A method and system for metrology detection of a coordinate measuring machine

Abstract

The application provides a kind of three coordinate measuring machine metrological detection method and system, related to three coordinate measurement field, including: the geometry model data of the workpiece to be measured is acquired, and the geometry model data is analyzed to determine the local geometric characteristics of the workpiece;According to the local geometric characteristics, generate path information and probe attitude information for measurement, simulate the contact between the measuring needle and the surface of the workpiece based on the path information and the probe attitude information, significantly improve the purity and reliability of the data source, the measurement data of the abnormal trigger signal removed is finely processed according to the local geometric characteristics to eliminate the deviation in the measurement data, and the geometric information of the key geometric details of the workpiece is maintained to ensure that the geometric information of the key features such as micro chamfer and sharp edge is not destroyed, which improves the accuracy, reliability and efficiency of three coordinate measuring machine in complex curved surface workpiece metrological detection, effectively solves the problem of low data acquisition efficiency.

Description

Technical Field

[0001] This application relates to the field of coordinate measuring technology, and more specifically, to a coordinate measuring machine measurement and testing method and system. Background Technology

[0002] In the field of high-end equipment manufacturing, three-dimensional dimensional inspection of precision parts is a crucial step in ensuring product quality. When performing coordinate measuring machine (CMM) measurements on complex curved workpieces with minute chamfers and areas of abrupt curvature changes, existing technologies rely on generic scan path generation rules designed for traditional workpiece geometry. This approach reveals its inherent limitations when dealing with the more complex geometries of new workpieces, particularly those with minute chamfers and abrupt curvature transitions. In these localized areas, especially where curvature changes are drastic or geometric features abruptly shift, the generic scan path often fails to achieve optimal normal contact between the probe and the workpiece surface. When the probe's contact angle deviates significantly from the surface normal, the probe may experience slight slippage or rolling at the moment of contact, leading to a discrepancy between the actual trigger point and the theoretical contact point.

[0003] This "pseudo-trigger" signal, caused by contact dynamics, has electrical characteristics very similar to the actual measurement point signal, but it contains random positional deviations. Existing probe trigger signal latching logic is typically designed to filter electrical noise caused by electromagnetic interference or mechanical vibration. Even if the measurement system reports the acquired data points, these data points themselves may have deviated from the true geometric position of the workpiece, thus contaminating the original measurement dataset and creating potential problems for subsequent data processing and analysis.

[0004] To address the aforementioned measurement point deviations, engineers attempted to adjust the parameters of the filtering module for the original measurement point data. However, since these deviations are not uniformly distributed random noise, but rather systematic errors strongly correlated with the scanning path of specific geometric regions, traditional filtering algorithms often prove ineffective. Excessive filtering strength, while eliminating some deviations, can also over-smooth realistic surface details, particularly minute chamfers and sharp edges, thus destroying the geometric information of these crucial features and preventing the measurement results from reflecting the true geometric contours of the workpiece. Summary of the Invention

[0005] This application discloses a measurement and testing method and system for a coordinate measuring machine (CMM), aiming to solve the technical problems of low data acquisition efficiency, excessive reliance on manual experience for error correction, and the inability of the probe to achieve optimal normal contact when measuring complex curved surface parts using conventional CMMs with small chamfers and areas of drastic curvature changes. Furthermore, the existing technology uses general scanning path generation rules, which leads to "false trigger" signals and contaminates the measurement dataset. In addition, traditional filtering algorithms struggle to find a balance between eliminating errors and preserving details, seriously affecting the reliability of measurement results.

[0006] The technical solution of this application is as follows:

[0007] In a first aspect, this application discloses a measurement and testing method for a coordinate measuring machine, which specifically includes the following steps:

[0008] Obtain the geometric model data of the workpiece to be tested, and analyze the geometric model data to determine the local geometric features of the workpiece;

[0009] Based on local geometric features, path information and probe posture information are generated for measurement. The path information and probe posture information are used to make the probe and the workpiece surface form a contact state close to the normal direction.

[0010] Based on path information and probe posture information, the contact between the probe and the workpiece surface is simulated to assess the probability that the probe contact point deviates from the preset contact conditions.

[0011] When the probability of deviating from the preset contact conditions is greater than the preset threshold, an instruction is generated to perform repeated measurements at multiple angles on the probe contact point.

[0012] When executing instructions for multi-angle repetitive measurements, the position information of the probe contact point, the probe attitude information, and the waveform characteristics of the probe trigger signal are collected.

[0013] Based on the results of repeated measurements from multiple angles, as well as the position information of the probe contact point, the probe posture information, and the waveform characteristics of the probe trigger signal, it is determined whether there is an abnormal trigger signal.

[0014] When an abnormal trigger signal is present, the measurement data corresponding to the abnormal trigger signal is removed from the measurement dataset; and

[0015] Measurement data for which abnormal trigger signals have been removed are processed based on local geometric features to eliminate deviations in the measurement data and maintain geometric information of key geometric details of the workpiece.

[0016] Secondly,

[0017] This application also discloses a coordinate measuring machine (CMM) metrology and testing system, which includes:

[0018] The geometric model acquisition and analysis module is used to acquire the geometric model data of the workpiece to be tested and analyze the geometric model data to determine the local geometric features of the workpiece.

[0019] The path and attitude generation module is used to generate path information and probe attitude information for measurement based on local geometric features. The path information and probe attitude information are used to make the probe and the workpiece surface form a contact state close to the normal direction.

[0020] The contact simulation and evaluation module is used to simulate the contact between the probe and the workpiece surface based on path information and probe posture information, so as to evaluate the probability that the probe contact point deviates from the preset contact conditions.

[0021] The multi-angle measurement instruction generation module is used to generate instructions for repeated multi-angle measurements of the contact point when the probability of deviating from the preset contact conditions is greater than the preset threshold.

[0022] The data acquisition module is used to acquire the position information of the probe contact point, the probe attitude information, and the waveform characteristics of the probe trigger signal when executing instructions for multi-angle repetitive measurement.

[0023] The abnormal trigger judgment module is used to determine whether there is an abnormal trigger signal based on the results of repeated multi-angle measurements, the position information of the probe contact point, the probe posture information, and the waveform characteristics of the probe trigger signal.

[0024] The abnormal data removal module is used to remove the measurement data corresponding to the abnormal trigger signal from the measurement dataset when an abnormal trigger signal is present; and

[0025] The data processing module is used to process the measurement data after the abnormal trigger signals have been removed, based on local geometric features, in order to eliminate deviations in the measurement data and maintain the geometric information of key geometric details of the workpiece.

[0026] Beneficial Effects: This application discloses a coordinate measuring machine (CMM) metrology and inspection method. By acquiring the geometric model data of the workpiece to be measured and analyzing its local geometric features, it can specifically generate path information and probe posture information that allow the probe to make near-normal contact with the workpiece surface. This effectively avoids the measurement point deviation and omission problems caused by non-optimal contact in complex curved surface areas due to traditional general scanning paths. It effectively compensates for the limitations of single-angle measurement and avoids data points with deviations from contaminating the original measurement dataset, significantly improving the purity and reliability of the data source. For measurement data with abnormal trigger signals removed, it performs fine processing based on local geometric features to eliminate deviations in the measurement data and maintain the geometric information of key geometric details of the workpiece, ensuring that the geometric information of key features such as small chamfers and sharp edges is not destroyed. This improves the accuracy, reliability, and efficiency of CMM in metrology and inspection of complex curved surface workpieces. It effectively solves many technical problems in the prior art, such as low data acquisition efficiency, reliance on manual experience for error correction, difficulty in identifying false trigger signals, and the dilemma of filtering processing. It provides a more advanced and reliable solution for the inspection of precision parts in the field of high-end equipment manufacturing. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of a coordinate measuring machine measurement and testing method provided in this application.

[0028] Figure 2 This is a schematic diagram of a coordinate measuring machine metrology and testing system provided in this application. Detailed Implementation

[0029] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0030] Reference Figure 1 The diagram illustrates an embodiment of a coordinate measuring machine metrology and testing method according to the present invention, which may specifically include the following steps:

[0031] S101, acquire the geometric model data of the workpiece to be tested, and analyze the geometric model data to determine the local geometric features of the workpiece;

[0032] S102, Based on the local geometric features, generate path information and probe posture information for measurement, the path information and the probe posture information are used to make the probe and the workpiece surface form a contact state close to the normal direction;

[0033] S103, based on the path information and the probe posture information, simulate the contact between the probe and the workpiece surface to assess the probability that the probe contact point deviates from the preset contact conditions;

[0034] S104, when the probability of deviating from the preset contact condition is greater than the preset threshold, an instruction is generated to perform repeated multi-angle measurements on the probe contact point.

[0035] S105, when executing the instruction for multi-angle repetitive measurement, collects the position information of the probe contact point, the probe attitude information, and the waveform characteristics of the probe trigger signal;

[0036] S106. Based on the results of the multi-angle repeated measurements, the position information of the probe contact point, the probe posture information, and the waveform characteristics of the probe trigger signal, determine whether there is an abnormal trigger signal.

[0037] S107, when an abnormal trigger signal exists, the measurement data corresponding to the abnormal trigger signal is removed from the measurement dataset;

[0038] S108, the measurement data of the removed abnormal trigger signal is processed according to the local geometric features to eliminate the deviation in the measurement data and maintain the geometric information of the key geometric details of the workpiece.

[0039] This application aims to reduce measurement errors at the source and improve the measurement accuracy and data reliability of complex curved workpieces through a series of steps, including optimizing the measurement path and probe posture, simulating contact conditions, performing repeated measurements from multiple angles, identifying and removing abnormal trigger signals, and data processing based on local geometric features. This effectively solves the problems of inaccurate measurement data and loss of key geometric details in the prior art.

