A method for intelligent geometric error identification and adaptive compensation based on three-coordinate detection

Abstract

The application discloses an intelligent geometric error identification and self-adaptive compensation method based on three-coordinate detection, and relates to the technical field of three-coordinate measurement. The method comprises the following steps: setting a detection track, controlling a probe to perform forward and reverse scanning respectively, collecting actual space coordinates and displacement instructions of each sampling point, and forming original scanning data; performing differential processing on the actual coordinates of the sampling points corresponding to the forward and reverse scanning, identifying and removing a dynamic offset component according to a scanning speed variation rate, and obtaining a net differential component sequence; extracting geometric errors based on the space position of the detection track and mapping the geometric errors into a full-track compensation vector sequence along the continuous detection track, performing compensation and re-measurement after smoothing processing, and performing self-adaptive iterative updating according to residual errors. The method solves the technical problems that the prior art cannot accurately identify geometric errors in combination with the motion characteristics of the scanning process and the compensation continuity is poor, and improves the precision and stability of three-coordinate detection.

Description

Technical Field

[0001] This application relates to the field of coordinate measuring technology, and in particular to an intelligent geometric error identification and adaptive compensation method based on coordinate measuring machine detection. Background Technology

[0002] Coordinate measuring machine (CMM) systems, as an important means of precision geometric measurement, are widely used in aerospace, precision manufacturing, automotive parts, and high-end equipment manufacturing to achieve the measurement of the dimensions, shape, and positional accuracy of complex workpieces. With the ever-increasing precision requirements of the measured objects, various geometric and dynamic errors introduced by CMM systems during actual operation have become key factors restricting further improvements in measurement accuracy.

[0003] During operation, existing coordinate measuring machines (CMMs) experience deviations between the actual spatial trajectory and the ideal commanded trajectory of their probes as the probes move along the detection path. These deviations are influenced by mechanical structural errors, assembly errors, guide rail straightness errors, attitude errors, and the dynamic characteristics of the drive and control systems. These deviations are not only related to the inherent static geometric errors of the equipment but also closely related to dynamic factors such as speed changes, acceleration / deceleration, and direction switching during the probe's movement, and are particularly pronounced in high-speed scanning or complex trajectory detection scenarios.

[0004] To improve detection accuracy, existing technologies typically establish geometric error models and compensate for them during the detection process to correct equipment operating errors. However, these methods mostly focus on modeling static geometric errors and do not adequately consider dynamic offset errors caused by speed changes, acceleration / deceleration, and direction switching during the scanning process, making it difficult to achieve high-precision compensation on the actual detection trajectory.

[0005] Furthermore, existing error compensation methods mostly focus on discrete measurement points, lacking a continuous compensation mechanism along the complete detection trajectory. This makes the compensation results prone to local abrupt changes, affecting detection stability and repeatability. Therefore, how to achieve refined error compensation along the complete detection trajectory while ensuring compensation continuity remains a pressing problem in the field of coordinate measuring machine (CMM) inspection.

[0006] In summary, existing technologies struggle to effectively distinguish between geometric errors and dynamic offset errors caused by velocity variations during coordinate measuring machine (CMM) inspection. Furthermore, they lack refined error identification based on the spatial position of the detection trajectory and a continuous compensation and adaptive update mechanism for the complete detection trajectory. Therefore, there is an urgent need for a CMM method that can accurately identify geometric errors by incorporating the motion characteristics of the scanning process and achieve continuous, adaptive compensation along the detection trajectory, thereby further improving detection accuracy and stability. Summary of the Invention

[0007] To address the aforementioned issues, this invention proposes an intelligent geometric error identification and adaptive compensation method based on three-coordinate detection. By combining forward and reverse scanning with velocity change characteristics, dynamic offset errors are identified and eliminated. Geometric errors are segmented based on the spatial position of the detection trajectory, and a continuous compensation vector sequence is constructed along the complete detection trajectory. Combined with retest feedback, the compensation parameters are adaptively updated.

