consistency determination process, the overall geometric measurement results of the substitute inspection surface model and the execution information of the inspection operation operator chain are uniformly expressed, so that the inspection operation structure has comparable quantitative indicators under different execution paths, thereby providing a stable structural determination basis for subsequent on-machine precision control and adjustment.

[0227] It should be further clarified that the "combining the overall residual on the specification side with the operator execution metric" mentioned in step S6 refers to the calculation of the reference operator's conserved invariants. At that time, using the same method as the inspection side, the overall residual metric and operator metric of the specification side are calculated based on the nominal geometric model and the GPS specification operation operator chain, and then a reference invariant is generated. This reference invariant is used in conjunction with the operator-conserved invariant calculated during the current inspection operation. A comparison is made to determine the operational consistency status. In other words, the overall residuals on the specification side do not directly participate. It is not used for calculation, but rather as a comparison benchmark for consistency determination.

[0228] Furthermore, in step S7, the step of adjusting the in-machine precision control based on the geometric error and operational consistency determination results includes:

[0229] The adjustment amount of the machining parameters is determined based on the geometric error.

[0230] Adjust the processing parameters based on the results of the operational consistency assessment.

[0231] Input the corrected processing parameter adjustments into the processing equipment;

[0232] Then, based on the processed workpiece, the in-machine measurement and inspection operations of steps S1 to S6 are re-executed.

[0233] Specifically, the geometric error has been obtained in step S5, and let the geometric error be expressed as... The geometric error is derived from the calculated difference between the inspection geometric characteristic value and the design geometric characteristic value, and is used to characterize the deviation between the current processing state and the design specifications.

[0234] First, determine the adjustment amount of the machining parameters based on the geometric error. Let the set of adjustable parameters of the machining equipment be represented as: ;in: Let represent the k-th machining parameter; q represents the number of machining parameters; machining parameters include feed rate, depth of cut, tool compensation, or tool axis posture parameters. Establish the mapping relationship between geometric error and machining parameters. Let the error response matrix be represented by H, then the machining parameter adjustment is represented as:

[0235] ;

[0236] in: This represents a vector of processing parameter adjustment amounts; Represents the error response matrix; Indicates the amount of geometric error;

[0237] The error response matrix H is obtained by calibrating the response relationship between changes in machining parameters and changes in geometric errors. Matrix elements This represents the influence coefficient of the i-th processing parameter on the j-th geometric error component. Subsequently, the processing parameter adjustment is corrected based on the operational consistency determination result obtained in step S6. Let the difference in the operator's conserved invariant corresponding to the operational consistency determination result be . Then define the correction factor. Represented as:

[0238] ;

[0239] in: Indicates the correction factor; This represents the difference between the operator's conserved invariants.

[0240] In practical engineering applications, the above-mentioned operational consistency correction coefficient The mechanism of action has a prerequisite tolerance criterion: only if the difference between the above operator's conserved invariants is... When the deviation is less than the preset consistency tolerance threshold (i.e., the characterization and testing operations basically conform to the design specifications, with only minor structural deviations), the system uses the above formula to calculate... And adjust the processing parameters. Perform image stabilization fine-tuning (at this time) Close to 1); if the difference in the judgment is If the value is greater than or equal to the preset consistency tolerance threshold, it indicates that a serious deviation or measurement failure has occurred during the on-machine inspection operation. In this case, the system will actively intercept the current processing parameter adjustment command. Instead of inputting the data into the processing equipment, the alarm prompt process for re-execution of in-machine inspection is triggered, thereby avoiding invalid or even harmful processing error compensation based on erroneous measurement data with inconsistent structures.

[0241] The machining parameter adjustment amount is corrected using a correction factor. The corrected machining parameter adjustment amount is expressed as follows: ;in: This represents the adjusted machining parameters. After obtaining the adjusted machining parameters, they are input into the machining equipment control system. The machining equipment control system updates the corresponding machining parameters according to the adjustment and executes the next round of machining. After machining is completed, in-machine measurement is re-executed to obtain new in-machine measurement point data, and steps S1 to S6 are executed sequentially to calculate the new geometric error and operator conservation invariants, thus forming a cyclic control process. During the cyclic control process, when the geometric error meets the preset tolerance judgment condition and the operation consistency judgment result meets the consistency requirement, the machining parameter adjustment is stopped, and the in-machine accuracy control process ends.