[0040] To better understand the technical solution proposed in this application, some key terms are explained first. "Geometric model data" refers to the digital three-dimensional representation of the workpiece to be measured, typically existing in the form of CAD models, point cloud data, or mesh models, containing complete geometric information of the workpiece. "Local geometric features" refer to areas on the workpiece surface with specific geometric attributes, such as chamfers, arcs, holes, grooves, and bosses; these features are often crucial to the function and performance of the workpiece. "Path information" refers to the spatial trajectory followed by the probe when measuring on the workpiece surface. "Probe posture information" refers to the direction and angle of the probe relative to the workpiece surface during measurement. "Preset contact conditions" refer to the ideal state that the probe should meet when in contact with the workpiece surface; for example, the angle between the probe axis and the normal direction of the workpiece surface should be less than a certain value to ensure measurement accuracy. "Abnormal trigger signal" refers to a trigger signal caused by non-ideal contact (such as probe slippage or lateral wobbling), but whose electrical characteristics are similar to the actual measurement point signal; these signals can cause the measurement data to deviate from the true geometric position of the workpiece.

[0041] The core of the coordinate measuring machine measurement and inspection method proposed in this application lies in ensuring the accuracy and reliability of measurement data and effectively preserving the key geometric details of the workpiece through a series of refined steps.

[0042] Specifically, in the step of acquiring and analyzing the geometric model data of the workpiece to be tested to determine its local geometric features, various methods can be used to acquire the geometric model data. For example, a precise 3D model of the workpiece can be obtained by importing CAD design files (such as STEP or IGES formats). Alternatively, a preliminary scan of the workpiece can be performed to generate high-density point cloud data, which can then be fitted into a geometric model using reverse engineering software. When analyzing the geometric model data, specialized geometric analysis software can be used to automatically identify and extract various local geometric features on the workpiece surface. For example, it can identify all fillets with radii smaller than a specific value, all regions with a rate of curvature greater than a specific threshold, or all features with specific topological structures (such as holes, slots, etc.).

[0043] In the step of generating path information and probe posture information for measurement based on the local geometric features, whereby the path information and probe posture information are used to ensure that the probe and the workpiece surface form a contact state close to the normal direction, an adaptive path planning algorithm can be employed based on the identified local geometric features. For example, for areas with gentle curvature changes, an equidistant scanning path can be used; while for areas with drastic curvature changes or small chamfers, a more refined scanning path along the curvature direction can be generated. The probe posture information can be generated by calculating the normal direction of each point on the workpiece surface and ensuring that the probe's axial direction is consistent with or nearly consistent with the surface normal direction. For example, the angle between the probe axis and the surface normal direction can be set to always be less than 5 degrees.

[0044] In the step of simulating the contact between the probe and the workpiece surface based on the path information and the probe posture information to assess the probability that the probe contact point deviates from the preset contact conditions, virtual measurement software can be used for simulation. This software can simulate the dynamic contact process of the probe as it executes the measurement path, based on preset probe geometry, probe kinematics model, and workpiece geometry model. For example, it can simulate the change in the angle between the probe's axis and the surface normal direction when it contacts the workpiece surface, and calculate in which areas this angle might exceed a preset threshold. Furthermore, it can simulate the minute sliding or rolling that the probe may experience at the moment of contact, and based on these simulation results, assess the probability that the probe contact point deviates from the preset contact conditions.

[0045] In the step of generating an instruction to perform repeated multi-angle measurements on the probe contact point when the probability of deviating from the preset contact condition is greater than a preset threshold, a probability threshold can be set, such as 20%. When the simulation results show that the probability of a probe contact point deviating from the preset contact condition exceeds 20%, the system will automatically generate a multi-angle repeated measurement instruction for that contact point. For example, the probe can be instructed to perform repeated measurements on the point at three or more different angles (e.g., deflected ±15 degrees from the original normal direction) to obtain more comprehensive data.

[0046] When executing instructions for multi-angle repetitive measurements, the data acquisition module records in real time the three-dimensional coordinates (position information) of the probe contact point, the spatial attitude of the probe at the moment of triggering (probe attitude information), and the waveform data of the electrical signal generated by the internal sensor of the probe as it changes over time (waveform characteristics of the trigger signal) during the steps of acquiring the position information of the probe contact point, the probe attitude information, and the waveform data of the electrical signal generated by the probe internal sensor as it changes over time. For example, it can record characteristic parameters such as the rise time, fall time, pulse width, and peak voltage of the trigger signal.

[0047] In the step of determining whether an abnormal trigger signal exists based on the results of repeated multi-angle measurements, the position information of the probe contact point, the probe posture information, and the waveform characteristics of the probe trigger signal, a consistency analysis can be performed on the collected multi-angle measurement data. For example, if the position information obtained from the same contact point under different angle measurements shows significant differences, or if the waveform characteristics of the trigger signal deviate significantly from the waveform characteristics of the normal trigger signal (e.g., pulse width is too narrow or too wide, peak voltage is abnormal), then it can be determined that an abnormal trigger signal exists.

[0048] In the step of removing the measurement data corresponding to the abnormal trigger signal from the measurement dataset when an abnormal trigger signal is detected, the system automatically deletes the measurement data, such as the position information and probe attitude information corresponding to these abnormal signals, from the original measurement dataset once the abnormal trigger signal is identified. For example, if a measurement point is determined to be abnormally triggered, all relevant data of that measurement point (X, Y, Z coordinates, probe attitude, etc.) will be removed to avoid negatively impacting subsequent data processing.

[0049] In the step of processing the measurement data of the removed abnormal trigger signals according to the local geometric features to eliminate deviations in the measurement data and maintain the geometric information of the key geometric details of the workpiece, an adaptive data processing algorithm based on local geometric features can be used. For example, for the measurement data of the removed abnormal trigger signals, the most suitable fitting algorithm can be selected according to the type of local geometric feature (such as plane, cylindrical surface, spherical cap surface, chamfer, etc.). For key geometric detail areas, higher weights can be assigned to ensure that the geometric information of these details is not over-smoothed or lost during data processing. For example, when fitting a surface to a small chamfer area, a locally weighted least squares method can be used, and the measurement data of the chamfer area can be assigned higher weights to ensure that the fitted surface can accurately reflect the geometry of the chamfer.

[0050] To further reduce potential measurement risks, the system simulates the contact between the probe and the workpiece surface based on the generated path and orientation information before actual measurement. By assessing the probability that the probe contact point deviates from the preset contact conditions, the system can pre-identify high-risk areas.

[0051] During multi-angle repetitive measurements, the system not only collects the position information of the probe contact point and the probe attitude information, but more importantly, it also collects the waveform characteristics of the probe trigger signal. These waveform characteristics are the key basis for judging "false trigger" signals. Through comprehensive analysis of the multi-angle measurement results and the waveform characteristics of the trigger signal, the system can accurately determine whether there is an abnormal trigger signal. Once an abnormal trigger signal is identified, its corresponding measurement data will be immediately removed from the measurement dataset, thereby effectively purifying the original measurement data and avoiding contamination of subsequent processing by erroneous data.

[0052] Finally, for the measurement data after the abnormal trigger signals have been removed, the system performs fine-tuning based on the previously determined local geometric features. This processing method is adaptive; it employs the most suitable algorithm to eliminate residual deviations in the measurement data for different types of local geometric features. Simultaneously, by paying special attention to and processing key geometric details, it ensures that this important geometric information is maintained throughout the data processing, avoiding the detail loss problems that may occur with traditional filtering methods. The entire process forms a complete closed loop from prevention, detection, and correction, ensuring the high accuracy of the final measurement results and a faithful reflection of the workpiece's true geometric contour.

[0053] This application effectively overcomes many limitations of existing technologies in measuring complex curved workpieces by introducing innovative technologies such as contact simulation, multi-angle repeated measurement, abnormal trigger signal waveform feature analysis, and adaptive data processing based on local geometric features, and significantly improves the accuracy, efficiency and reliability of coordinate measuring machine metrology and inspection.

[0054] This application further proposes the following steps for generating instructions to perform multi-angle repeated measurements on the probe contact point when the probability of deviating from the preset contact conditions is greater than a preset threshold:

[0055] Identify the geometric features at the probe contact point to obtain geometric feature information at the probe contact point;

[0056] Based on the geometric feature information, a first probe posture combination is generated;

[0057] Identify the type of instability at the probe contact point to obtain information about the type of instability at the probe contact point;

[0058] Based on the unstable condition type information, the first probe posture combination is adjusted to obtain the second probe posture combination; and based on the second probe posture combination, the multi-angle repetitive measurement command is generated.

[0059] Specifically, identifying the geometric features at the probe contact point refers to determining the geometric attributes of the area where the probe contact point is located by analyzing the geometric model data of the workpiece under test. These attributes include curvature, slope, edge type (such as sharp edges or rounded corners), hole depth, and groove width. This geometric feature information can guide subsequent probe posture planning. For example, for high-curvature areas, more probe postures may be needed to ensure full coverage; for sharp edges, it is necessary to avoid secondary contact between the probe and the edge.

[0060] The generation of a first probe posture combination based on the geometric feature information can be understood as initially planning a set of probe postures based on the identified geometric features. For example, for a planar region, a small number of postures can be generated around the normal direction; for a circular arc surface, postures varying along the tangent direction of the arc can be generated. This first probe posture combination aims to provide a basic, comprehensive set of measurement angles.