[0008] To achieve the above objectives, this invention provides an intelligent geometric error identification and adaptive compensation method based on coordinate measuring machine (CMM) detection, comprising the following steps: Set the detection trajectory, control the probe to perform forward and reverse scanning along the detection trajectory, collect the actual spatial coordinates and displacement commands of each sampling point in the two scanning directions, and form the original scanning data; Based on the original scanning data, the corresponding sampling points of the forward and reverse scanning are differentially processed to obtain a differential error sequence. According to the rate of change of the speed of the probe during the scanning process along the detection trajectory, the dynamic offset component in the differential error sequence is identified and removed to obtain a net difference component sequence. The detection trajectory is divided into several spatial segments according to spatial location. Within each spatial segment, the coordinate deviation between the actual spatial coordinates of the sampling point and the command coordinates corresponding to its displacement command is calculated. Based on the net difference component sequence, the coordinate deviation between forward and reverse scanning is corrected for directional consistency. Geometric error is extracted from the corrected coordinate deviation. The detection trajectory is discretized into several continuous micro-displacement units with a preset step size. Based on the geometric error of the sampling points associated with the position of each micro-displacement unit in the spatial segment to which it belongs, the compensation vector corresponding to each micro-displacement unit is determined. The compensation vectors are then arranged in the order of the detection trajectory to obtain the full trajectory compensation vector sequence. After smoothing the full trajectory compensation vector sequence, it is loaded into a three-coordinate measuring machine. Geometric error compensation is performed on the detected trajectory and the measurement is repeated to obtain the residual error sequence. The full trajectory compensation vector sequence is then updated based on the residual error sequence. This process is repeated until convergence is achieved, thus realizing adaptive iterative compensation.

[0009] The technical solution provided in this invention has at least the following technical effects or advantages: By combining forward and reverse scanning, and utilizing the velocity variation characteristics of the probe along the detection trajectory for differential error processing, dynamic offset errors are effectively identified and eliminated, resulting in purer geometric error information. This avoids misjudging dynamic errors as geometric errors and improves the accuracy of geometric error identification. A spatial segmentation mechanism based on the spatial position of the detection trajectory is introduced. Within each spatial segment, the coordinate deviation is corrected for directional consistency and decomposed into components, allowing geometric errors to be extracted according to different physical meanings such as tangential, attitude, and local distortion, enhancing the specificity and physical interpretability of error identification. The detection trajectory is discretized into continuous small displacement units, and a full trajectory compensation vector sequence is constructed based on the geometric errors within the spatial segments, ensuring that the compensation result covers the entire detection trajectory. After the compensation vector sequence is generated, continuity constraints are introduced for smoothing, ensuring that compensation changes between adjacent positions are controlled by the command resolution capability and dynamic characteristics of the detection equipment. This improves the stability of the compensation execution process and reduces the risk of introducing new dynamic errors due to abrupt compensation changes. By employing an adaptive iterative compensation mechanism based on retesting and residual error feedback, the full-trajectory compensation vector sequence is updated and optimized. This allows the compensation parameters to adaptively adjust according to changes in equipment operating status and detection conditions, further improving the long-term stability and detection accuracy of the coordinate measuring machine (CMM). Compared with existing technologies, this invention accurately identifies geometric errors by combining the motion characteristics of the scanning process, realizing a CMM method with continuous and adaptive compensation along the detection trajectory, thus improving detection accuracy and stability.

