A method and system for intelligently verifying dimensional deviations in the precision machining of steel structural components
By establishing processing benchmark states and cumulative dimensional states, the source of deviations in steel structural components is determined, solving the problem of unclear deviation sources in existing technologies and achieving more accurate dimensional verification and rework positioning.
Patent Information
- Application Number
- CN202611246490.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-17
- Publication Date
- 2026-11-10
AI Technical Summary
Existing technologies struggle to distinguish whether the final dimensional deviation of a target machining feature in a steel structure originates from its own machining deviation, the step-by-step transmission of dimensional relationships, or changes in the machining datum. Furthermore, the lack of corresponding relationships in the retesting methods leads to inaccurate dimensional verification.
By establishing the correspondence between the designed processing features and the measured features, processing reference states and reference alignment feature sets are generated, local deviations and cumulative dimensional states are calculated, the source of deviation is determined, and a retest configuration is generated based on the source of deviation. The status is updated to redetermine the source of deviation and rework positioning.
It enables accurate identification of the sources of deviation in steel structural components, reduces errors caused by measuring target features individually, and improves the accuracy of dimensional verification and the effectiveness of rework positioning.
Smart Images

Figure CN123065403A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of geometric measurement and machining accuracy detection of mechanical components, and in particular to an intelligent verification method and system for precision machining dimensional deviations of steel structural components. Background Technology
[0002] In the field of geometric measurement and machining accuracy inspection of mechanical components, after drilling, milling, end face machining, connection surface machining, or groove machining of steel structural components, it is usually necessary to inspect the dimensions of holes, end faces, connection surfaces, and positioning features according to design dimensions, dimensional tolerances, and machining datums. Existing solutions mostly use coordinate measuring machines, laser measuring machines, or three-dimensional measuring machines to obtain measured features, compare the measured positions and dimensions with the corresponding design positions and dimensions, and identify dimensional deviations or geometric deviations based on the comparison results. In cases involving different measurement coordinates, coordinate transformation relationships are established based on datum features before unified measurement is performed.
[0003] For steel structural components with multiple machining features and dimensional relationships between them, the final dimensional deviation exhibited by the target machining feature does not necessarily originate solely from that feature itself. Small dimensional deviations between adjacent machining features may propagate step-by-step along the dimensional relationships. Furthermore, changes in the position or orientation of the machining datum during measurement can cause multiple machining features to exhibit similar directional shifts. Existing solutions typically focus on a direct comparison between the measured dimensions of the target machining feature and the design dimensions, or determine the error type based on a single measurement result. This makes it difficult to distinguish between similar out-of-tolerance behaviors caused by local machining deviations, cumulative dimensional chain deviations, and datum drift.
[0004] Existing re-measurement methods often involve re-measuring target machining features that have already exceeded tolerances, lacking a direct correlation between the re-measured object and the actual location of the deviation. When the deviation originates from associated machining features in the dimensional correlation path or from changes in the machining datum, simply repeating the measurement of the target machining feature is insufficient to verify the location of the deviation. Subsequent dimensional verification and rework positioning are also easily influenced by the initial deviation assessment results. Therefore, it is necessary to address the issue of how to distinguish the source of deviation and determine the corresponding re-measurement feature when the machining datum state, the deviation of the target machining feature itself, and the dimensional correlation relationship all change, focusing on the dimensional verification of steel structural components. Summary of the Invention
[0005] To address the aforementioned technical problems, this invention provides an intelligent verification method for dimensional deviations in the precision machining of steel structural components, comprising: S100. Obtain the design and machining features, machining datum, dimensional relationships, and measured features of the steel structural components; establish the correspondence between the design and machining features and the measured features; perform coordinate transformation according to the machining datum; and generate the machining datum state and datum alignment feature set based on the differences between the datum features and the machining datum. S200. Calculate the local deviation of the target processing feature based on the reference alignment feature set; determine the size association path according to the size association relationship, take the difference between the actual relative size and the design relative size of adjacent associated processing features as the association deviation and synthesize it along the path to generate the local deviation state and the size accumulation state. S300: Based on the local deviation state, the cumulative size state, and the machining reference state, compare the direction and amplitude relationship between the local deviation and the cumulative size state to determine the source of the deviation; for local machining deviation, cumulative size chain deviation, and reference drift, select the target machining feature, the associated machining feature in the size association path, and the reference feature as features to be retested, and generate a retest configuration. S400. Obtain the retest data of the feature to be retested according to the retest configuration; update the corresponding local deviation status, dimensional cumulative status or processing reference status according to the category of the feature to be retested; redetermine the source of deviation based on the updated status; and generate dimensional verification results and rework positioning information.
[0006] Furthermore, the design and processing features include feature identifiers, geometric types, design locations, design dimensions, and dimensional tolerances; the geometric types of the design and processing features are compared with those of the measured features, and the locations of the measured features are compared with those of the design locations to establish the corresponding relationship.
[0007] Furthermore, the machining datum includes a datum surface, a datum edge, and a positioning hole; a coordinate system transformation relationship is established based on the coordinates of the measurement points of the datum surface, the datum edge, and the positioning hole and the design position, and the measured features are transformed according to the coordinate system transformation relationship.
[0008] Furthermore, the reference surface after coordinate transformation is fitted to generate a fitting residual; the directional deviation of the reference edge and the positional deviation of the positioning hole are calculated, and the fitting residual, directional deviation and positional deviation are compared with the corresponding allowable range of the reference to generate the machining reference state.
[0009] Furthermore, the measured features are generated by fitting data from multiple measurement points; residuals are generated based on the differences between each measurement point data and the measured features, the residuals are compared with a preset threshold to determine the validity of the measurement points, measurement point data whose residuals exceed the preset threshold are identified as abnormal measurement points, and the remaining measurement point data are used to refit the measured features.
[0010] Furthermore, the dimension association path connecting the associated processing feature to the target processing feature is determined according to the design size and dimension association relationship; the difference between the actual relative size and the design relative size of adjacent associated processing features is calculated as the association deviation, and the association deviation is synthesized according to the directional relationship of the dimension association path to generate the dimension cumulative state.
[0011] Furthermore, when the fitting residual, direction deviation, and position deviation are all within the allowable range of the corresponding reference, and the direction of the local deviation is inconsistent with the cumulative size state or the amplitude difference exceeds a preset threshold, the source of the deviation is determined to be a local processing deviation; when the fitting residual, direction deviation, and position deviation are all within the allowable range of the corresponding reference, and the direction of the local deviation is consistent with the cumulative size state and the amplitude difference is within a preset threshold, the source of the deviation is determined to be a cumulative deviation of the size chain; when the fitting residual, direction deviation, or position deviation exceeds the allowable range of the corresponding reference, and the local deviations of multiple processing features are consistent in direction, the source of the deviation is determined to be reference drift.