[0061] In practical applications, identifying the type of instability at the probe contact point refers to further analyzing the contact simulation results or combining them with historical measurement data to determine potential measurement instability factors at that contact point. For example, instability type information may include the risk of probe slippage, the risk of the probe forming non-ideal contact with the workpiece surface (such as lateral contact or multi-point contact), and the risk of jitter or delay in the probe trigger signal. This instability type information can be evaluated based on parameters such as the contact force distribution between the probe and the workpiece surface, the size of the contact area, and the probe deflection in the contact simulation.

[0062] Furthermore, based on the unstable condition type information, the first probe posture combination is adjusted to obtain a second probe posture combination. This adjustment process aims to optimize the probe posture to avoid or mitigate the effects of unstable conditions. For example, if a probe slippage risk is identified, the probe posture can be adjusted to make its contact direction closer to the surface normal, or the diversity of contact angles can be increased to disperse the risk; if there is a risk of non-ideal contact, the probe posture can be adjusted to avoid such contact at specific angles. The second probe posture combination is an optimized and more robust set of measurement postures.

[0063] Therefore, generating the multi-angle repetitive measurement instruction based on the second probe posture combination means converting the optimized probe posture set into an executable measurement program instruction, which guides the coordinate measuring machine to repeatedly measure the probe contact point with these specific probe postures during the actual measurement process.

[0064] In some preferred embodiments, a specific example is given below. Assume the workpiece to be tested has a deep groove structure with sharp edges. In the initial contact simulation, the system assesses that the probability of a probe contact point at the bottom edge of the groove deviating from preset contact conditions is greater than a preset threshold; for example, the simulation shows that the probe may slightly slip or make secondary contact with the edge at that point.

[0065] At this point, the solution of this application will first identify the geometric features at the probe contact point, namely, that it is located at the bottom of a deep groove with a sharp edge. Based on this geometric feature, the system will generate a first probe posture combination, for example, including several postures tilted along the groove wall normal and along the groove length.

[0066] Next, the system will identify the type of instability at the probe contact point, such as determining whether there is a "probe slippage risk" or a "secondary contact risk at the edge".

[0067] Based on this information about the types of instability, the system will adjust the first probe posture combination. For example, to reduce the risk of slippage, it may increase the posture closer to the surface normal and fine-tune the contact angle to avoid parallelism with sharp edges; to avoid secondary contact, it may restrict certain overly tilted postures or increase the safety distance between the probe and the edge. This results in a second probe posture combination that includes more optimized and safer measurement angles.

[0068] Finally, based on this combination of second probe postures, multi-angle repeatable measurement commands are generated, guiding the coordinate measuring machine to repeatedly measure the probe contact point at the bottom edge of the groove using these finely adjusted postures during actual measurement. In this way, even in complex areas prone to unstable contact, the accuracy and reliability of the measurement can be ensured.

[0069] This application further proposes a method for processing measurement data of removed abnormal trigger signals, the steps of which include:

[0070] Acquire information on the force change experienced by the probe when it contacts the surface of the workpiece;

[0071] Based on the geometric model of the workpiece and the material properties of the workpiece, determine the normal contact force curve of the probe;

[0072] Compare the force change information with the normal contact force curve to determine whether there is a risk of stable displacement at the probe contact point;

[0073] If the risk of instability exists, adjust the contact dynamics parameters of the probe to perform a second measurement;

[0074] Obtain the probe contact point position information and force change information obtained from the second measurement;

[0075] By comparing the position information of the probe contact point obtained from the original measurement point with that obtained from the remeasurement, and combining the force change information obtained from the remeasurement, it is possible to determine whether there is a stable offset at the original measurement point, and obtain the measurement data with a stable offset.

[0076] The measurement data containing the stable offset are removed from the measurement data of the removed abnormal trigger signals; and

[0077] Based on the geometric model data of the workpiece and the confidence level of the measurement data, surface fitting is performed on the measurement data to reconstruct the geometric features of the workpiece.

[0078] Specifically, acquiring information on the force changes experienced by the probe when it contacts the workpiece surface refers to real-time monitoring of the changes in force and torque experienced by the probe during contact with the workpiece surface using force sensors, strain gauges, or other devices integrated inside the probe or on the probe itself. This force change information reflects the dynamic contact behavior between the probe and the workpiece surface, such as the instantaneous magnitude, direction, and duration of the contact force. Specifically, this force change information can be understood as the trajectory of the three-dimensional force vector experienced by the probe over time, from initial contact to stable contact, and then to the issuance of the probe trigger signal. Its purpose is to capture minute physical forces that may cause probe deviation.

[0079] Specifically, determining the normal contact force curve of the probe based on the workpiece's geometric model and material properties involves using the workpiece's CAD model data and its known material properties (e.g., elastic modulus, Poisson's ratio, hardness, etc.), combined with the probe's geometry and material properties, to calculate and predict the typical mechanical response curve that should occur when the probe contacts the workpiece surface under ideal, non-offset contact conditions, through finite element analysis (FEA) or empirical formulas. This normal contact force curve serves as a benchmark for subsequent comparison with force variation information obtained in actual measurements. The aim is to establish a theoretically ideal contact force model to identify abnormal mechanical behavior in actual measurements.

[0080] In practical applications, comparing the force change information with the normal contact force curve to determine whether there is a risk of stable offset at the probe contact point involves comparing the actually collected force change information with a pre-determined normal contact force curve. This comparison can be achieved by calculating the difference, correlation, or deviation of specific characteristic points (such as peak value or duration) between the two. When the deviation between the actual force change information and the normal contact force curve exceeds a preset threshold, and this deviation has a certain degree of persistence or regularity, a risk of stable offset is considered to exist. Stable offset risk refers to the deviation of the probe's actual contact point from the expected theoretical contact point due to continuous, non-instantaneous mechanical action when the probe contacts the workpiece surface, and this deviation remains relatively stable. Its purpose is to identify systematic deviations that are not abnormal trigger signals but may lead to inaccurate measurement point positions.

[0081] In a preferred embodiment, when the risk of stable offset exists, adjusting the contact dynamics parameters of the probe for re-measurement means that the system automatically or manually modifies relevant parameters of the coordinate measuring machine when performing the measurement task, such as the probe's contact speed, acceleration, trigger force threshold, or probe posture, based on the identified risk of stable offset. The adjusted parameters aim to optimize the contact mode between the probe and the workpiece, in order to reduce or eliminate stable offset during re-measurement. The purpose is to verify and correct any potential stable offset by changing the measurement conditions, thereby obtaining more accurate measurement data.

[0082] Furthermore, obtaining the probe contact point position information and force change information obtained from the remeasurement refers to re-performing the measurement operation on the same probe contact point after adjusting the contact dynamics parameters, and again collecting the precise position data of the probe and the force change information experienced during the contact process. This remeasured data will be compared with the original measurement data to confirm the existence and extent of stable offset. The purpose is to obtain comparative data for verifying and quantifying stable offset through repeated measurements.

[0083] The process involves comparing the original measurement point with the probe contact point position obtained from a subsequent measurement, and combining this with the force change information obtained from the subsequent measurement, to determine whether there is a stable offset at the original measurement point. This results in measurement data showing a stable offset. Specifically, it involves a precise comparison between the probe contact point position obtained from the first measurement and the probe contact point position obtained after parameter adjustment. Simultaneously, the force change information obtained from both measurements is used to comprehensively analyze and determine whether the original measurement point has undergone a systematic, non-random deviation due to continuous mechanical action. If there is a significant and consistent displacement between the two measurement points, and this displacement is consistent with the stable offset risk indicated by the force change information, then the original measurement point is marked as measurement data showing a stable offset. The aim is to accurately identify and quantify the measurement error caused by stable offset through multi-dimensional data comparison and verification.

[0084] Therefore, removing measurement data with stable offsets from the measurement data with removed abnormal trigger signals means that after identifying measurement data with stable offsets, these data are removed from the current measurement dataset used for subsequent processing. This step ensures that subsequent data processing and geometric reconstruction are based only on high-quality measurement data without stable offsets. Its purpose is to further purify the measurement dataset and improve the overall accuracy and reliability of the data.

[0085] Specifically, based on the workpiece's geometric model data and the confidence level of the measurement data, surface fitting is performed on the measurement data to reconstruct the workpiece's geometric features. This involves using the remaining, high-quality measurement data, combined with the workpiece's original geometric model data, after removing abnormal trigger signals and stable offset measurement data, to perform a high-precision surface fitting operation. The confidence level of the measurement data can be evaluated based on factors such as the repeatability of the measurement points, the stability of the measurement environment, the accuracy of the probe calibration, and the deviation from the theoretical model. During the surface fitting process, different weights can be assigned to each measurement point based on its confidence level, allowing high-confidence data to have a greater impact on the fitting result, thereby more accurately reconstructing the workpiece's geometric features, especially at critical geometric details. The aim is to use the most reliable data to reconstruct the workpiece's geometry in the most optimized way, ensuring the accuracy of the reconstruction result and faithful reproduction of key details.

[0086] This application effectively identifies and eliminates stable offsets caused by minute physical interactions between the probe and the workpiece surface, thereby significantly improving the accuracy and reliability of measurement data. Surface fitting, combined with the confidence level of the measurement data, results in more accurate reconstructed workpiece geometry, particularly in maintaining key geometric details, thus more faithfully reflecting the true shape of the workpiece and enhancing the overall accuracy and reliability of coordinate measuring machine (CMM) metrology and inspection.

[0087] According to the scheme of this application, firstly, during the measurement process, the force sensor inside the probe acquires the force change information of the probe when it contacts the blade surface in real time. For example, when the probe contacts the blade edge or the inner wall of the hole, the force sensor may detect a mechanical response that deviates continuously from the normal contact force curve.