[0010] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart illustrating an intelligent geometric error identification and adaptive compensation method based on coordinate measuring machine detection, provided in an embodiment of this application. Detailed Implementation

[0013] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0014] Example 1, as Figure 1 As shown, this application provides an intelligent geometric error identification and adaptive compensation method based on coordinate measuring machine (CMM) detection, wherein the method includes: S1: Set the detection trajectory, control the probe to perform forward and reverse scanning along the detection trajectory, collect the actual spatial coordinates and displacement commands of each sampling point in the two scanning directions, and form the original scanning data; Specifically, firstly, an appropriate trajectory type is selected based on the geometry of the object being inspected. For example, a straight-line scan is used for measuring objects with relatively regular surfaces, a curve scan is used for curved or complex-shaped objects, or a spiral scan is used for circular or spiral-shaped objects. Next, a global coordinate system is established in three-dimensional space to describe the position of the object. Using the working reference point of the coordinate measuring machine as the origin, the coordinate axis directions are determined based on the placement of the object in the measurement space and its geometric reference characteristics, ensuring that the coordinate system orientation aligns with the main geometric features or measurement reference of the object. The inspection trajectory is then set within the global coordinate system, covering all key parts of the object. The trajectory setting is based on a comprehensive consideration of factors such as the surface shape of the object, scanning requirements, and measurement accuracy. The probe's movement range and scanning speed should also be considered to ensure the probe can successfully complete the entire scan. Sampling points are then set, with the spacing and distribution density determined based on the geometric characteristics of the object and the measurement requirements. Finally, the probe is controlled to perform forward and reverse scans along the inspection trajectory. Forward scanning refers to the probe scanning from the starting point to the ending point along the inspection trajectory, while reverse scanning is the scanning process from the ending point to the starting point. At each sampling point, the probe acquires its actual spatial coordinates and records its actual displacement command. The displacement command is a target position command sent by the control system of the coordinate measuring machine to the probe according to the preset detection trajectory, indicating the theoretical spatial position that the probe should reach at each sampling point. Finally, the data from all sampling points in both forward and reverse scans are integrated to form the raw scan data.

[0015] S2: Based on the original scanning data, differential processing is performed on the corresponding sampling points of forward and reverse scanning to obtain a differential error sequence. According to the speed change rate of the probe during the scanning process along the detection trajectory, the dynamic offset component in the differential error sequence is identified and removed to obtain a net difference component sequence. Specifically, based on the acquired raw scan data, a differential error sequence is obtained according to the spatial coordinate differences of corresponding sampling points in forward and reverse scans. Then, based on the speed changes during probe scanning, the impact of speed changes on the error is analyzed to identify dynamic offset errors, which are temporary errors caused by speed changes during scanning. In this process, by analyzing the correlation between the differential error and the rate of speed change, dynamic offset errors are eliminated, thus obtaining the net difference component sequence.

[0016] Furthermore, the acquisition of the net difference component sequence includes: Based on the original scan data, the actual spatial coordinates of the corresponding sampling points of the forward and reverse scans are differentially calculated to obtain the differential error sequence. Based on the rate of change of the probe's speed along the detection trajectory, the scanning process is divided into different segments, including acceleration segment, constant speed segment and deceleration segment; Within each segment, the differential error sequence is segmented and statistically analyzed. The correlation between the differential error of a sampling point and its corresponding rate of change is calculated. Based on the correlation, the differential error sequence is filtered for velocity correlation. When the correlation of a sampling point exceeds a preset threshold, the corresponding differential error is determined to be a dynamic offset component. The dynamic offset component is removed from the differential error sequence to obtain the net difference component sequence.

[0017] Furthermore, the method for calculating the correlation degree is as follows: The correlation degree is calculated according to the following formula: ; Where i is the sampling point index, Let be the correlation between the differential error at the i-th sampling point and the rate of change of velocity. This represents the trend of the differential error sequence within the segment containing the i-th sampling point. Let be the rate of change of velocity corresponding to the i-th sampling point. The variation amplitude of the differential error within the segment containing the i-th sampling point. For speed weighting index, This is the trend smoothing coefficient. The coupling coefficient is denoted as ; wherein the trend quantity is obtained by linearly fitting the differential error sequence within the segment, and the fluctuation amplitude is obtained by calculating the standard deviation of the differential error within the segment. The settings are based on the dynamic characteristics of the coordinate measuring machine and the rate of change of the probe scanning speed.