[0012] Furthermore, when the source of deviation is local machining deviation, the retest configuration includes the target machining feature and the associated machining features adjacent to the target machining feature; when the source of deviation is cumulative deviation of the dimensional chain, the retest configuration includes the associated machining features in the dimensional association path, and the retest order is determined according to the dimensional association path; when the source of deviation is reference drift, the retest configuration includes the reference feature.
[0013] Furthermore, when the retest configuration includes a reference feature, the machining reference is re-established and the machining reference status is updated based on the retest data; when the retest configuration includes associated machining features in the dimension association path, the associated deviation is updated and the dimension accumulation status is regenerated based on the retest data; when the retest configuration includes a target machining feature, the local deviation status is updated based on the retest data; the source of deviation is re-determined based on the updated machining reference status, dimension accumulation status, and local deviation status, and a retest consistency result is generated.
[0014] This invention also provides an intelligent verification system for dimensional deviations in the precision machining of steel structural components, comprising: The feature correspondence module is used to obtain the design and processing features, processing references, dimensional relationships and measured features of steel structural components, and to establish the correspondence between the design and processing features and the measured features; A reference state generation module, connected to the feature correspondence module, is used to perform coordinate transformation on the measured features with corresponding relationships according to the processing reference, and generate a processing reference state and a reference alignment feature set based on the difference between the reference features and the processing reference. The deviation correlation calculation module, connected to the reference state generation module, is used to calculate the local deviation of the target processing feature based on the reference alignment feature set, determine the size correlation path according to the size correlation relationship, take the difference between the actual relative size and the design relative size of adjacent correlated processing features as the correlation deviation and synthesize it along the size correlation path to generate the local deviation state and the size cumulative state. The deviation source determination module is connected to the reference state generation module and the deviation correlation calculation module, respectively. It is used to determine the deviation source by comparing the direction and amplitude relationship between the local deviation and the size cumulative state based on the processing reference state, the local deviation state, and the size cumulative state; and to select the feature to be retested from the target processing feature, the associated processing feature in the size correlation path, and the reference feature according to the deviation source, and to generate the retest configuration. The retesting and verification module is connected to the baseline state generation module, the deviation correlation calculation module, and the deviation source determination module, respectively. It is used to obtain the retesting data of the feature to be retested according to the retesting configuration, update the corresponding processing baseline state, local deviation state, or dimensional cumulative state according to the category of the feature to be retested, redetermine the deviation source based on the updated processing baseline state, local deviation state, and dimensional cumulative state, and generate dimensional verification results and rework positioning information.
[0015] The key innovations of this invention include: (1) Establish a continuous calling relationship between the coordinate unification result of the processing datum and the subsequent deviation verification. Form the processing datum state based on the difference between the datum feature and the processing datum. At the same time, form the local deviation state based on the measured result of the target processing feature. Take the difference between the actual relative size and the design relative size of the adjacent related processing features as the related deviation. Synthesize the size accumulation state along the size association path so that the processing datum, the target processing feature's own state and the size association path can respectively form the measured basis that can participate in the subsequent deviation source determination.
[0016] (2) The local deviation state, the cumulative size state and the processing reference state are combined for the determination of the source of deviation. By comparing the direction and magnitude relationship between the local deviation and the cumulative size state, and combining the processing reference state, the local processing deviation, the cumulative size chain deviation and the reference drift are distinguished. The source of deviation is not used as a separate classification result, but the selection of the target processing feature, the associated processing feature in the size association path and the reference feature are constrained respectively to form a retest configuration corresponding to the location where the deviation is formed.
[0017] (3) Establish the corresponding update relationship between the feature category to be retested and the previous deviation state. The retest data of the target processing feature updates the local deviation state. The retest data of the associated processing feature in the dimension association path updates the associated deviation and re-forms the dimension accumulation state. The retest data of the reference feature is used to re-establish the processing reference and update the processing reference state. The updated state participates in the deviation source determination again, so that the retest result returns to the deviation judgment link corresponding to its physical measurement object.
[0018] The following are its main beneficial effects: (1) In view of the situation that the final deviation of the target machining feature in the steel structure may be affected by its own machining deviation, the step-by-step transmission of related dimensions and the change of machining datum, by forming machining datum state, local deviation state and dimension cumulative state respectively, the measured deviation of the target machining feature can be established to correspond with the actual deviation transmission in the datum change and dimension association path, reducing the situation that the deviation formation location is not clear when checking only the measured value of a single target machining feature.
[0019] (2) To address the issue that local processing deviations, cumulative deviations in the dimensional chain, and reference drift may manifest as similar dimensional deviations, the source is determined by combining three states, and different deviation sources are respectively assigned to target processing features, associated processing features, and reference features. This allows the determination of the deviation source to directly form a specific retest object, reducing the situation where a single target processing feature is repeatedly measured when the actual source of the deviation is located in the dimensional association path or processing reference.
[0020] (3) To address the problem of insufficient connection between the retest results and the previous deviation judgment basis, the local deviation status, dimensional cumulative status or processing reference status are updated according to the category of the feature to be retested, and the source of deviation is re-determined using the updated status, so that the retest data acts on the deviation evidence corresponding to its actual measurement object; when the relationship between the initial deviation source and the status after retest is inconsistent, the dimensional verification results and rework positioning information are formed based on the updated deviation source, reducing the situation where the initial judgment result directly fixes the subsequent rework position. Attached Figure Description
[0021] Figure 1 A flowchart illustrating an intelligent verification method for precision machining dimensional deviations of steel structural components, provided in an embodiment of this application; Figure 2 The structural block diagram of an intelligent verification system for precision machining dimensional deviations of steel structural components provided in this application embodiment is shown. Detailed Implementation
[0022] Example 1: Refer to Figure 1This is a flowchart illustrating an intelligent verification method for precision machining dimensional deviations of steel structural components provided in an embodiment of the present invention. The process may include at least steps S100-S400: S100. Obtain the design and machining features, machining datum, dimensional relationships, and measured features of the steel structural components; establish the correspondence between the design and machining features and the measured features; perform coordinate transformation according to the machining datum; and generate the machining datum state and datum alignment feature set based on the differences between the datum features and the machining datum. S200. Calculate the local deviation of the target processing feature based on the reference alignment feature set; determine the size association path according to the size association relationship, take the difference between the actual relative size and the design relative size of adjacent associated processing features as the association deviation and synthesize it along the path to generate the local deviation state and the size accumulation state. S300: Based on the local deviation state, the cumulative size state, and the machining reference state, compare the direction and amplitude relationship between the local deviation and the cumulative size state to determine the source of the deviation; for local machining deviation, cumulative size chain deviation, and reference drift, select the target machining feature, the associated machining feature in the size association path, and the reference feature as features to be retested, and generate a retest configuration. S400. Obtain the retest data of the feature to be retested according to the retest configuration; update the corresponding local deviation status, dimensional cumulative status or processing reference status according to the category of the feature to be retested; redetermine the source of deviation based on the updated status; and generate dimensional verification results and rework positioning information.