[0088] Next, the system pre-calculates the normal contact force curve under ideal contact conditions based on the blade's geometric model data and the properties of its aerospace-grade alloy materials. By comparing the actual force variation information with this normal contact force curve, the system identifies certain measurement points at the blade edge and the inner wall of the orifice that have a risk of stable offset.

[0089] When a risk of stable offset is identified, the system automatically adjusts the probe's contact speed and trigger force threshold, and remeasures these risk points. For example, it reduces the contact speed and fine-tunes the trigger force to minimize the dynamic interaction between the probe and the workpiece.

[0090] After the second measurement, the system acquires new information on the probe contact point location and force change. By comparing the positions of the original and second measurement points, and combining the force change information from the two measurements, the system confirms that the original measurement point does indeed have a stable offset. For example, the original measurement point has a systematic deviation of 0.5 micrometers in a certain direction relative to the second measurement point, and the force change curves in both measurements show a continuous non-ideal contact force.

[0091] Subsequently, these measurement data that were confirmed to have stable offsets will be removed from the measurement dataset where the abnormal trigger signals have been removed.

[0092] Finally, using the remaining high-quality measurement data, combined with the blade's geometric model data and the confidence level of each measurement point (e.g., points with high repeatability have high confidence), a surface fitting was performed on the blade surface. During the fitting process, high-confidence data points were given higher weights, ensuring that the geometric features of the blade's complex surface and micro-holes could be accurately reconstructed, and eliminating the bias introduced by stable offsets, thereby maintaining the geometric information of the blade's key geometric details.

[0093] This application further proposes a coordinate measuring machine (CMM) metrology and inspection method, wherein the step of performing surface fitting on the measurement data based on the workpiece's geometric model data and the confidence level of the measurement data to reconstruct the workpiece's geometric features includes:

[0094] The geometric model data of the workpiece is analyzed to identify areas in the geometric model data that differ from the micro-geometric features allowed by the manufacturing tolerances of the workpiece.

[0095] A geometric deformation range is set for the region of difference in microscopic geometric features;

[0096] When performing surface fitting, based on the confidence level of the measured data, within the allowable geometric deformation range, the constraint strength of the geometric model data in the region of microscopic geometric feature difference is adjusted to obtain the fitted surface;

[0097] Monitor the local deviations between the fitted surface and the geometric model at key geometric details;

[0098] When the deviation of the fitted surface from the high-confidence measurement data at the key geometric details meets a deviation threshold, or when the difference between the geometric features of the fitted surface at the key geometric details and the geometric model meets specific conditions, a detail correction operation is triggered; and

[0099] Using the high-confidence measurement data and combining it with the topological information of the geometric model at the key geometric details, the fitted surface is locally adjusted to obtain a local adjustment region, so that the geometric information of the key geometric details is consistent with the measurement data and the geometric model.

[0100] Specifically, the geometric model data of the workpiece is analyzed to identify areas where slight differences may arise due to tolerances during manufacturing; these areas are referred to as micro-geometric feature difference areas. For example, in areas with acute angles, small-radius fillets, or thin-walled structures, manufacturing tolerances are more likely to cause slight deviations between the actual geometry and the ideal geometric model. Setting a geometric deformation range for these micro-geometric feature difference areas means defining an allowable range of geometric change for these potentially differing areas, ensuring that the deformation of these areas does not exceed the limits of actual manufacturing tolerances during surface fitting.

[0101] During surface fitting, the constraint strength of the geometric model data in the microscopic geometric feature difference region is adjusted according to the confidence level of the measured data and within the allowable geometric deformation range. This means that for high-confidence measurement data, the constraint of the geometric model on the fitted surface can be appropriately relaxed in the microscopic geometric feature difference region to better reflect the actual measurement data; while for low-confidence data, the constraint of the geometric model can be strengthened to avoid introducing inaccurate deformation. Thus, a preliminary fitted surface can be obtained.

[0102] Furthermore, it is necessary to monitor local deviations between the fitted surface and the geometric model at key geometric details. Key geometric details typically refer to features crucial to the functionality, assembly, or aesthetics of the workpiece, such as holes, slots, bosses, and chamfers. Monitoring deviations in these areas helps ensure the accuracy of these critical features. A detail correction operation is triggered when the deviation of the fitted surface from the high-confidence measurement data at the key geometric details meets a deviation threshold, or when the difference between the geometric features of the fitted surface at the key geometric details and the geometric model meets specific conditions. This indicates that a more refined correction process needs to be initiated when the initial fitting results show unexpected or potential problems in critical areas.

[0103] Finally, using the high-confidence measurement data and combining it with the topological information of the geometric model at the key geometric details, the fitted surface is locally adjusted to obtain the local adjustment region. The topological information includes the connection relationships between points, lines, and surfaces, which is crucial for maintaining the integrity and consistency of the geometric details. Through this local adjustment, the geometric information of the key geometric details can be made consistent with the measurement data and the geometric model, thereby accurately reconstructing the key geometric features of the workpiece based on the overall fitted surface.

[0104] This application overcomes the problem that traditional surface fitting methods may lead to distortion of key geometric details or excessive smoothing when dealing with workpieces with complex geometric features and manufacturing tolerances. By identifying areas of difference in microscopic geometric features and setting the range of geometric deformation, the fitting process can better adapt to actual manufacturing tolerances, avoiding unnecessary geometric constraints. Furthermore, by adjusting the constraint strength according to the confidence level of the measurement data, the effective use of high-precision measurement data in key areas is ensured, while reducing the negative impact of low-precision data on the overall fitting. Most importantly, by monitoring local deviations at key geometric details and triggering detail correction operations, this application can precisely adjust these important areas in a targeted manner. Combining high-confidence measurement data and the topological information of the geometric model, it ensures that the reconstructed workpiece geometric model is not only smooth overall but also highly consistent with the actual workpiece at key details. This significantly improves the accuracy and reliability of coordinate measuring machine (CMM) metrology and provides a more accurate geometric data foundation for subsequent quality control and product development.

[0105] In some preferred embodiments, a specific example is given below. Assume the workpiece to be tested is a complex mechanical part with multiple holes and chamfers. First, the system analyzes the geometric model data of the workpiece, identifying these hole and chamfer areas as regions with microscopic geometric feature differences, as these areas are easily affected by tolerances during manufacturing. For example, the diameter of a hole may deviate slightly within the design tolerance range, and the chamfer angle may also have minor variations. An allowable geometric deformation range is set for these areas; for example, the hole diameter is allowed to vary within ±0.02 mm, and the chamfer angle is allowed to vary within ±1 degree. When performing surface fitting, if the measurement data for a certain hole area has a high confidence level, indicating that the measurement results are very reliable, then when fitting that hole, the constraint strength of the geometric model on the hole diameter will be appropriately relaxed, allowing the fitted surface to better fit the actual measurement data within a deformation range of ±0.02 mm.

[0106] This application further proposes a method to optimize the triggering mechanism of detail correction operations, which improves the accuracy and adaptability of correction operations by dynamically adjusting the deviation threshold or specific conditions that trigger detail correction operations.

[0107] The steps to trigger the detail correction operation when the deviation between the fitted surface and the high-confidence measurement data at key geometric details meets the deviation threshold, or when the difference between the geometric features of the fitted surface at key geometric details and the geometric model meets specific conditions, include:

[0108] Obtain manufacturing tolerance and material property information for the critical geometric detail areas to be corrected;

[0109] Based on the manufacturing tolerance information and the material property information, adjust the deviation threshold or specific conditions that trigger the detail correction operation;

[0110] Based on the adjusted deviation threshold or specific conditions, determine whether the deviation between the fitted surface and the high-confidence measurement data at the key geometric details meets the deviation threshold, or whether the difference between the geometric features of the fitted surface at the key geometric details and the geometric model meets the specific conditions; and when the determination result shows that the deviation meets the deviation threshold or the difference meets the specific conditions, trigger the detail correction operation.

[0111] Specifically, manufacturing tolerance information for the critical geometric details area to be corrected can include dimensional tolerances, form tolerances, and positional tolerances for that area. This information is typically derived from the workpiece's design drawings or manufacturing specifications. Material property information can include the area's elastic modulus, yield strength, hardness, and coefficient of thermal expansion. This information reflects the material's response characteristics under stress or temperature changes. This information can be obtained by consulting material databases, conducting material tests, or extracting it from the workpiece's design documents.

[0112] Adjusting deviation thresholds or specific conditions based on manufacturing tolerance and material property information refers to dynamically setting or modifying the judgment criteria used to determine whether to trigger detail correction operations based on the actual physical and manufacturing properties of key geometric detail areas. For example, for areas with strict tolerance requirements, the deviation threshold can be set smaller; for materials with low elastic modulus, they may produce larger local deformations under measurement forces, and specific conditions can be appropriately adjusted to account for this deformation. This adjustment can be achieved through preset rules, lookup tables, or machine learning-based models to ensure the rationality and adaptability of the judgment criteria.

[0113] In practical applications, various geometric comparison algorithms can be used to determine whether the deviation between the fitted surface and the high-confidence measurement data at key geometric details meets the adjusted deviation threshold, or whether the differences between the geometric features of the fitted surface at key geometric details and the geometric model meet the adjusted specific conditions. For example, the maximum distance, average distance, or root mean square distance between the fitted surface and the high-confidence measurement data points can be calculated and compared with the adjusted deviation threshold. For differences in geometric features, characteristic parameters such as curvature, normal direction, and edge sharpness can be compared and matched with the adjusted specific conditions.