[0018] Specifically, firstly, based on the original scanning data, the actual spatial coordinates of the corresponding sampling points under forward and reverse scanning are obtained, and differential calculation is performed to obtain the differential error sequence, which reflects the error caused by the movement of the probe and its dynamic characteristics during the scanning process; Because the velocity change characteristics of the probe vary under different motion states during scanning, and the impact of the velocity change rate on the differential error varies, the scanning process is divided into segments based on the velocity change rate during probe scanning, including acceleration, constant velocity, and deceleration segments. This segmentation uses a threshold segmentation method, determined based on the overall trend of the velocity change rate within continuous sampling points. Specifically, the segmentation is based on the velocity change rate between adjacent sampling points during probe scanning along the detection trajectory. When the overall velocity change rate within continuous sampling points exceeds a preset acceleration threshold, it is determined to be an acceleration segment; when the velocity change rate is close to zero and its absolute value is lower than a preset constant velocity threshold, it is determined to be a constant velocity segment; when the overall velocity change rate shows a negative trend and its absolute value is lower than a preset deceleration threshold, it is determined to be a deceleration segment. These thresholds can be set according to the dynamic characteristics of the coordinate measuring machine and the probe, such as the probe's acceleration and deceleration capabilities, equipment response time, equipment calibration results, and historical data, and adjusted in conjunction with the target detection accuracy requirements. Preferably, the acceleration threshold and deceleration threshold are 0.2 m / s², and the constant velocity threshold is 0.01 m / s; After completing the segmentation of the scanning process, the differential error sequence is statistically analyzed within each segment. The correlation between the differential error at each sampling point and the rate of change of velocity at the corresponding sampling point is calculated using the following formula: Where i is the sampling point index; The correlation between the differential error at the i-th sampling point and the rate of change of velocity; The trend of the differential error sequence within the segment containing the i-th sampling point is obtained by linear fitting of the differential error sequence within the segment, that is, by fitting a straight line (or an approximate linear trend) using the least squares method. The slope of this line is the trend, which reflects the changing trend of the differential error within a certain segment as the scanning process progresses. The velocity change rate corresponding to the i-th sampling point; The fluctuation range of the differential error within the segment containing the i-th sampling point is obtained by calculating the standard deviation of the differential error within the segment, reflecting the dispersion of the data, i.e., the fluctuation range of the differential error within that segment. All parameters are preset based on the dynamic characteristics of the coordinate measuring machine and the rate of change of the scanning speed of the probe along the detection trajectory. This is a velocity weighting index used to control the impact of the velocity change rate on the differential error. It is adjusted according to the velocity change amplitude in different scanning segments (acceleration, constant speed, deceleration). Values ​​can be set differently depending on the speed. For example, in acceleration sections where speed changes are significant, the differential error is highly sensitive to speed variations and can be set to a larger value (e.g., 0.8) to enhance its response to speed changes. In uniform speed sections where speed changes are minimal or constant, the differential error is less affected by speed variations and can be set to a smaller value (e.g., 0.2). In deceleration sections where speed changes are large in magnitude but short in duration, the impact on the differential error is relatively limited, therefore... Set to a moderate value (e.g., 0.5) to account for changes in differential error during deceleration, but not to overreact to speed changes. This is the trend smoothing coefficient, used to smooth the trend of the differential error. It is set based on the equipment's response characteristics and the fluctuation of the differential error. If the differential error fluctuates significantly or the equipment itself has strong dynamic characteristics, then... Set to a larger value to smooth out errors; if the equipment is highly stable and the differential error is small, then a smaller value can be set. This is the coupling coefficient, used to determine the degree of coupling between the rate of change of speed and the differential error. It is set based on the rate of change of scanning speed. During high-speed scanning, the rate of change of speed is large, and the change in the differential error is more dependent on the speed. The value is relatively large; the differential error is less sensitive to speed changes during low-speed scanning, therefore... The value is set to a small value. Subsequently, the differential error sequence is filtered for velocity correlation based on the calculated correlation degree. If the correlation degree of a sampling point exceeds the preset correlation degree threshold, the differential error corresponding to that sampling point is determined to be a dynamic offset component. The correlation degree threshold is preset before detection based on equipment calibration or historical measurement data. Specifically, the correlation degree is calculated from historical detection data obtained by the coordinate measuring machine under normal scanning conditions, and the statistical distribution of the correlation degree is analyzed to determine the typical range of correlation degree values ​​under conditions without significant dynamic offset. The correlation degree threshold is then set as the upper limit of this typical range for subsequent detection to identify dynamic offset components. The identified dynamic offset components are removed from the differential error sequence to obtain the net difference component sequence, which is the pure geometric error after removing the dynamic error caused by velocity changes.