[0023] S100. Obtain the design and machining features, machining datum, dimensional relationships, and measured features of the steel structural components; establish the correspondence between the design and machining features and the measured features; perform coordinate transformation according to the machining datum; and generate the machining datum state and datum alignment feature set based on the differences between the datum features and the machining datum. After the steel structural component completes the predetermined finishing process and enters the dimensional verification stage, the feature corresponding unit reads the design processing features, processing references and dimensional relationships corresponding to the steel structural component, and receives the measurement point data generated by the dimensional measuring equipment. The design processing features are organized according to the design dimensions of the same steel structural component. Each design processing feature has a feature identifier, geometric type, design position, design dimension and dimensional tolerance. The dimensional relationship represents the relative positional constraints between different design processing features formed by the design dimensions, so that subsequent processing retains both the design requirements of the individual processing feature itself and the dimensional transfer relationship between different processing features.
[0024] The measured features are formed by fitting multiple measurement point data within the corresponding processing area. The measurement point data are collected according to the actual measured object, and measurement points located on different processing surfaces are not mixed to form the same measured feature. For hole-type measured features, the position and size of the hole are formed based on the measurement points on the hole wall. For end face or connecting surface-type measured features, the planar position is formed based on the coordinates of the measurement points on the corresponding surface. For measured features corresponding to the reference edge, the direction and position are formed based on the measurement points obtained along the reference edge. Subsequent feature correspondence and reference establishment all call the above fitting results.
[0025] When fitting the measured features, the residuals are first calculated based on the differences between the data of each measuring point and the current fitting result, and then compared with the preset threshold configured for this measurement task. Measuring points whose residuals do not exceed the preset threshold continue to participate in the fitting of the measured features. Measuring points whose residuals exceed the preset threshold are identified as abnormal measuring points and removed from the current fitting data. The remaining measuring point data are then used to refit the measured features. When the remaining measuring point data does not meet the fitting conditions of the corresponding geometric type, the current fitting result is not used to enter the subsequent size calculation, and the measuring point data of the processing area is re-acquired.
[0026] When establishing a correspondence between designed processing features and measured features, the feature correspondence unit first compares the geometric types, forming candidate objects from measured features with the same geometric type. Then, it compares the position of the candidate measured features with the design position of the designed processing features. For multiple holes or multiple processing surfaces with repetitive geometric types, the correspondence is not only based on a single position value, but also checked in conjunction with the positional order between the feature and the reference feature and adjacent designed processing features. After both the geometric type and position match, the correspondence is established. Measured features that do not obtain a unique correspondence result are not included in the reference alignment feature set, but their measurement point data are retained and the feature correspondence is re-executed or the measurement is re-measured.
[0027] After completing the correspondence between the designed machining features and the measured features, the reference state generation unit reads the reference features corresponding to the machining reference from the measured features. The machining reference consists of a reference surface, a reference edge, and a positioning hole. In one embodiment, the reference plane of the steel structure is determined by the measured plane corresponding to the reference surface, the coordinate axis direction is constrained by the measured direction corresponding to the reference edge, and the coordinate position is constrained by the measured position corresponding to the positioning hole. The coordinate system transformation relationship from the measurement coordinate system to the machining reference coordinate system is established based on the coordinates of the measurement points of the three types of reference features and their respective design positions.
[0028] After the coordinate system transformation relationship is established, the corresponding measured features that have been completed are transformed from the original measurement coordinate system to the machining reference coordinate system, so that the target machining features, associated machining features and reference features are under the same coordinate relationship. For steel structural parts that are measured in different regions, the measuring points in each region are first unified according to their respective coordinate system transformation relationship with the machining reference. The original coordinates in different measurement regions are not directly compared to prevent changes in the placement of the measuring equipment or the clamping posture of the steel structural parts from directly entering the subsequent associated deviation calculation.
[0029] After the coordinate transformation is completed, the transformed reference surface is refitted and the fitting residual is obtained. At the same time, the directional deviation between the measured direction and the design direction of the reference edge, and the positional deviation between the measured position and the design position of the positioning hole are calculated. The fitting residual, directional deviation and positional deviation are compared with the corresponding reference allowable range. When all three are within the corresponding reference allowable range, the machining reference state indicates that the current reference relationship is consistent with the design machining reference. When any one of them exceeds the corresponding reference allowable range, the machining reference state retains the reference features and their deviations that exceed the range, so that S300 can distinguish between the local deviation of the target machining feature and the reference drift.
[0030] In one implementation, the machining reference state does not replace the fitting residual, direction deviation, and position deviation with a single total deviation value, but retains the comparison results corresponding to the three. This is because the fitting anomaly of the reference surface, the direction change of the reference edge, and the position change of the positioning hole have different effects on the subsequent measured feature coordinates. When S300 needs to check whether multiple machining features have a common directional offset, the reference deviation in the reference state corresponding to that direction directly participates in the deviation source determination.
[0031] When the coordinate system transformation relationship cannot be stably established due to the abnormality of the reference feature measurement points, the reference state generation unit does not use the transformation result formed by the abnormal reference to process the target machining feature; the system re-acquires the measurement point data of the corresponding reference feature, and refits the reference surface, reference edge or positioning hole according to the measurement point validity rule. After re-acquiring the reference feature that meets the allowable range of the reference, the coordinate system transformation relationship is re-established. The original measurement point data keeps the original measurement coordinates unchanged before the reference is re-established.
[0032] The reference alignment feature set formed by the above processing preserves the correspondence between the designed processing features and the measured features, as well as the position and size of the measured features in the processing reference coordinate system; the processing reference state also retains the fitting residual, direction deviation and position deviation on which the coordinate unification is based. S200 reads the reference alignment feature set to calculate the local deviation and cumulative size status of the target processing feature, and S300 continues to read the processing reference state to participate in the deviation source determination.
[0033] S200. Calculate the local deviation of the target processing feature based on the reference alignment feature set; determine the size association path according to the size association relationship, and use the difference between the actual relative size and the design relative size of adjacent associated processing features as the association deviation and synthesize them along the path to generate a local deviation state and a size accumulation state: The deviation correlation calculation unit receives the reference alignment feature set formed by S100 and reads the size correlation relationship corresponding to the current verification task. The target machining feature in this step is the machining feature that needs to be checked for size deviation. The measured position and measured size of the target machining feature are both the result of S100 coordinate transformation, and the values in the original measurement coordinate system are no longer used, so that the deviation of the target machining feature itself and the correlation deviation in the subsequent size correlation path have the same coordinate reference.
[0034] For the target machining feature, first read its corresponding design position, design size and dimensional tolerance. Then compare the measured position or measured size of the target machining feature with the corresponding design value to obtain the local deviation. When the design size has a clear direction, the local deviation retains the positive and negative relationship and magnitude in that direction. When the design machining feature is described by positional relationship, the positional difference is obtained according to the direction of the corresponding design size, so that when S300 compares the local deviation with the cumulative size state, it uses the same direction definition instead of just comparing the absolute values of the two deviations.