[0114] When the judgment result shows that the deviation meets the adjusted deviation threshold or the difference meets the adjusted specific conditions, it is considered that there is a geometric inconsistency in the key geometric detail area that needs to be corrected, and the detail correction operation will be triggered. This operation may include local reconstruction of the fitted surface, adjustment of the topology, or application of other correction algorithms, so that the reconstructed geometric information more accurately reflects the actual geometric features of the workpiece.

[0115] This application achieves significant optimization of the triggering mechanism for detail correction operations in coordinate measuring machine (CMM) metrology and inspection methods. Specifically, by acquiring and utilizing manufacturing tolerance information and material property information of key geometric detail areas, the system can dynamically adjust the deviation threshold or specific conditions for triggering detail correction operations, thereby making the triggering of correction operations more precise and intelligent. This adaptive judgment mechanism effectively avoids the problems of over-correction or under-correction caused by using static thresholds, improving the accuracy and reliability of measurement data processing. Furthermore, this scheme can better adapt to workpieces of different materials and manufacturing processes, enhancing the versatility and robustness of the method, and ensuring that the geometric information of key geometric details can be more accurately maintained and reconstructed in the inspection of complex workpieces.

[0116] In some preferred embodiments, a specific example is given below. Assume the workpiece under test is an aero-engine blade, whose key geometric details include the leading and trailing edges. After surface fitting, it is necessary to determine whether to trigger detail correction operations for these areas.

[0117] First, the system acquires manufacturing tolerance information for the leading edge region of the blade, such as a dimensional tolerance of ±0.02 mm and a shape tolerance of 0.01 mm. Simultaneously, it acquires material property information for this region; for example, the blade material is a high-temperature alloy with a high elastic modulus, but localized stress concentrations may lead to minor deformations.

[0118] Based on this information, the system will adjust the deviation threshold or specific conditions that trigger detailed correction operations. For example, for the leading edge of a blade, because it has a significant impact on aerodynamic performance and has strict tolerance requirements, the system will adjust the deviation threshold from the default 0.05mm to 0.015mm and set a more stringent curvature difference specific condition to ensure that even small geometric deviations can be identified.

[0119] Subsequently, the system will determine whether the deviation between the fitted surface at the leading edge of the blade and the high-confidence measurement data meets the threshold, or whether the difference between the geometric features of the fitted surface at the leading edge of the blade and the geometric model meets the specific condition, based on the adjusted deviation threshold of 0.015mm and the more stringent curvature difference specific condition.

[0120] If the assessment results show that the local deviation of the fitted surface at the blade leading edge exceeds 0.015 mm, or that its curvature difference does not meet the adjusted specific conditions, a detailed correction operation for the blade leading edge will be immediately triggered. In this way, the solution proposed in this application can ensure more accurate detection and correction of key geometric details of high-precision workpieces such as aero-engine blades, thereby improving product quality and performance.

[0121] This application further proposes a method for metrological testing using a coordinate measuring machine, wherein the step of using the high-confidence measurement data, combined with the topological information of the geometric model at the key geometric details, to locally adjust the fitted surface to obtain a locally adjusted region, so that the geometric information of the key geometric details is consistent with the measurement data and the geometric model, includes:

[0122] The consistency between the high-confidence measurement data and the topological information of the geometric model at the key geometric details is evaluated to identify minor inconsistencies between the two.

[0123] Based on the geometric features and topological relationships of the minor inconsistency regions, the local adjustment region is divided into multiple sub-regions, and a local adjustment strategy is generated for each sub-region.

[0124] Within each sub-region, based on the local adjustment strategy, local parametric surface reconstruction is performed using the high-confidence measurement data and the geometric model topology information to generate local adjustment surface patches.

[0125] The smoothness of the connection between the generated local adjustment surface patch and the fitted surface at the boundary of the sub-region is evaluated to obtain the smoothness evaluation result;

[0126] When the smoothness evaluation result does not meet the preset smoothness threshold, the parameters of the locally parameterized surface reconstruction are adjusted; and

[0127] The local adjustment surface patch whose smoothness evaluation result meets the preset smoothness threshold is seamlessly spliced ​​with the fitted surface.

[0128] Specifically, assessing the consistency between high-confidence measurement data and the topological information of the geometric model at key geometric details involves comparing the geometric attributes (e.g., location, normal direction) and topological relationships (e.g., edges, face connections) of the corresponding regions in the measurement data points and the geometric model to quantify the differences between them. This allows for the precise identification of minor inconsistencies that exceed allowable tolerances or exhibit structural deviations. These inconsistencies may be caused by manufacturing errors, measurement noise, or model simplification.

[0129] Based on the geometric characteristics and topological relationships of the minor inconsistencies, the local adjustment region is divided into multiple sub-regions. The purpose is to decompose the complex adjustment task into smaller, more manageable units. Each sub-region can be assigned a customized local adjustment strategy based on its specific geometric shape (e.g., plane, curved surface, edge, hole, etc.) and its connection relationship with other regions. For example, different surface fitting algorithms, control point distributions, or weights can be used.

[0130] In practical applications, within each sub-region, local parametric surface reconstruction is performed using high-confidence measurement data and geometric model topology information, based on a local adjustment strategy, to generate locally adjusted surface patches. Local parametric surface reconstruction can be understood as refining the surface model within the sub-region using high-confidence measurement data while maintaining the original topology of the geometric model. This can be achieved by employing techniques such as B-splines, NURBS, or other freeform surface techniques, adjusting control points or weights to approximate the measurement data, while simultaneously ensuring the smoothness and continuity of the surface.

[0131] Furthermore, the smoothness of the connection between the generated locally adjusted surface patch and the fitted surface at the sub-region boundary is evaluated. The purpose is to ensure that the locally adjusted surface patch can achieve a smooth and seamless transition with the surrounding fitted surface, avoiding sharp edges, breaks, or discontinuous curvature changes. The smoothness evaluation results can be judged based on the continuity of the first, second, or even higher-order derivatives of the surface.

[0132] When the smoothness evaluation result does not meet the preset smoothness threshold, the parameters of the local parametric surface reconstruction are adjusted. For example, the number, position, and weight of control points can be adjusted, or the surface order and the number of fitting iterations can be changed to optimize the shape of the surface patch so that it can better connect with the fitted surface at the boundary.

[0133] Finally, the local adjustment surface patch whose smoothness evaluation results meet the preset smoothness threshold is seamlessly spliced ​​with the fitted surface, thereby forming a reconstructed surface that conforms to high-confidence measurement data and maintains the integrity of the geometric model's topology.

[0134] In some preferred embodiments, it is assumed that the workpiece to be tested is a turbine blade with a complex curved surface and multiple sharp edges. After performing coordinate measuring machine (CMM) measurements on the blade and completing preliminary surface fitting, it was found that in a key blade leading edge region, despite having high-confidence measurement data, there was a slight geometric deviation between the fitted surface and the geometric model in this region, and there may be slight unevenness at the boundary.

[0135] At this point, the solution proposed in this application will be applied. First, the system will assess the consistency between the high-confidence measurement data of the leading edge region of the blade and the topological information of the geometric model at that location. For example, by calculating the distance from the measurement point to the model surface, the angle between the normal of the measurement point and the normal of the model, several minor inconsistencies in the leading edge region can be identified. For instance, the measurement data may slightly deviate from the model at a small radius fillet, or there may be slight ripples at a transition surface.

[0136] Next, based on the geometric characteristics (e.g., fillets, transition surfaces) and topological relationships of these minor inconsistencies, the local adjustment area of ​​the entire blade leading edge is divided into multiple sub-regions. For example, the fillet region is divided into one sub-region, and the adjacent transition surface is divided into another sub-region. Then, a local adjustment strategy is generated for each sub-region; for example, a higher-order B-spline surface reconstruction might be used for the fillet region, while a NURBS surface that emphasizes smoothness might be used for the transition surface.

[0137] Subsequently, within each sub-region, based on its specific local adjustment strategy, local parametric surface reconstruction is performed using high-confidence measurement data and geometric model topology information to generate a series of locally adjusted surface patches. For example, in the rounded corner sub-region, by adjusting the control points of the B-spline, the reconstructed surface patches are made to precisely fit the high-confidence measurement data while maintaining the geometric characteristics of the rounded corners.

[0138] After generating these surface patches, the system evaluates the smoothness of the connection between each local adjustment surface patch and the surrounding fitted surface at the sub-region boundary. If the connection between a surface patch and the fitted surface does not meet the preset smoothness threshold (e.g., poor curvature continuity), the system will automatically adjust the parameters of the reconstructed local parametric surface, such as fine-tuning the position of the control points or increasing the number of iterations, until the smoothness evaluation result meets the requirements.

[0139] Finally, all locally adjusted surface patches that meet the preset smoothing threshold are seamlessly stitched with the fitted surface. This allows for precise correction of critical geometric details at the blade leading edge. The reconstructed surface not only highly matches the high-confidence measurement data but also maintains consistency with the topology of the geometric model, while ensuring a smooth transition with surrounding surfaces. This significantly improves the accuracy and reliability of turbine blade geometry inspection.

[0140] This application further proposes the following steps for assessing the consistency between the high-confidence measurement data and the topological information of the geometric model at key geometric details, in order to identify minor inconsistencies between the two:

[0141] Obtain manufacturing tolerance range information and measurement uncertainty information at the key geometric details;

[0142] The high-confidence measurement data is projected onto the corresponding surface of the geometric model to obtain the projection point, and the spatial distance between the projection point and the original measurement point is calculated, as well as the angle between the surface normal direction of the geometric model at the projection point and the normal direction of the measurement point.