[0019] S3: Divide the detection trajectory into several spatial segments according to spatial location, calculate the coordinate deviation between the actual spatial coordinates of the sampling point and the command coordinates corresponding to its displacement command in each spatial segment, and perform directional consistency correction on the coordinate deviation between forward and reverse scanning based on the net difference component sequence, and extract the geometric error from the corrected coordinate deviation. Specifically, the detection trajectory is divided into several spatial segments according to spatial location, and the coordinate deviation between the actual spatial coordinates and the displacement command coordinates of the sampling points within each spatial segment is calculated. Next, the coordinate deviations of the forward and reverse scans are corrected for directional consistency using a net difference component sequence, eliminating error differences caused by different scanning directions, allowing the coordinate deviations of the forward and reverse scans to be processed uniformly within the same coordinate system. After correction, geometric errors are extracted, including tangential error components, attitude error components, and local distortion error components.

[0020] Furthermore, the extraction of the geometric error includes: The detection trajectory is divided into several spatial segments according to spatial location, and the coordinate deviation between the actual spatial coordinates of the sampling point and the command coordinates corresponding to its displacement command is calculated in each spatial segment. Based on the net difference component sequence, the coordinate deviation between the forward and reverse scans is corrected for directional consistency to obtain the coordinate deviation after directional correction. Within each spatial segment, the coordinate deviation after direction correction is divided into components according to spatial direction, and the geometric error is extracted, including: Tangential error component, wherein the tangential error component is the projection component of the coordinate deviation after direction correction on the tangential direction of the detection trajectory; The attitude error component is a trend component obtained by low-frequency filtering or fitting of the sequence formed by the projection components of the coordinate deviation after direction correction onto the normal direction of the detection trajectory. The local distortion error component is the residual component obtained by removing the tangential error component and the attitude error component from the coordinate deviation after direction correction.

[0021] Specifically, firstly, based on the shape of the inspection trajectory and the distribution of sampling points, the detection trajectory is divided into several spatial segments according to its spatial location. For example, if the detection trajectory is a curve, it can be divided into several equally spaced segments according to the density of sampling points or the curvature of the trajectory. Then, within each spatial segment, the actual spatial coordinates of each sampling point are compared with its corresponding displacement command coordinates, and the coordinate deviation between them is calculated. These deviations reflect the difference between the actual measured position and the target position (command position) of each sampling point. Because the probe may be affected by environmental factors and inertia during the scanning process, resulting in non-symmetrical deviations between forward and reverse scans, after obtaining the coordinate deviations of each sampling point, directional consistency correction is performed on the coordinate deviations of the forward and reverse scans based on the net difference component sequence. Specifically, within the same spatial segment, the forward and reverse coordinate deviations are compared point-to-point to determine the sampling point index corresponding to the spatial position in the forward and reverse scans. Then, the net difference component corresponding to the sampling point index is used as the correction value in the net difference component sequence to correct the forward and reverse coordinate deviations of that sampling point, thereby offsetting the directional offset between the forward and reverse scans, ensuring data consistency between the forward and reverse scans, eliminating possible errors and deviations during the scanning process, and obtaining the directional corrected coordinate deviations. Next, within each spatial segment, based on the coordinate deviation after direction correction, the space is divided into components according to spatial direction, and the geometric error is extracted, including: The tangential error component is usually related to the geometric characteristics of the device's motion trajectory and detection path. Therefore, the projection component obtained by projecting the coordinate deviation after direction correction along the tangential direction of the detection trajectory (the direction of movement along the detection trajectory) is the tangential error component. The attitude error component refers to the trend component obtained by projecting the coordinate deviation after orientation correction onto the normal direction of the detection trajectory, and then processing the sequence of projection components according to the detection trajectory through low-frequency filtering or fitting. This processing removes short-term noise and focuses on the long-term trend of the error, used to characterize the continuous offset caused by changes in the attitude of the equipment or probe.