[0242] Through the above steps, the geometric error and the operator's conserved invariants are incorporated into the calculation of machining parameter adjustment, so that the machining parameter adjustment reflects both the geometric deviation state and the consistency state of the inspection operation structure, thereby realizing in-machine precision control and adjustment based on the inspection results.

[0243] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. 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 method for on-machine inspection and accuracy control of precision machining geometric errors based on GPS dual mapping, characterized in that, Includes the following steps: S1. Obtain the GPS design specification data, nominal geometric model, and on-machine measurement point set data of the workpiece, and perform coordinate unification processing on the on-machine measurement point set data; S2. Based on the GPS design specification data, analyze the geometric constraints of the design semantics and construct a chain of GPS specification operation operators with a specific execution order; S3. Based on the machine tool measurement constraints, and while keeping the specific execution order unchanged, perform GPS dual mapping on each operator in the GPS standard operation operator chain by operator parameter transformation to generate a verification operation operator chain that is consistent with the design semantic structure. S4. Construct a multi-scale skin model based on the on-machine measurement point set data, and process the multi-scale skin model based on the inspection operation operator chain to generate an alternative inspection surface model. S5. Calculate the inspection geometric feature value based on the alternative inspection surface model, and at the same time calculate the design geometric feature value of the nominal geometric model according to the GPS specification operation operator chain to obtain the geometric error amount; S6. Calculate the overall residual and operator execution metric on the test side based on the alternative test surface model, and calculate the operator conservation invariants by combining the overall residual and operator execution metric on the specification side. The operator-conserved invariants are compared with the reference operator-conserved invariants to determine the operational consistency between the inspection process and the design specifications. S7. Based on the geometric error and the operational consistency determination result, perform in-machine precision control adjustment.

2. The method according to claim 1, characterized in that, The coordinate unification process described in step S1 includes: The coordinate transformation relationship between the machine tool coordinate system and the measurement coordinate system is obtained by minimizing the spatial deviation between the set of measurement reference points and the set of nominal reference points and fitting a rigid body transformation matrix. Translation vector To establish; Based on the coordinate transformation relationship, the on-machine measurement point set data is spatially transformed and unified to the workpiece reference coordinate system; Perform a benchmark consistency check on the converted benchmark measurement points, assuming the first... The nominal benchmark points are , No. The converted measurement reference points are: Its deviation is defined as ,in Represents the Euclidean norm, when all reference points deviate... All are less than the preset benchmark consistency tolerance threshold. At that time, the coordinates were determined and the unified processing was completed.

3. The method according to claim 1, characterized in that, The step S2, which involves constructing a chain of GPS canonical operation operators with a specific execution order, includes: Analyze the tolerance types and datum information in GPS design specification data; Based on the tolerance type and reference information, the selected operators are arranged and combined according to the operation sequence specified in the GPS design specifications to construct the GPS standard operation operator chain.

4. The method according to claim 1 or 3, characterized in that, The GPS dual mapping mentioned in step S3 includes: Obtain machine tool measurement path constraint information and measurement reachability information; Match each operator in the GPS standard operation operator chain with the machine tool measurement path constraint information; The execution parameters and execution methods of each operator are converted based on the matching results; The parameter transformation of the filter operator includes: based on the distribution function of the distance between measurement points. Statistical characteristics can be used to reset the filter window size or cutoff frequency. The resetting of the filter window size or cutoff frequency includes: using the median of the measurement point spacing. As a statistical characteristic, the cutoff wavelength of the filter operator is set. ,in The preset empirical coefficients are determined in advance based on the accuracy of the machine tool measuring device, the surface characteristics of the workpiece being processed, and the stage of the processing procedure, so that the filtering parameters are adapted to the local density of the measurement points; The parameter transformation of the geometric association operator includes: calculating the angle between the measurement direction vector and the nominal reference direction vector based on the measurement accessibility information; if the angle exceeds the preset attitude constraint range determined by the machine tool measurement attitude limit, then adjusting the reference association direction or replanning the measurement attitude.