[0143] Based on the spatial distance and the included angle, combined with the manufacturing tolerance range information and the measurement uncertainty information, a multi-dimensional quantitative evaluation is performed on the minor inconsistencies to obtain multi-dimensional quantitative evaluation results, thereby distinguishing inconsistencies caused by measurement noise, minor deformations allowed by manufacturing tolerances, and defects in the actual geometric model; and

[0144] Based on the multi-dimensional quantitative evaluation results, the minor inconsistencies are identified, and each identified minor inconsistency is assigned an inconsistency type identifier.

[0145] Specifically, acquiring the manufacturing tolerance range information and measurement uncertainty information for the key geometric details means that, before conducting a consistency assessment, the system pre-loads or calculates the manufacturing tolerance range data and the inherent uncertainty data of the measuring equipment or method related to the key geometric details to be assessed. Manufacturing tolerance range information can be understood as the allowable dimensional and shape deviations of the workpiece during manufacturing, its purpose being to provide an acceptable benchmark for subsequent inconsistency assessments. Measurement uncertainty information reflects the accuracy limitations of the measurement system itself, its purpose being to help distinguish between true deviations and measurement errors.

[0146] The process of projecting the high-confidence measurement data onto the corresponding surface of the geometric model to obtain projection points, and calculating the spatial distance between the projection points and the original measurement points, as well as the angle between the surface normal direction of the geometric model at the projection point and the normal direction of the measurement point, can be understood as a geometric matching and deviation quantification process. High-confidence measurement data refers to measurement point cloud data with high reliability after preliminary screening and processing. The purpose of the projection operation is to find the point on the geometric model closest to each measurement point for direct geometric comparison. The spatial distance is used to quantify the normal or approximate normal deviation between the measurement point and the model surface, while the angle of the normal direction is used to quantify the deviation between the direction of the measurement point and the direction of the model surface. The aim is to comprehensively capture the differences between the measurement data and the geometric model from both positional and directional dimensions.

[0147] In practical applications, based on the spatial distance and the included angle, combined with the manufacturing tolerance range information and the measurement uncertainty information, a multi-dimensional quantitative evaluation of the minor inconsistencies is performed to obtain multi-dimensional quantitative evaluation results, distinguishing inconsistencies caused by measurement noise, minor deformations allowed by manufacturing tolerances, and defects in the actual geometric model. This is a comprehensive deviation analysis method. By comparing the spatial distance and the included angle with preset manufacturing tolerance range information and measurement uncertainty information, a multi-dimensional evaluation model can be established. For example, if the spatial distance and the included angle are both within the measurement uncertainty range, it may be attributed to measurement noise; if they exceed the measurement uncertainty but are still within the manufacturing tolerance range, they may belong to minor deformations allowed by manufacturing tolerances; if they exceed the manufacturing tolerance range, it may indicate a defect in the actual geometric model. The purpose is to accurately classify the sources of inconsistencies, providing an accurate basis for subsequent local adjustments.

[0148] Furthermore, based on the multi-dimensional quantitative evaluation results, the minor inconsistency regions are identified, and each identified minor inconsistency region is assigned an inconsistency type identifier. This means that after completing the quantitative evaluation, the system will cluster or divide the measurement data points into regions based on the evaluation results, thereby identifying continuous regions with similar inconsistency characteristics. The purpose of assigning an inconsistency type identifier (e.g., "noise region," "tolerance deformation region," or "defect region") to each region is to provide clear guidance for subsequent local adjustment strategies, ensuring that the most appropriate handling method is adopted for different types of inconsistencies.

[0149] This application's solution, by introducing manufacturing tolerance range information and measurement uncertainty information, and combining spatial distance and normal angle for multi-dimensional quantitative evaluation, enables a more refined analysis of subtle inconsistencies between high-confidence measurement data and the geometric model. It is precisely this multi-dimensional, multi-factor comprehensive evaluation that allows the system to accurately distinguish inconsistencies caused by measurement noise, minor deformations allowed by manufacturing tolerances, and defects in the actual geometric model. Traditional consistency assessments may only focus on the magnitude of geometric deviations, failing to effectively differentiate between these different sources of deviation. This could lead to unnecessary corrections for measurement noise or deformations within allowable tolerances, or failure to promptly detect and correct actual geometric defects. This solution, by accurately classifying inconsistencies, ensures the targetedness and effectiveness of subsequent local adjustment operations, avoiding the problems of blindly correcting or overlooking critical defects.

[0150] In some preferred embodiments, a specific example is given below. Assume the workpiece to be measured is an aero-engine blade with a complex curved surface and sharp edges. After performing coordinate measuring machine (CMM) measurements on the blade, high-confidence measurement data is obtained. To identify minor inconsistencies between the measurement data and the geometric model, the manufacturing tolerance range and measurement uncertainty information of key geometric details of the blade (e.g., the leading and trailing edges) are first obtained. For example, the manufacturing tolerance of the leading edge might be ±0.02 mm, and the measurement uncertainty might be ±0.005 mm.

[0151] Subsequently, the high-confidence measurement data points are projected onto the surface of the blade's geometric model to obtain projection points. For each measurement point, the spatial distance between it and the corresponding projection point, as well as the angle between the normal direction of the measurement point and the normal direction of the model surface, are calculated. For example, the spatial distance of measurement point A is 0.003 mm, and the normal angle is 0.5 degrees. The spatial distance of another measurement point B is 0.015 mm, and the normal angle is 2 degrees.

[0152] Next, based on these spatial distances and angles, combined with manufacturing tolerance information and measurement uncertainty information, a multi-dimensional quantitative evaluation is performed. For measurement point A, since its spatial distance of 0.003 mm and normal angle of 0.5 degrees are both within the measurement uncertainty range of ±0.005 mm and a small angle, it is evaluated as an inconsistency caused by measurement noise. For measurement point B, its spatial distance of 0.015 mm exceeds the measurement uncertainty, but is still within the manufacturing tolerance range of ±0.02 mm, and the normal angle of 2 degrees may also be within the allowable range, so it is evaluated as a small deformation allowed by the manufacturing tolerance. If there is a measurement point C whose spatial distance reaches 0.03 mm, far exceeding the manufacturing tolerance range, it will be evaluated as a true geometric model defect.

[0153] Ultimately, based on these multi-dimensional quantitative evaluation results, the system identifies regions with minor inconsistencies caused by measurement noise, regions with minor deformations allowed by manufacturing tolerances, and regions with defects in the actual geometric model, assigning a corresponding inconsistency type label to each region. For example, a region on the leading edge of the blade is labeled as a "tolerance deformation region," while a recessed area in the middle of the blade is labeled as a "defect region." This precise classification provides clear guidance for subsequent local adjustments; for example, "noise regions" may only require smoothing, "tolerance deformation regions" may require fitting within acceptable limits, while "defect regions" require more stringent corrections.

[0154] This application further proposes the steps of dividing the local adjustment region into multiple sub-regions based on the geometric features and topological relationships of the minor inconsistency regions, and generating a local adjustment strategy for each sub-region, including:

[0155] Analyze the geometric features and topological connections of the aforementioned minor inconsistencies;

[0156] The boundaries of the sub-regions are determined based on the geometric features and the topological connectivity.

[0157] Evaluate the geometric continuity at the boundaries of the sub-regions;

[0158] When there are potential discontinuities in the geometric continuity or conflicts in topological information, the division of the sub-regions is adjusted; and

[0159] Generate a local adjustment strategy for each adjusted sub-region.

[0160] Specifically, analyzing the geometric features and topological connections of minor inconsistencies involves conducting in-depth geometric analysis of the identified minor inconsistencies. This includes extracting geometric attributes such as curvature, normal direction, and boundary curves, and analyzing topological information such as their connection methods and adjacency relationships with other regions. This step aims to comprehensively understand the morphology and structure of inconsistencies. Determining the boundaries of sub-regions based on geometric features and topological connections can be understood as initially defining the scope of each sub-region based on the above analysis results. For example, initial boundaries can be determined based on regions with significant curvature changes, regions with altered topological connections, or preset geometric size thresholds. In practical applications, evaluating the geometric continuity at the boundaries of sub-regions specifically refers to checking whether the geometric attributes of adjacent sub-regions transition smoothly at the boundaries. For example, zero-order continuity (positional continuity), first-order continuity (tangential continuity), and second-order continuity (curvature continuity) at the boundaries can be evaluated. The purpose is to ensure the rationality of sub-region division and avoid sharp edges or unnatural transitions at the boundaries. When potential discontinuities exist in geometric continuity or topological information conflicts exist, the sub-region partitioning is adjusted. For example, iterative optimization algorithms can be used to eliminate discontinuities or resolve conflicts by fine-tuning boundary positions, changing the number of sub-regions, or merging / splitting existing sub-regions. The goal is to obtain high-quality sub-region partitioning, providing a solid foundation for subsequent local adjustments. Furthermore, generating local adjustment strategies for each adjusted sub-region refers to developing personalized adjustment schemes based on each sub-region's specific geometric characteristics, inconsistency type identifiers, and its importance in the overall geometric model. For example, low-order surface fitting can be used for regions with gentle curvature changes; while more refined parametric reconstruction methods may be needed for regions with critical geometric details.

[0161] Through the above technical solution, this application can significantly improve the quality and reliability of local adjustment region division. By rigorously evaluating and dynamically adjusting the geometric continuity of the sub-region boundaries, the overall accuracy and reliability of coordinate measuring machine (CMM) metrology and inspection are improved.