[0035] The local deviation status is formed by comparing the local deviation with the corresponding dimensional tolerance, and retains the direction and magnitude of the local deviation. When the local deviation is within the dimensional tolerance range, the current target machining feature does not directly cause out-of-tolerance due to its own size. When the local deviation exceeds the dimensional tolerance range, the formation position is checked by combining the cumulative dimensional status and the machining reference status. Therefore, the local deviation status is not directly equivalent to the final deviation source determination result.
[0036] Subsequently, the dimension association path corresponding to the target machining feature is determined according to the dimension association relationship. The dimension association path is formed by connecting associated machining features with continuous design dimension relationships. For example, in the same steel structure, machining feature A corresponding to the machining reference has a first design dimension with hole B, hole B has a second design dimension with hole C, and hole C has a third design dimension with end face D. When end face D is the target machining feature, A, B, C, and D form a dimension association path corresponding to D according to the above design dimension relationship. The path direction is unified from the reference machining feature to the target machining feature according to each design dimension.
[0037] For adjacent associated processing features in the dimension association path, the deviation association calculation unit obtains the actual relative size of the two based on the reference alignment feature set, and reads the design relative size between the same two design processing features; the difference between the actual relative size and the design relative size is used as the association deviation of the associated segment. The direction of the association deviation is saved according to the direction of the dimension association path. When the actual relative size is greater than the design relative size along the path direction, it is identified differently from the association deviation that decreases along the opposite direction, so that multiple small deviations can be distinguished during subsequent synthesis when they accumulate in the same direction and cancel each other out.
[0038] In one implementation, when the positional relationship between associated processing features is located in different spatial directions, the directional components of each associated deviation are first obtained according to the direction of the corresponding design dimension, and then synthesized according to the directional relationship of the dimension association path. For dimension relationships in the same straight line direction, the correlation deviation is synthesized according to the positive and negative directions. For dimension association paths with turning relationships, the dimensions corresponding to the target processing feature are processed separately according to the direction of each segment of the design dimension to form the dimension accumulation state. The deviation amplitudes in different spatial directions are not directly added without direction.
[0039] The cumulative size status reflects the cumulative change of the actual relative size relative to the design relative size from the reference processing feature to the target processing feature in the size association path. This status retains the direction and magnitude of the cumulative deviation after synthesis, while maintaining its corresponding size association path, so that when the local deviation has a corresponding relationship with the direction and magnitude of the cumulative size status, S300 can still obtain the associated processing features that need to be retested from the path.
[0040] When a target processing feature has multiple dimension-related paths formed by different design dimension relationships, the paths are not mixed and synthesized without distinction. The deviation correlation calculation unit selects the path corresponding to the current design dimension to be checked and calculates the dimension accumulation state separately. When the calculation results of different paths conflict in direction, the corresponding path and its associated deviation are retained. In S300, the final deviation source is not directly formed based on a single path, but the associated processing feature that caused the conflict is selected by the retest configuration for remeasurement.
[0041] In the aforementioned implementation scenarios A, B, C, and D, if the actual relative dimensions of the three associated segments A to B, B to C, and C to D are all offset relative to the corresponding design relative dimensions in the direction from A to D, then the three associated deviations are combined along the dimension association path to form a dimension accumulation state in the same direction. If the local deviation of the target processing feature D also shows the same direction, and the magnitude relationship between the two is within the preset threshold range adopted in subsequent S300, then this state combination enters the determination process of dimension chain accumulation deviation. If the direction of the local deviation of D is opposite to the direction of the dimension accumulation state, then the local deviation of D cannot be directly explained by the dimension association path.
[0042] When the measured features of a certain associated processing feature are missing or the validity of its measurement points does not meet the fitting conditions of S100, the associated deviation connected to the processing feature does not participate in path synthesis; the deviation association calculation unit retains the currently obtained local deviation state and effective association segment, and treats the missing processing feature as the object to be reacquired measurement point data. After the measured feature is re-formed, the corresponding association deviation calculation is restored to avoid using the cumulative results of incomplete paths to enter S300.
[0043] This step ultimately forms a local deviation state and a cumulative dimension state. The local deviation state corresponds to the measured change of the target machining feature itself, and the cumulative dimension state corresponds to the deviation transmission formed by multiple related machining features along the design dimension relationship. S300 reads the above two states and the machining reference state retained by S100, compares the correspondence between the three to determine the source of the deviation, and changes the subsequent retest object accordingly.
[0044] S300. Based on the local deviation state, the cumulative dimension state, and the machining reference state, compare the direction and amplitude relationship between the local deviation and the cumulative dimension state to determine the source of the deviation; for local machining deviation, cumulative dimension chain deviation, and reference drift, select the target machining feature, the associated machining feature in the dimension association path, and the reference feature as features to be retested, and generate a retest configuration: The deviation source determination unit receives the local deviation state and dimensional cumulative state formed by S200, and reads the machining reference state formed by S100. The deviation source determination does not classify the deviation solely according to whether the target machining feature exceeds the dimensional tolerance. Instead, it first checks the machining reference state, and then compares the direction and amplitude relationship between the local deviation and the dimensional cumulative state, so that the changes in the target machining feature itself, the transmission of the dimensional correlation path, and the changes in the machining reference correspond to different measurement evidence.
[0045] When the fitting residual of the datum plane, the directional deviation of the datum edge, and the positional deviation of the positioning hole are all within the allowable range of the corresponding datum, the current machining datum serves as the common measurement basis for local deviation and cumulative dimensional state. Under this condition, if the local deviation of the target machining feature exceeds its dimensional tolerance, and the direction of the local deviation is inconsistent with the cumulative dimensional state, or the difference in their magnitudes exceeds a preset threshold, then the cumulative change formed by the dimensional correlation path cannot correspond to the current out-of-tolerance of the target machining feature, and the deviation source determination result is recorded as local machining deviation.
[0046] For local machining deviations, the remeasurement configuration selects the target machining feature and the associated machining features adjacent to the target machining feature; the target machining feature remeasures its own size and position, and the adjacent associated machining features remeasures the position that directly forms the design size relationship with the target machining feature. This remeasurement configuration enables S400 to check whether the anomaly of the target machining feature is maintained after repeated measurement, and at the same time check whether the size relationship directly connected to it is still in the original state, without reacquiring all measurement point data along the entire size association path.
[0047] When the processing reference state is within the allowable range of the corresponding reference, if the local deviation is in the same direction as the cumulative dimension state and the difference in their amplitudes is within a preset threshold, then the deviation of the target processing feature corresponds to the result of the synthesis of multiple associated deviations in the dimension association path, and the deviation source determination result is recorded as the cumulative deviation of the dimension chain. The preset threshold is configured in the measurement task according to the current design dimension and dimensional tolerance, and remains the same in the initial verification and re-verification to avoid the deviation source changing due to the use of different amplitude comparison conditions in the two rounds.