5. The method according to claim 1, characterized in that, The steps in step S4 for constructing the multi-scale skin model include: The on-machine measurement point set data is subjected to scale decomposition processing to obtain geometric deviation data at multiple scale levels; The geometric deviation data of each scale layer are combined with the nominal geometric model to generate a multi-scale skin model; The scale decomposition process is achieved by constructing an adjacency graph according to spatial neighborhood relationships and performing hierarchical decomposition of deviation data based on neighborhood scale parameters determined by workpiece geometric features or measurement point density. The division of each scale layer is determined based on the spatial neighborhood radius or equivalent filtering parameters.

6. The method according to claim 5, characterized in that, The step of generating the alternative test surface model in step S4 includes: According to the execution parameters determined by the test operation operator chain, the multi-scale skin model is sequentially subjected to geometric feature extraction processing, filtering processing, geometric association processing and geometric construction processing; Based on the processing results, an alternative verification surface model corresponding to the GPS standard operation operator chain structure is constructed.

7. The method according to claim 1, characterized in that, The steps in step S6 for calculating the operator-conserved invariants include: The overall geometric measurement of the alternative test surface model is performed to obtain the overall residual measurement value. ,in To determine the total number of measurement points involved in the calculation of the geometric elements, For the first The residual vector from each measurement point to the test geometric element; Operator metrics are generated based on the operator execution information recorded in the operator chain of the inspection operation. ,in The total number of operators participating in the generation of alternative test surface models, To pre-calibrate and determine the first operator based on its type and its influence on the final error evaluation result. The weights of each operator According to the first Execution parameters of the operators norm and the scale of the data involved The operator metric function for computation; The overall residual metric and the operator metric are respectively passed through After performing dimensionless normalization on the function and summing the results, we obtain the operator-conserved invariants. .

8. The method according to claim 7, characterized in that, The The function is a dimensionless normalization processing function, defined as follows: ; in, The metric to be normalized and These are the minimum and maximum values ​​of the metric obtained statistically from historical data, respectively; if historical data is insufficient or it is the first time running, the... and Experienced or theoretical values ​​preset according to machining accuracy requirements and machine tool characteristics can be used to replace them, thereby eliminating the dimensional differences between different measurement values, so that the overall residual measurement value and the operator measurement value after processing can be summed under the same dimension.

9. The method according to claim 7, characterized in that, Step S6, which involves determining the operational consistency between the inspection process and the design specifications, includes: Obtain the operator-conserved invariants obtained from the current test operation. Conserved invariants with reference operator , wherein The conserved invariants of the standard side operators are derived from historical processing data, calibration data, or calculated based on the nominal geometric model and GPS standard operating operator chain. Calculate the difference ; like Less than the preset consistency tolerance threshold If so, the inspection process is deemed to have acceptable consistency with the design specifications; like Greater than or equal to the tolerance threshold If so, it is determined that a structural deviation has occurred in the inspection process.

10. The method according to claim 9, characterized in that, The step of performing in-machine accuracy control adjustment as described in step S7 includes: If the operational consistency determination result of step S6 indicates that the inspection operation process has acceptable consistency with the design specifications, then: Based on geometric error Determine the adjustment amount of processing parameters ,in This is the error response matrix established based on the dynamic characteristics of the machine tool; according to Calculate the correction factor The processing parameters are adjusted to obtain the correct amount. ; Input the corrected processing parameter adjustment into the processing equipment, and re-execute the in-machine measurement and inspection operations of steps S1 to S6 based on the processed workpiece; If the consistency determination result of step S6 indicates that a structural deviation has occurred in the inspection operation process, then: If the processing parameters are not adjusted and the adjustment amount is not input into the processing equipment, the process of re-executing the in-machine measurement and inspection operations of steps S1 to S6 will be triggered or an alarm message will be output.

Images