[0022] The normal direction is perpendicular to the tangent of the detection trajectory, so this attitude error component reflects the error of the device in the vertical direction; Local distortion error components refer to the residual components obtained after removing the tangential and attitude error components from the coordinate deviations after orientation correction. The residual components reflect the errors remaining after removing known error sources (tangential and attitude errors), which are usually caused by local distortion or nonlinear factors.

[0023] Through the above steps, the geometric error of the detection trajectory can be accurately decomposed, extracting three error components: tangential error, attitude error, and local distortion error. Each error component reflects different sources of error during the measurement process, which helps to further analyze and compensate for measurement errors, ensuring the high precision and stability of the coordinate measuring system and providing a foundation for subsequent compensation and error correction.

[0024] S4: Discretize the detection trajectory into several continuous micro-displacement units according to a preset step size. Based on the geometric error of the sampling points associated with the position of each micro-displacement unit in the spatial segment to which it belongs, determine the compensation vector corresponding to each micro-displacement unit. Arrange the compensation vectors in the order of the detection trajectory to obtain the full trajectory compensation vector sequence. Furthermore, obtaining the full trajectory compensation vector sequence includes: The detection trajectory is discretized into several continuous micro-displacement units according to a preset step size; For each micro-displacement unit, the spatial segment to which it belongs on the detection trajectory is determined, and the geometric error of the sampling point associated with the spatial position of the micro-displacement unit within the spatial segment is called to determine the compensation vector corresponding to the micro-displacement unit. The compensation vector includes tangential error compensation vector components, attitude error compensation vector components, and local distortion error compensation vector components. Wherein, when the position of the micro-displacement unit coincides with that of a sampling point, the geometric error of that sampling point is the compensation vector; When the small displacement unit is located between adjacent sampling points, the geometric error of the adjacent sampling points is interpolated to obtain the compensation vector; When the micro-displacement unit covers multiple sampling points, the geometric error of the multiple sampling points is weighted averaged or fitted to obtain the compensation vector. The compensation vectors corresponding to each small displacement unit are arranged in the order of the detected trajectory to obtain the full trajectory compensation vector sequence.