[0048] When the deviation originates from the cumulative deviation of the dimension chain, the retest configuration is not limited to the target processing feature, but selects associated processing features from the dimension association path and forms a retest sequence according to the dimension association path. In one embodiment, the retest sequence proceeds sequentially from the reference processing feature to the target processing feature, so that the corresponding associated deviation can be updated after the actual relative size of the adjacent processing feature is obtained again each time. When the deviation direction changes after a certain associated segment is retested, the subsequent S400 recalculates the cumulative dimension state according to the updated path.
[0049] When the fitting residual of the datum surface, the directional deviation of the datum edge, or the positional deviation of the positioning hole exceeds the allowable range of the corresponding datum, the machining datum status indicates that the current coordinate unification relationship itself has a part that is inconsistent with the design machining datum. At this time, the deviation source determination unit reads the local deviation of multiple machining features in the datum alignment feature set. If multiple machining features show the same deviation direction, then the common offset is related to the abnormal machining datum status, and the deviation source determination result is recorded as datum drift.
[0050] For datum drift, the retest configuration selects the datum features corresponding to the machining datum, that is, remeasures the datum surface, datum edge and positioning hole, instead of directly treating all the target machining features that exceed the tolerance as local rework objects; after the S400 re-obtains the datum feature measurement points, it first reconstructs the machining datum. If the positional relationship of the original measured features under the new machining datum changes synchronously, the deviation source determination is continued with the updated machining datum state, avoiding the initial coordinate datum anomaly being directly fixed as a local machining deviation of multiple independent machining features.
[0051] In one implementation, if the fitting residual, direction deviation, and position deviation exceed the allowable range of the reference, but the local deviations of multiple processing features do not show a consistent direction, the source of the deviation is not immediately fixed as reference drift. The retest configuration retains both the reference features that exceed the allowable range of the reference and the current target processing features. S400 first reacquires the reference feature measurement points and updates the processing reference state, and then recalculates the local deviation of the target processing features based on the updated coordinate relationship. This branch is used to handle the situation where local measurement anomalies of the reference features and the actual processing deviation of the target processing features coexist.
[0052] If the local deviation is in the same direction as the size accumulation state, but the amplitude difference is outside the preset threshold boundary, or different size association paths form conflicting size accumulation states for the same target processing feature, the current deviation source determination result is not used as the final size verification result. In one embodiment, the retest configuration selects the associated processing features that generate amplitude differences or path conflicts, so that S400 re-forms the corresponding associated deviation and compares again, instead of forcibly obtaining a certain deviation source by expanding the error category or changing the preset threshold.
[0053] The resulting retest configuration corresponds to the deviation source determination result. Local processing deviation corresponds to the target processing feature and adjacent related processing features, cumulative deviation of the dimension chain corresponds to the related processing features in the dimension association path, and reference drift corresponds to the reference feature. The retest configuration is sent to the retest verification unit. The retest data subsequently acquired updates the local deviation status, cumulative dimension status or processing reference status according to the category of the feature to be retested, instead of using the same retest data processing method to cover all the initial verification results.
[0054] S400. Obtain the retest data of the feature to be retested according to the retest configuration; update the corresponding local deviation state, dimensional cumulative state, or processing reference state according to the category of the feature to be retested; redetermine the source of deviation based on the updated state; and generate dimensional verification results and rework positioning information. After receiving the retest configuration formed by S300, the retest verification unit calls the corresponding measurement task according to the feature to be retested. The dimensional measuring device reacquires the measurement point data of the feature to be retested and forms retest data according to the measurement point validity rules in S100. The retest data retains the design and processing feature identifier and geometric type in the initial measurement, so that the initial measured feature and the retest measured feature maintain the same correspondence. When the measurement point residual exceeds the preset threshold, abnormal measurement points are first excluded and refitted. The original state is not updated using the retest data that has not completed the validity check.
[0055] When the feature to be retested in the retest configuration is the target machining feature, the retest verification unit reads the new measured position or measured size of the target machining feature and recalculates its local deviation relative to the designed machining feature in the current machining reference coordinate system. The new local deviation replaces the original local deviation in the local deviation state calculation, while the dimension association path and machining reference state remain at their current values if they are not remeasured in this round. Then, the direction and amplitude comparison in S300 are re-executed.
[0056] For the retesting configuration of local machining deviations, the associated machining features adjacent to the target machining feature are measured simultaneously. If the local deviation of the target machining feature after retesting still exceeds the dimensional tolerance, and the actual relative dimensions re-formed by the adjacent associated machining features do not change the dimensional cumulative state to a state that can explain the target local deviation, then the re-determined deviation source remains the local machining deviation, and the rework positioning information corresponds to the target machining feature. If the associated deviation of the adjacent associated machining features changes after retesting and forms a dimensional cumulative state corresponding to the target local deviation, then the determination of the dimensional chain cumulative deviation is re-executed, and the initial deviation source is not used.
[0057] When the feature to be retested in the retest configuration belongs to the associated processing feature in the dimension association path, the retest verification unit updates the corresponding measured feature item by item according to the retest order, and recalculates the actual relative size of adjacent associated processing features; the new actual relative size is compared with the same design relative size to obtain the new association deviation, and then the dimension cumulative state is resynthesized according to the directional relationship of the original dimension association path. Associated processing features not included in this retest configuration and whose measurement point validity has not changed continue to use the existing measured results.
[0058] During the dimensional chain cumulative deviation retesting process, if the original associated deviation changes significantly after the retesting of a certain associated processing feature, the new associated deviation directly participates in the current dimensional cumulative state, without using the average of the initial associated deviation and the retested associated deviation as a substitute; when the updated dimensional cumulative state and the local deviation of the target processing feature still maintain the same direction and the amplitude difference is within the preset threshold, the retest consistency result records that the source of the initial deviation is consistent with the source of the deviation after the retest, and the rework positioning information corresponds to the associated processing feature that still retains the associated deviation after the retest and its position in the dimensional association path.
[0059] If, after re-measuring along the dimension association path, the updated cumulative dimension status can no longer form a direction and magnitude correspondence with the local deviation of the target machining feature, then the initial cumulative dimension chain deviation is not considered as the final source of deviation. The re-measurement and verification unit uses the updated cumulative dimension status and the machining reference status to re-execute S300. If the machining reference status is normal and the target local deviation still exists, it switches to the re-measurement configuration corresponding to the local machining deviation. If the machining reference status is also abnormal, the reference feature is re-measured according to the re-measurement configuration corresponding to the reference drift.
[0060] When the retest configuration includes datum features, the retest verification unit reacquires the measurement point data of the datum surface, datum edge, and positioning hole, and recalculates the fitting residual of the datum surface, the directional deviation of the datum edge, and the positional deviation of the positioning hole; based on the new datum feature measurement point coordinates and the design position, the coordinate system transformation relationship is re-established, and the machining datum state is updated according to the new coordinate system transformation relationship, so that the subsequent local deviations and dimensional cumulative states are compared based on the retested machining datum.