[0162] In some preferred embodiments, a specific example is given below. Suppose that after measuring a complex curved surface workpiece, a minor inconsistency is identified in a certain area, requiring local adjustment. This minor inconsistency area may contain an acute-angled edge and a smooth transition surface. First, the system analyzes the geometric features (such as the curvature of the edge, the normal direction of the plane) and topological connections (such as the connection between the edge and adjacent surfaces) of the minor inconsistency area. Then, based on these analysis results, the system initially determines the boundaries of the sub-regions. For example, the acute-angled edge area may be divided into one sub-region, and the smooth transition surface into another. Subsequently, the system evaluates the geometric continuity of these two sub-regions at the boundary. If a significant geometric discontinuity is found in the connection between the acute-angled edge and the smooth transition surface after the initial division (e.g., inconsistent tangent directions), or if there are conflicts in the topological information (e.g., the boundary segment is defined as one in the geometric model, but appears as multiple in the measurement data), the system will trigger an adjustment operation. Specifically, the system might fine-tune the boundary positions of sub-regions or insert a transition sub-region between sharp edges and smooth transition surfaces to ensure that the geometric continuity at the boundaries reaches a preset smoothness threshold. For example, continuity can be improved by introducing a small chamfer or rounded transition at the boundary. Finally, a corresponding local adjustment strategy is generated for each adjusted sub-region (e.g., sharp edge sub-region, transition sub-region, smooth transition surface sub-region). For example, for sharp edge sub-regions, a reconstruction algorithm that preserves sharp features might be used; for transition sub-regions, an algorithm that emphasizes smooth transitions might be used; and for smooth transition surface sub-regions, a conventional surface fitting algorithm might be used. In this way, the accuracy of local adjustments and the high quality of the overall surface are ensured.

[0163] The steps described above, in each sub-region, to perform local parametric surface reconstruction using the high-confidence measurement data and the geometric model topology information, to generate locally adjusted surface patches, include:

[0164] Obtain material property information of the sub-region, including the elastic modulus distribution and yield strength distribution of the sub-region;

[0165] Obtain manufacturing tolerance information for the sub-region, the manufacturing tolerance information including dimensional tolerances and shape tolerances of the sub-region;

[0166] Based on the material property information, determine the deformation sensitivity distribution of the sub-region under stress.

[0167] Based on the manufacturing tolerance information, determine the allowable geometric deformation range of the sub-region;

[0168] Based on the deformation sensitivity distribution and the geometric deformation range, the parameters of the local parametric surface reconstruction algorithm are adjusted, including control point weights, surface order, and the number of fitting iterations; and

[0169] Using the high-confidence measurement data and the topological information of the geometric model, local parametric surface reconstruction is performed to generate the local adjustment surface patch.

[0170] Specifically, the material property information refers to the inherent physical properties of the material constituting the sub-region. For example, the elastic modulus distribution describes how the material's ability to resist elastic deformation varies at different locations, while the yield strength distribution indicates the material's maximum stress-bearing capacity before permanent plastic deformation. This information can be obtained through material database queries, experimental testing, or finite element analysis, with the aim of providing basic data for subsequent deformation sensitivity analysis. The manufacturing tolerance information can be understood as the allowable deviation range of the geometric dimensions and shape of the sub-region during the workpiece manufacturing process. Dimensional tolerances define the maximum and minimum allowable values ​​for specific dimensions, while shape tolerances constrain the range of variation in geometric shapes such as flatness, roundness, and straightness. This information is typically derived from engineering drawings or design specifications, and its purpose is to define the acceptable variability of the workpiece during the manufacturing process.

[0171] In practical applications, the deformation sensitivity distribution of the sub-region under stress is determined based on the material property information. For example, regions with lower elastic modulus are more prone to deformation under the same load and are therefore considered to have higher deformation sensitivity. This distribution can be obtained through structural mechanics analysis or simulation, the purpose of which is to quantify the degree of response of different regions to external or internal stresses. Furthermore, the allowable geometric deformation range of the sub-region is determined based on the manufacturing tolerance information. For example, if the dimensional tolerance of a region is ±0.1 mm, then its allowable geometric deformation range is 0.2 mm. This range provides a physical boundary for surface reconstruction, aiming to ensure that the reconstructed surface is manufacturable.

[0172] Furthermore, based on the deformation sensitivity distribution and the geometric deformation range, the parameters of the local parametric surface reconstruction algorithm are adjusted. For example, for regions with high deformation sensitivity and a small allowable deformation range, the control point weights can be increased, the surface order can be improved, and the number of fitting iterations can be increased to ensure that the reconstructed surface can more accurately fit the high-confidence measurement data while strictly adhering to the geometric deformation range. The control point weights affect the control of the local shape of the surface, the surface order determines the complexity and smoothness of the surface, and the number of fitting iterations affects the accuracy and convergence of the fitting. The aim is to enable the reconstruction algorithm to adaptively handle the physical and manufacturing constraints of different regions. Thus, using the high-confidence measurement data and the geometric model topology information, local parametric surface reconstruction is performed to generate the locally adjusted surface patch.

[0173] In some preferred embodiments, a specific example is given below. Assume the workpiece to be tested is an aero-engine blade with a complex curved surface and thin-walled structure, where critical geometric details have extremely high requirements for material strength and manufacturing precision. After performing coordinate measuring machine (CMM) measurements on the blade, sub-regions requiring local adjustment are identified using the method described above. Specifically, firstly, material property information for this sub-region is obtained. For example, the blade material is a high-temperature alloy, and its elastic modulus and yield strength distribution may have slight differences in different locations. This data is obtained through material handbooks and finite element analysis. Simultaneously, manufacturing tolerance information for this sub-region is obtained. For example, the dimensional tolerance of the blade edge is ±0.05 mm, and the surface shape tolerance is 0.02 mm. Based on this information, the system determines the deformation sensitivity distribution of this sub-region under engine operating stress; for example, thin-walled regions have higher deformation sensitivity. Furthermore, combined with the manufacturing tolerance information, the allowable geometric deformation range for this sub-region is determined. Subsequently, during local parametric surface reconstruction, the parameters of the reconstruction algorithm are intelligently adjusted. For example, in blade edge regions with high deformation sensitivity and strict manufacturing tolerances, control point weights are set to higher values, the surface order is increased, and the number of fitting iterations is increased to ensure that the generated local adjustment surface patch can fit the high-confidence measurement data extremely accurately, while strictly adhering to the 0.05mm dimensional tolerance and 0.02mm shape tolerance, avoiding the generation of any geometric features that exceed these physical and manufacturing limitations. Finally, using the adjusted algorithm parameters, combined with the high-confidence measurement data and the topological information of the blade geometric model, a local parametric surface reconstruction is performed to generate a local adjustment surface patch that accurately reflects the measurement data and conforms to the material's physical properties and manufacturing tolerances. This surface patch is then seamlessly stitched with the fitted surface to obtain a highly accurate and physically realistic blade geometric model.

[0174] Secondly, referring to Figure 2This application further proposes a coordinate measuring machine (CMM) metrology and testing system, which includes:

[0175] The geometric model acquisition and analysis module 201 is used to acquire geometric model data of the workpiece to be tested and analyze the geometric model data to determine the local geometric features of the workpiece.

[0176] The path and posture generation module 202 is used to generate path information and probe posture information for measurement based on the local geometric features. The path information and the probe posture information are used to make the probe and the workpiece surface form a contact state close to the normal direction.

[0177] The contact simulation and evaluation module 203 is used to simulate the contact between the probe and the workpiece surface based on the path information and the probe posture information, so as to evaluate the probability that the probe contact point deviates from the preset contact conditions.

[0178] The multi-angle measurement instruction generation module 204 is used to generate an instruction to perform repeated multi-angle measurements on the contact point when the probability of deviating from the preset contact conditions is greater than a preset threshold.

[0179] The data acquisition module 205 is used to acquire the position information of the probe contact point, the probe posture information, and the waveform characteristics of the probe trigger signal when executing the instruction for multi-angle repetitive measurement;

[0180] The abnormal trigger judgment module 206 is used to determine whether there is an abnormal trigger signal based on the results of the multi-angle repeated measurement, the position information of the probe contact point, the probe posture information, and the waveform characteristics of the probe trigger signal.

[0181] The abnormal data removal module 207 is used to remove the measurement data corresponding to the abnormal trigger signal from the measurement dataset when an abnormal trigger signal exists.

[0182] The data processing module 208 is used to process the measurement data of the removed abnormal trigger signal according to the local geometric features, so as to eliminate the deviation in the measurement data and maintain the geometric information of the key geometric details of the workpiece.

[0183] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method of metrological detection of a coordinate measuring machine, characterized in that, Includes the following steps: Obtain the geometric model data of the workpiece to be tested, and analyze the geometric model data to determine the local geometric features of the workpiece; Based on the local geometric features, path information and probe posture information for measurement are generated. The path information and probe posture information are used to make the probe and the workpiece surface form a contact state close to the normal direction. Based on the path information and the probe posture information, the contact between the probe and the workpiece surface is simulated to assess the probability that the probe contact point deviates from the preset contact conditions. When the probability of deviating from the preset contact condition is greater than the preset threshold, an instruction is generated to perform repeated measurements at multiple angles on the probe contact point. When executing instructions for multi-angle repetitive measurements, the position information of the probe contact point, the probe attitude information, and the waveform characteristics of the probe trigger signal are collected. Based on the results of the multi-angle repeated measurements, the position information of the probe contact point, the probe posture information, and the waveform characteristics of the probe trigger signal, it is determined whether there is an abnormal trigger signal. When an abnormal trigger signal is present, the measurement data corresponding to the abnormal trigger signal is removed from the measurement dataset; Furthermore, an adaptive data processing algorithm based on local geometric features is employed to process the measurement data after the abnormal trigger signals have been removed, in order to eliminate the deviation in the measurement data. Based on the geometric model data of the workpiece and the confidence level of the measurement data, surface fitting is performed on the measurement data to reconstruct the geometric features of the workpiece.