[0025] Specifically, the detection trajectory is first discretized into several continuous small displacement units according to a preset step size. By dividing the entire detection trajectory into these small displacement units, every detail on the detection trajectory can be processed more accurately so that the compensation can be processed in a fine-grained manner in the future. Next, for each micro-displacement unit, its spatial segment is determined based on its position on the detection trajectory. Then, based on the geometric errors of the sampling points associated with the micro-displacement unit's position within that spatial segment, a compensation vector corresponding to that micro-displacement unit is determined. Each micro-displacement unit corresponds to one compensation vector, and the components of this compensation vector correspond to the components of the previously extracted geometric errors. Specifically, the tangential error compensation vector component corresponds to the tangential error component, with its direction opposite to the measured tangential error; that is, the compensation vector is used to cancel the original error. The attitude error compensation vector component corresponds to the attitude error component, with its direction opposite to the measured normal trend error. The local distortion error compensation vector component corresponds to the local distortion error component, with its direction opposite to the local residual error. Specifically, if the position of a small displacement element coincides with that of a sampling point within its spatial segment, the geometric error of that sampling point is directly used as the compensation vector for the small displacement element. If a small displacement element is located between two sampling points within its spatial segment, the compensation vector of the small displacement element is obtained by interpolating the geometric errors of the two sampling points. If a small displacement element covers multiple sampling points within its spatial segment, the geometric errors of these sampling points are weighted averaged or fitted to obtain the compensation vector of the small displacement element. The compensation vectors corresponding to each minute displacement unit obtained through the above steps are arranged in the order of the detected trajectory to form a full trajectory compensation vector sequence. These compensation vectors reflect the geometric error of the entire trajectory and provide basic data for subsequent smoothing processing.

[0026] S5: After smoothing the full trajectory compensation vector sequence, load it into the coordinate measuring machine, perform geometric error compensation on the detected trajectory and retest, obtain the residual error sequence, and update the full trajectory compensation vector sequence according to the residual error sequence. Repeat the process until convergence to achieve adaptive iterative compensation.

[0027] Specifically, the entire trajectory compensation vector sequence is first smoothed to reduce potential abrupt changes or discontinuities, resulting in a more stable compensation outcome across the entire trajectory. Next, the smoothed full-trajectory compensation vector sequence is loaded into a coordinate measuring machine (CMM) for geometric error compensation, and a residual error sequence is obtained through retesting. The full-trajectory compensation vector sequence is then updated based on the spatial distribution of the residual error sequence. This process employs an adaptive iterative compensation mechanism, allowing the compensation parameters to be optimized and adjusted according to the actual measurement conditions and scanning parameters until the residual error reaches a preset limit, thereby ensuring the operational accuracy and long-term stability of the CMM.

[0028] Further, smoothing the full trajectory compensation vector sequence includes: Based on the compensation vectors corresponding to adjacent small displacement elements, the rate of change of each compensation vector component is calculated. Based on the resolution of the displacement command of the coordinate measuring machine and the sampling fluctuation characteristics of the compensation vector components on adjacent micro-displacement units, a continuity constraint is established, which limits the maximum allowable variation of the compensation vector components between adjacent micro-displacement units. Under the continuity constraint, filtering or fitting processing is performed on each compensation vector component sequence in the full trajectory compensation vector sequence to obtain a smoothed full trajectory compensation vector sequence.

[0029] Furthermore, the adaptive iterative compensation includes: The smoothed full trajectory compensation vector sequence is loaded into the coordinate measuring machine, geometric error compensation is performed on the detection trajectory and the retest is carried out to obtain the residual error sequence. Based on the spatial distribution of the residual error sequence on the detection trajectory, the compensation vector of the corresponding small displacement unit is corrected to generate an updated full trajectory compensation vector sequence. The geometric error compensation and retest are performed repeatedly until the compensation meets the convergence condition and the iteration is terminated. The convergence condition is that the maximum value of the residual error sequence is less than a preset limit. The final converged full trajectory compensation vector sequence is stored as the operating error compensation parameters of the coordinate measuring machine, and is used for subsequent measurements of the same detection trajectory.