[0061] In the remeasurement of the reference drift, the target machining feature and related machining feature measurement point data formed in the initial measurement are retained as long as the measurement point validity is still valid, and are reconverted to the updated machining reference coordinate system according to the newly established coordinate system transformation relationship; thereby, the local deviation of each target machining feature and the actual relative size in the dimension association path are recalculated, so that the on-site measurement of all machining features is not mechanically repeated due to the remeasurement of the reference feature. When the existing measurement point data after reconversion cannot correspond to the new machining reference relationship, the measurement point data of the corresponding machining feature is reacquired.
[0062] For example, if the local deviations of multiple holes and end faces in the initial verification are all in the same direction, and the positional deviation of the positioning hole exceeds the allowable range of the corresponding datum, then S300 will include it in the datum drift re-measurement configuration; after re-measuring the positioning hole, datum surface and datum edge, if the new machining datum state is within the datum allowable range, and the local deviations of the original multiple holes and end faces after recalculation through the new coordinate system transformation relationship return to the corresponding dimensional tolerance range, then the re-measurement consistency result records the initial datum drift judgment and the datum change after the re-measurement, and the dimensional verification result adopts the calculation result after re-establishing the machining datum, instead of recording the original multiple holes and end faces as local machining deviations respectively.
[0063] If the target machining feature still maintains a local deviation that exceeds the dimensional tolerance after the datum is re-established, the retesting and verification unit will not maintain the original source of deviation because it was initially determined to be a datum drift; the system will use the updated machining datum state to reread the local deviation state and dimensional cumulative state of the target machining feature, and re-distinguish between local machining deviation and dimensional chain cumulative deviation according to the rules of S300. The new source of deviation will serve as the basis for the current dimensional verification result.
[0064] The consistency results of the retest are compared with the initial source of deviation and the source of deviation re-determined using the retest data. At the same time, it is checked whether the processing reference state, local deviation state and dimensional cumulative state supporting the source of deviation still meet the corresponding judgment conditions. When the initial and retested sources of deviation are consistent, the current source of deviation is included in the dimensional verification results. When they are inconsistent, the deviation source judgment is re-executed with the three states updated after the retest, and the next round of retest is carried out according to the new retest configuration. The rework positioning information is directly generated without using the initial source of deviation.
[0065] The dimensional verification results record the local deviation status, final deviation source, and retest consistency results of the target machining feature after retesting. The rework location information points to the actual machining feature that needs to be processed according to the final deviation source. When the final deviation source is a local machining deviation, the rework location information corresponds to the target machining feature. When the final deviation source is a cumulative deviation in the dimensional chain, the rework location information corresponds to the associated machining feature in the dimensional association path that still has associated deviations after retesting. When the initial manifestation is a datum drift and the target machining feature no longer exceeds the tolerance after re-establishing the machining datum, the original target machining feature is not used as the rework location object.
[0066] In the system implementation, the feature correspondence unit completes the correspondence between the designed processing features and the measured features in S100, the reference state generation unit completes the coordinate transformation and outputs the processing reference state and the reference alignment feature set, the deviation correlation calculation unit receives the reference alignment feature set to form the local deviation state and the dimension accumulation state, the deviation source determination unit receives the above two states and the processing reference state to form the retest configuration, the retest verification unit obtains the retest data and updates the corresponding state according to the category of the feature to be retested, and the updated state re-enters the deviation source determination unit, thereby maintaining the same data call relationship between the method steps and the system units.
[0067] When a steel structural component enters a new round of dimensional verification, the processing datum state and measured features that have been re-measured and confirmed are not directly reused as the measurement results of the new component. The new round of verification re-obtains the measured features and datum feature measurement point data corresponding to the current steel structural component. The dimensional correlation and design processing features remain consistent with the current design processing data. The datum alignment feature set for this round is re-established in S100, and then the dimensional deviation verification for this round is completed in S200, S300 and S400 in sequence.
[0068] Example 2: Figure 2 A structural block diagram of an intelligent verification system for precision machining dimensional deviations of steel structural components according to an embodiment of the present invention is shown. Figure 2 As shown, the structure may include: Feature correspondence module 01 is used to acquire the design and processing features, processing datum, dimensional relationships, and measured features of the steel structural components, and to establish the correspondence between the design and processing features and the measured features. The feature correspondence module accesses the design and processing data and dimensional measurement results of the steel structural components, whereby the design and processing data includes design and processing features, processing datum, and dimensional relationships, and the dimensional measurement results include measured features formed by fitting data from multiple measurement points. The design and processing features are organized according to feature identifier, geometric type, design position, design size, and dimensional tolerance, while the measured features retain the corresponding geometric type, measured position, and measured size. The feature correspondence module first filters candidate measured features according to geometric type, then compares the position of the candidate measured features with the design position, and establishes a correspondence when both the geometric type and positional relationship match. For multiple machining features with the same geometric type in the same steel structural component, the feature matching module continues to check the positional order between each candidate measured feature and the machining datum and adjacent machining features. When there are multiple candidate objects and a unique correspondence cannot be formed, the current measured feature is not passed to the datum state generation module. Instead, the corresponding measurement point data is retained and feature matching is re-executed. After the correspondence is established, the feature matching module provides the datum state generation module with the measured feature with the design machining feature identifier, the machining datum, and the dimensional association relationship. This ensures that subsequent coordinate transformations, local deviation calculations, and dimensional association path calculations all reference the same design machining feature and measured feature correspondence relationship.
[0069] The reference state generation module 02, connected to the feature correspondence module, is used to perform coordinate transformation on the corresponding measured features according to the processing reference, and generate processing reference state and reference alignment feature set based on the difference between the reference features and the processing reference. After obtaining the corresponding measured features from the feature correspondence module, the reference state generation module reads the reference surface, reference edge and positioning hole corresponding to the processing reference. The coordinates of the measurement points of the reference surface participate in the reference plane fitting, the coordinates of the measurement points of the reference edge participate in the direction relationship calculation, and the coordinates of the measurement points of the positioning hole participate in the position relationship calculation. Based on the measured position and the design position of the above reference features, a coordinate system transformation relationship is established. Then, according to the coordinate system transformation relationship, the target processing feature, associated processing feature and reference feature are uniformly transformed to the processing reference coordinate system. The transformed measured features maintain the correspondence with the design processing features according to the original feature identifier and form a reference alignment feature set.
[0070] After completing the coordinate transformation, the datum state generation module refits the datum surface and obtains the fitting residual. It calculates the directional deviation between the measured direction of the datum edge and the design direction, and simultaneously calculates the positional deviation between the measured position and the design position of the positioning hole. The fitting residual, directional deviation, and positional deviation are compared with the corresponding allowable range of the datum. The comparison results are retained as the machining datum state, so that the machining datum state reflects both the current coordinate transformation basis and the offset of the datum feature itself relative to the design machining datum. When there are insufficient effective measurement points on the datum surface, the datum edge direction cannot form a stable fitting result, or the positioning hole position data is missing, the datum state generation module pauses the output of the current datum alignment result and rereads the measurement point data of the corresponding datum feature. After obtaining the effective datum feature again, the coordinate system transformation relationship is updated. The datum alignment feature set is passed to the deviation association calculation module, and the machining datum state is passed to the deviation source determination module and the re-measurement verification module, respectively.