2. The method of claim 1, wherein, The step of generating an instruction to perform repeated multi-angle measurements on the probe contact point when the probability of deviating from the preset contact condition is greater than a preset threshold includes: Identify the geometric features at the probe contact point to obtain geometric feature information at the probe contact point; Based on the geometric feature information, a first probe posture combination is generated; Identify the type of instability at the probe contact point to obtain information about the type of instability at the probe contact point; Based on the unstable condition type information, adjust the first probe posture combination to obtain the second probe posture combination; and Based on the second probe posture combination, a multi-angle repeatable measurement command is generated.

3. The method of claim 1, wherein, The steps of employing an adaptive data processing algorithm based on local geometric features to process the measurement data after the abnormal trigger signals have been removed to eliminate deviations in the measurement data, and performing surface fitting on the measurement data based on the geometric model data of the workpiece and the confidence level of the measurement data to reconstruct the geometric features of the workpiece, include: Acquire information on the force change experienced by the probe when it contacts the surface of the workpiece; Based on the geometric model of the workpiece and the material properties of the workpiece, determine the normal contact force curve of the probe; Compare the force change information with the normal contact force curve to determine whether there is a risk of stable displacement at the probe contact point; If the aforementioned risk of instability exists, adjust the contact dynamics parameters of the probe to perform a second measurement; Obtain the probe contact point position information and force change information obtained from the second measurement; By comparing the position information of the probe contact point obtained from the original measurement point with that obtained from the remeasurement, and combining the force change information obtained from the remeasurement, it is determined whether there is a stable offset at the original measurement point, and the measurement data with stable offset is obtained. The measurement data containing the stable offset are removed from the measurement data of the removed abnormal trigger signals; Based on the geometric model data of the workpiece and the confidence level of the measurement data, surface fitting is performed on the measurement data to reconstruct the geometric features of the workpiece.

4. A method of metrological detection of a coordinate measuring machine according to claim 3, characterized in that, The step of performing surface fitting on the measurement data based on the geometric model data of the workpiece and the confidence level of the measurement data to reconstruct the geometric features of the workpiece includes: The geometric model data of the workpiece is analyzed to identify areas in the geometric model data that differ from the micro-geometric features allowed by the workpiece manufacturing tolerances; A geometric deformation range is set for the region of difference in microscopic geometric features; When performing surface fitting, based on the confidence level of the measured data, within the allowable range of geometric deformation, the constraint strength of the geometric model data in the region of microscopic geometric feature difference is adjusted to obtain the fitted surface; Monitor the local deviations between the fitted surface and the geometric model at key geometric details; When the deviation of the fitted surface from the high-confidence measurement data at the key geometric details meets a deviation threshold, or when the difference between the geometric features of the fitted surface at the key geometric details and the geometric model meets specific conditions, a detail correction operation is triggered; and Using the high-confidence measurement data and combining it with the topological information of the geometric model at the key geometric details, the fitted surface is locally adjusted to obtain a local adjustment region, so that the geometric information of the key geometric details is consistent with the measurement data and the geometric model.

5. A method of metrological testing of a coordinate measuring machine according to claim 4, characterized in that, The step of triggering the detail correction operation when the deviation of the fitted surface from the high-confidence measurement data at the key geometric details meets a deviation threshold, or when the difference between the geometric features of the fitted surface at the key geometric details and the geometric model meets a specific condition, includes: Obtain manufacturing tolerance and material property information for the critical geometric detail areas to be corrected; Based on the manufacturing tolerance information and the material property information, adjust the deviation threshold or specific conditions that trigger the detail correction operation; Based on the deviation threshold or specific conditions of the adjusted trigger detail correction operation, determine whether the deviation between the fitted surface and the high-confidence measurement data at the key geometric details meets the deviation threshold, or whether the difference between the geometric features of the fitted surface at the key geometric details and the geometric model meets specific conditions; and When the judgment result shows that the deviation meets the deviation threshold or the difference meets specific conditions, the detailed correction operation is triggered.

6. The coordinate measuring machine metrology and testing method according to claim 4, characterized in that, The step of using the high-confidence measurement data, combined with the topological information of the geometric model at the key geometric details, to locally adjust the fitted surface to obtain a local adjustment region, so that the geometric information of the key geometric details is consistent with the measurement data and the geometric model, includes: The consistency between the high-confidence measurement data and the topological information of the geometric model at the key geometric details is evaluated to identify minor inconsistencies between the two. Based on the geometric features and topological relationships of the minor inconsistency regions, the local adjustment region is divided into multiple sub-regions, and a local adjustment strategy is generated for each sub-region. Within each sub-region, based on the local adjustment strategy, local parametric surface reconstruction is performed using the high-confidence measurement data and geometric model topology information to generate local adjustment surface patches. The smoothness of the connection between the generated local adjustment surface patch and the fitted surface at the boundary of the sub-region is evaluated to obtain the smoothness evaluation result; When the smoothness evaluation result does not meet the preset smoothness threshold, the parameters of the locally parameterized surface reconstruction are adjusted; and The local adjustment surface patch whose smoothness evaluation result meets the preset smoothness threshold is seamlessly spliced ​​with the fitted surface.

7. A method of metrological detection of a coordinate measuring machine according to claim 6, characterized in that, The step of evaluating the consistency between the high-confidence measurement data and the topological information of the geometric model at the key geometric details, in order to identify minor inconsistencies between the two, includes: Obtain manufacturing tolerance range information and measurement uncertainty information at the key geometric details; The high-confidence measurement data is projected onto the corresponding surface of the geometric model to obtain the projection point, and the spatial distance between the projection point and the original measurement point is calculated, as well as the angle between the surface normal direction of the geometric model at the projection point and the normal direction of the measurement point. Based on the spatial distance and the included angle, combined with the manufacturing tolerance range information and the measurement uncertainty information, a multi-dimensional quantitative evaluation is performed on the minor inconsistencies to obtain multi-dimensional quantitative evaluation results, thereby distinguishing inconsistencies caused by measurement noise, minor deformations allowed by manufacturing tolerances, and defects in the actual geometric model; and Based on the multi-dimensional quantitative evaluation results, the minor inconsistency regions are identified, and each identified minor inconsistency region is assigned an inconsistency type identifier.

8. A coordinate measuring machine metrology and testing method according to claim 6, characterized in that, The step of dividing the local adjustment region into multiple sub-regions based on the geometric features and topological relationships of the minor inconsistency regions, and generating a local adjustment strategy for each sub-region, includes: analyzing the geometric features and topological connectivity of the minor inconsistency regions; The boundaries of the sub-regions are determined based on the geometric features and the topological connectivity. Evaluate the geometric continuity at the boundaries of the sub-regions; When there are potential discontinuities in the geometric continuity or conflicts in topological information, the division of the sub-regions is adjusted; and Generate a local adjustment strategy for each adjusted sub-region.

9. The method of claim 6, wherein, The step of reconstructing a local parametric surface within each sub-region, based on the local adjustment strategy and utilizing the high-confidence measurement data and the geometric model topology information, to generate a locally adjusted surface patch includes: Obtain material property information of the sub-region, including the elastic modulus distribution and yield strength distribution of the sub-region; Obtain manufacturing tolerance information for the sub-region, the manufacturing tolerance information including dimensional tolerances and shape tolerances of the sub-region; Based on the material property information, determine the deformation sensitivity distribution of the sub-region under stress. Based on the manufacturing tolerance information, determine the allowable geometric deformation range of the sub-region; Based on the deformation sensitivity distribution and the geometric deformation range, the parameters of the local parametric surface reconstruction algorithm are adjusted, including control point weights, surface order, and the number of fitting iterations; and Using the high-confidence measurement data and the topological information of the geometric model, local parametric surface reconstruction is performed to generate the local adjustment surface patch.

10. A coordinate measuring machine metrology inspection system, characterized by, The system includes: The geometric model acquisition and analysis module is used to acquire geometric model data of the workpiece to be tested and analyze the geometric model data to determine the local geometric features of the workpiece. The path and posture generation module is used to generate path information and probe posture information for measurement based on the local geometric features. The path information and the probe posture information are used to make the probe and the workpiece surface form a contact state close to the normal direction. The contact simulation and evaluation module is used to simulate the contact between the probe and the workpiece surface based on the path information and the probe posture information, so as to evaluate the probability that the probe contact point deviates from the preset contact conditions. A multi-angle measurement instruction generation module is used to generate an instruction to perform repeated multi-angle measurements on the contact point when the probability of deviating from the preset contact conditions is greater than a preset threshold. The data acquisition module is used to acquire the position information of the probe contact point, the probe attitude information, and the waveform characteristics of the probe trigger signal when executing instructions for multi-angle repetitive measurement. The abnormal trigger judgment module is used to determine whether there is an abnormal trigger signal based on the results of the multi-angle repeated measurement, the position information of the probe contact point, the probe posture information, and the waveform characteristics of the probe trigger signal. An abnormal data removal module is used to remove the measurement data corresponding to an abnormal trigger signal from the measurement dataset when an abnormal trigger signal is present; and The data processing module is used to process the measurement data after the abnormal trigger signals have been removed using an adaptive data processing algorithm based on local geometric features, so as to eliminate the deviation in the measurement data, and to perform surface fitting on the measurement data according to the geometric model data of the workpiece and the confidence level of the measurement data, so as to reconstruct the geometric features of the workpiece.

Images