[0071] The deviation correlation calculation module 03, connected to the reference state generation module, is used to calculate the local deviation of the target processing feature based on the reference alignment feature set, determine the size correlation path according to the size correlation relationship, and use the difference between the actual relative size and the design relative size of adjacent related processing features as the correlation deviation and synthesize it along the size correlation path to generate a local deviation state and a size cumulative state. The deviation correlation calculation module reads the measured position or measured size corresponding to the target processing feature in the reference alignment feature set and compares it with the design position or design size of the corresponding design processing feature to form a local deviation with direction and magnitude. The local deviation is then compared with the corresponding size tolerance to form a local deviation state. The local deviation state retains the deviation direction, deviation magnitude and its comparison relationship with the size tolerance of the target processing feature itself, and is not directly used as the final deviation source.
[0072] For the same target processing feature, the deviation correlation calculation module reads the design dimensions connected to the target processing feature from the dimensional correlation relationship, and establishes a dimensional correlation path according to the continuous connection relationship between the design dimensions. For two adjacent related processing features in the path, the corresponding measured positions are obtained from the benchmark alignment feature set, and the actual relative dimensions are calculated according to the spatial relationship between the two. Then, the design relative dimensions between the two identical design processing features are read, and the difference between the actual relative dimensions and the design relative dimensions is taken as the correlation deviation of the correlation segment. Multiple correlation deviations are synthesized according to the directional relationship specified by the dimensional correlation path. Correlation deviations in the same direction participate in the accumulation according to the same direction, and correlation deviations in opposite directions participate in the synthesis according to the corresponding directional relationship, thereby forming a dimensional accumulation state corresponding to the target processing feature. When there are multiple dimensional correlation paths, each path retains its corresponding correlation deviation and accumulation result, and the data from different paths are not directly mixed.
[0073] When a certain associated processing feature in the dimension association path lacks a valid measured feature, the associated deviation connected to that processing feature will not be included in the current path synthesis. The deviation association calculation module retains the remaining valid association segments and provides the missing processing feature identifier to the preceding feature corresponding module. After the corresponding measured feature is reconstructed from the dimension measurement results, the path calculation is restored. After the local deviation state and the dimension cumulative state are formed, they are jointly transmitted to the deviation source determination module. The dimension cumulative state also retains the corresponding dimension association path, so that the deviation source determination module can directly obtain the associated processing feature corresponding to the current cumulative result when it needs to form a retest configuration.
[0074] The deviation source determination module 04 is connected to the reference state generation module and the deviation association calculation module, respectively. It is used to determine the deviation source by comparing the direction and amplitude relationship between the local deviation and the cumulative size state based on the processing reference state, the local deviation state, and the cumulative size state. According to the deviation source, it selects the feature to be retested from the target processing feature, the associated processing feature in the size association path, and the reference feature, and generates the retest configuration. The deviation source determination module first reads the comparison results of the fitting residual, direction deviation, and position deviation in the processing reference state with the corresponding reference allowable range. If the processing reference state does not exceed the reference allowable range, it then compares the direction relationship and amplitude difference between the local deviation of the target processing feature and the cumulative size state. When the two directions are inconsistent, or the directions are consistent but the amplitude difference exceeds the preset threshold, the current deviation source is recorded as a local processing deviation. When the two directions are consistent and the amplitude difference is within the preset threshold, the current deviation source is recorded as a cumulative size chain deviation.
[0075] When the fitting residual, direction deviation, or position deviation in the processing reference state exceeds the allowable range of the corresponding reference, the deviation source determination module continues to read the local deviation direction of multiple processing features in the reference alignment feature set. When multiple processing features show a consistent directional offset, the deviation source is recorded as reference drift. If the local deviation directions of multiple processing features do not have a consistent relationship, the reference drift result is not used to replace the original local deviation state. Instead, the reference features that exceed the allowable range of the reference and the current target processing features are retained in the retest configuration. The retest verification module obtains the corresponding retest data again and performs the deviation source determination again.
[0076] Once the source of deviation is identified, the deviation source determination module organizes the retest configuration according to the correspondence between the deviation source and the feature to be retested. Local processing deviation corresponds to the target processing feature and the associated processing features that directly form a dimensional relationship with the target processing feature. Cumulative deviation in the dimensional chain corresponds to the associated processing features in the current dimensional association path, and the retest objects are arranged according to the connection order of the dimensional association path. Reference drift corresponds to the reference feature in the processing reference. The retest configuration transmits the feature to be retested and its category to the retest verification module, which then determines whether the subsequent retest data enters the update process of local deviation state, dimensional accumulation state, or processing reference state.
[0077] The retesting and verification module 05 is connected to the baseline state generation module, the deviation correlation calculation module, and the deviation source determination module, respectively. It is used to acquire retesting data of the feature to be retested according to the retesting configuration, update the corresponding processing baseline state, local deviation state, or dimensional cumulative state according to the category of the feature to be retested, redetermine the deviation source based on the updated processing baseline state, local deviation state, and dimensional cumulative state, and generate dimensional verification results and rework positioning information. After receiving the retesting configuration output by the deviation source determination module, the retesting and verification module matches the feature to be retested with the original design processing feature identifier and reads the retesting data regenerated by the dimensional measurement equipment. The measurement points in the retesting data are first checked according to the same measurement point validity conditions as the initial measurement. Abnormal measurement points do not participate in the refitting of the corresponding measured features. When the remaining measurement points cannot form valid measured features, the current retesting branch is maintained, and incomplete retesting results are not used to replace existing states.
[0078] When the feature to be retested is a target machining feature, the retesting and verification module recalculates its local deviation relative to the designed machining feature according to the current machining reference coordinate system, and updates the local deviation status with the new local deviation. When the feature to be retested belongs to the associated machining feature in the dimension association path, the actual relative size and associated deviation between adjacent machining features are recalculated based on the updated measured feature, and then the deviation association calculation module re-forms the dimension accumulation status according to the original dimension association path. When the feature to be retested belongs to the reference feature, the retesting and verification module provides the new reference feature measurement points to the reference status generation module, re-establishes the coordinate system transformation relationship and updates the machining reference status. Existing measured features, if the validity of the measurement points has not changed, re-form the reference alignment result according to the new coordinate system transformation relationship.
[0079] The updated machining datum state, local deviation state, and cumulative dimensional state are sent back to the deviation source determination module. The deviation source is re-determined according to the same datum allowable range, directional relationship, and amplitude comparison conditions as the initial determination. When the re-determined deviation source is consistent with the initial deviation source, the re-measurement and verification module generates a dimensional verification result based on the current state. If the re-determined deviation source changes, the re-measurement configuration is re-formed with the updated deviation source, and the original deviation source is not directly included in the rework positioning information. When the final deviation source is a local machining deviation, the rework positioning information corresponds to the target machining feature that still has a local deviation after re-measurement. When the final deviation source is a cumulative dimensional chain deviation, the rework positioning information corresponds to the associated machining feature in the dimensional association path that still retains the associated deviation after re-measurement. When the original target machining feature no longer exceeds the dimensional tolerance after the datum feature is re-measured and the machining datum is re-established, the original target machining feature is not used as the rework positioning object. This forms a feedback connection relationship from feature correspondence, datum state generation, deviation association calculation, deviation source determination to re-measurement and verification, and then back to the aforementioned state calculation module.
Claims
1. A method for intelligently verifying dimensional deviations in the precision machining of steel structural components, characterized in that, include: S100. Obtain the design and machining features, machining datum, dimensional relationships, and measured features of steel structural components; Establish the correspondence between the designed processing features and the measured features, perform coordinate transformation according to the processing datum, and generate the processing datum state and datum alignment feature set based on the difference between the datum features and the processing datum; S200. Calculate the local deviation of the target processing feature based on the reference alignment feature set; determine the size association path according to the size association relationship, take the difference between the actual relative size and the design relative size of adjacent associated processing features as the association deviation and synthesize it along the path to generate the local deviation state and the size accumulation state. S300: Based on the local deviation state, the cumulative size state, and the machining reference state, compare the direction and amplitude relationship between the local deviation and the cumulative size state to determine the source of the deviation; for local machining deviation, cumulative size chain deviation, and reference drift, select the target machining feature, the associated machining feature in the size association path, and the reference feature as features to be retested, and generate a retest configuration. S400. Obtain the retest data of the feature to be retested according to the retest configuration; update the corresponding local deviation status, dimensional cumulative status or processing reference status according to the category of the feature to be retested; redetermine the source of deviation based on the updated status; and generate dimensional verification results and rework positioning information.
2. The intelligent verification method for precision machining dimensional deviations of steel structural components according to claim 1, characterized in that, The design and processing features include feature identifier, geometric type, design location, design size, and dimensional tolerance; the geometric types of the design and processing features are compared with those of the measured features, and the locations of the measured features are compared with those of the design locations to establish the corresponding relationship.
3. The intelligent verification method for precision machining dimensional deviations of steel structural components according to claim 2, characterized in that, The machining datum includes a datum surface, a datum edge, and a positioning hole; a coordinate system transformation relationship is established based on the coordinates of the measurement points of the datum surface, the datum edge, and the positioning hole and the design position, and the measured features are transformed according to the coordinate system transformation relationship.
4. The intelligent verification method for precision machining dimensional deviations of steel structural components according to claim 3, characterized in that, The reference surface after coordinate transformation is fitted to generate a fitting residual; the directional deviation of the reference edge and the positional deviation of the positioning hole are calculated, and the fitting residual, directional deviation and positional deviation are compared with the corresponding allowable range of the reference to generate the machining reference state.
5. The intelligent verification method for precision machining dimensional deviations of steel structural components according to claim 2, characterized in that, The measured features are generated by fitting data from multiple measurement points; residuals are generated based on the differences between each measurement point data and the measured features, and the validity of the measurement points is determined by comparing the residuals with a preset threshold. Measurement point data whose residuals exceed the preset threshold are identified as abnormal measurement points, and the measured features are refitted using the remaining measurement point data.
6. The intelligent verification method for precision machining dimensional deviations of steel structural components according to claim 1, characterized in that, The dimension association path connecting the associated processing feature to the target processing feature is determined according to the design size and dimension association relationship; the difference between the actual relative size and the design relative size of adjacent associated processing features is calculated as the association deviation, and the association deviation is synthesized according to the directional relationship of the dimension association path to generate the dimension cumulative state.
7. The intelligent verification method for precision machining dimensional deviations of steel structural components according to claim 4, characterized in that, When the fitting residual, direction deviation, and position deviation are all within the allowable range of the corresponding reference, and the direction of the local deviation is inconsistent with the cumulative size state or the amplitude difference exceeds a preset threshold, the source of the deviation is determined to be a local processing deviation; when the fitting residual, direction deviation, and position deviation are all within the allowable range of the corresponding reference, and the direction of the local deviation is consistent with the cumulative size state and the amplitude difference is within a preset threshold, the source of the deviation is determined to be a cumulative deviation of the size chain; when the fitting residual, direction deviation, or position deviation exceeds the allowable range of the corresponding reference, and the local deviations of multiple processing features are consistent in direction, the source of the deviation is determined to be reference drift.
8. The intelligent verification method for precision machining dimensional deviations of steel structural components according to claim 7, characterized in that, When the source of deviation is local machining deviation, the retest configuration includes the target machining feature and the associated machining features adjacent to the target machining feature; when the source of deviation is cumulative deviation of the dimensional chain, the retest configuration includes the associated machining features in the dimensional association path, and the retest order is determined according to the dimensional association path; when the source of deviation is datum drift, the retest configuration includes the datum feature.
9. The intelligent verification method for precision machining dimensional deviations of steel structural components according to claim 8, characterized in that, When the retest configuration includes datum features, the machining datum is re-established and the machining datum status is updated based on the retest data; when the retest configuration includes associated machining features in the dimension association path, the associated deviation is updated and the dimension accumulation status is regenerated based on the retest data; when the retest configuration includes target machining features, the local deviation status is updated based on the retest data; the source of deviation is re-determined based on the updated machining datum status, dimension accumulation status, and local deviation status, and a retest consistency result is generated.
10. An intelligent verification system for dimensional deviations in the precision machining of steel structural components, applied to the method described in any one of claims 1 to 9, characterized in that, include: The feature correspondence module is used to obtain the design and processing features, processing references, dimensional relationships and measured features of steel structural components, and to establish the correspondence between the design and processing features and the measured features; A reference state generation module, connected to the feature correspondence module, is used to perform coordinate transformation on the measured features with corresponding relationships according to the processing reference, and generate a processing reference state and a reference alignment feature set based on the difference between the reference features and the processing reference. The deviation correlation calculation module, connected to the reference state generation module, is used to calculate the local deviation of the target processing feature based on the reference alignment feature set, determine the size correlation path according to the size correlation relationship, take the difference between the actual relative size and the design relative size of adjacent correlated processing features as the correlation deviation and synthesize it along the size correlation path to generate the local deviation state and the size cumulative state. The deviation source determination module is connected to the reference state generation module and the deviation correlation calculation module, respectively. It is used to determine the deviation source by comparing the direction and amplitude relationship between the local deviation and the size cumulative state based on the processing reference state, the local deviation state, and the size cumulative state; and to select the feature to be retested from the target processing feature, the associated processing feature in the size correlation path, and the reference feature according to the deviation source, and to generate the retest configuration. The retesting and verification module is connected to the baseline state generation module, the deviation correlation calculation module, and the deviation source determination module, respectively. It is used to obtain the retesting data of the feature to be retested according to the retesting configuration, update the corresponding processing baseline state, local deviation state, or dimensional cumulative state according to the category of the feature to be retested, redetermine the deviation source based on the updated processing baseline state, local deviation state, and dimensional cumulative state, and generate dimensional verification results and rework positioning information.