[0030] Specifically, after obtaining the full trajectory compensation vector sequence, to avoid unreasonable abrupt changes in the compensation vectors between adjacent micro-displacement units and to maintain the spatial continuity of the compensation vectors on the detection trajectory, the full trajectory compensation vector sequence is further smoothed. First, based on the compensation vectors of adjacent micro-displacement units, the rate of change of each compensation vector component is calculated to characterize the local change amplitude of the compensation vector in the detection trajectory direction. The rate of change of each compensation vector component includes the rate of change of the tangential error compensation vector component, the rate of change of the attitude error compensation vector component, and the rate of change of the local distortion error compensation vector component. Next, based on the resolution capability of the coordinate measuring machine's displacement commands and the sampling fluctuation characteristics of the compensation vector components on adjacent small displacement units, continuity constraints are established for each compensation vector component. The displacement command resolution capability characterizes the smallest displacement change that the machine can distinguish at the command level, and its value can be determined by the minimum displacement command increment of the coordinate measuring machine. The sampling fluctuation characteristics of the compensation vector components reflect random fluctuations caused by measurement noise, environmental disturbances, and system micro-vibrations, and their scale can be characterized by the statistical dispersion of the corresponding compensation vector component sequence. The continuity constraints are as follows: ;in, To compensate for the rate of change of vector components between adjacent small displacement units, res is the minimum resolvable increment of the displacement command of the three-coordinate detection device. To detect the distance of the walking distance from the trajectory, To correspond to the standard deviation of the sampling fluctuation of the compensation vector component on adjacent micro-displacement units, k is a proportionality coefficient. k is determined by repeatedly scanning and calibrating the coordinate measuring machine under preset scanning conditions and based on the statistical relationship between the rate of change of the compensation vector component and its sampling fluctuation standard deviation. Through this continuity constraint, the maximum allowable variation range of the compensation vector component between adjacent micro-displacement units is limited, ensuring that the variation of the compensation vector component is confined within the resolution capability of the equipment and its sampling stability range, thus providing constraints for subsequent smoothing processing. Under the premise of satisfying the continuity constraint, filtering (such as removing high-frequency noise through low-pass filtering) or curve fitting (such as fitting using the least squares method) is used to smooth each component sequence of the full trajectory compensation vector sequence to remove abnormal fluctuations and improve the compensation effect. Finally, the smoothed tangential error compensation vector component sequence, attitude error compensation vector component sequence, and local distortion error compensation vector component sequence are combined to form the smoothed full trajectory compensation vector sequence.

[0031] After obtaining the smoothed full-trajectory compensation vector sequence, this sequence is loaded into the error compensation module of the coordinate measuring machine (CMM). During the scanning of the detection trajectory, the CMM applies this compensation vector sequence to the actual movement of the probe, causing it to shift accordingly in each direction when executing displacement commands. This corrects the actual trajectory to a near-ideal trajectory close to the commanded coordinates, thus compensating for geometric errors. In this way, the measurement points of each minute displacement unit can be corrected, ensuring that the probe's actual position on the entire detection trajectory remains consistent with the commanded position, thereby guaranteeing detection accuracy. After the compensation operation is completed, the probe is controlled to perform a second scan along the same detection trajectory. The compensated detection results are collected, and based on the difference between the actual spatial coordinates obtained from the re-measurement and the commanded coordinates of the corresponding displacement command, a residual error sequence is obtained. This residual error sequence reflects the distribution of incompletely eliminated geometric errors still existing on the detection trajectory under the current compensation vector. Subsequently, based on the spatial distribution of the residual error sequence on the detection trajectory, the residual error is matched with each small displacement unit obtained by discretizing the detection trajectory, and the compensation vector of the corresponding small displacement unit is corrected to take into account the deviation of the residual error in the spatial direction, thereby forming an updated full trajectory compensation vector sequence. Based on the above, the geometric error compensation and retesting process is repeated, and the residual error sequence is evaluated after each round of retesting until the compensation meets the convergence condition. When the maximum residual error in the residual error sequence is less than a preset limit, the compensation process is deemed to have reached the convergence condition. The preset limit is determined based on the inherent measurement accuracy of the coordinate measuring machine. The iteration terminates when the maximum value of the residual error is less than the error level that the machine can distinguish. The error level that the machine can distinguish can be selected as any one of the nominal measurement uncertainty, repeatability error, or measurement noise amplitude, and can be set as a predetermined proportion (between 0.1 and 1) of its corresponding value, thereby ensuring that the residual error below the limit will not have a distinguishable or meaningful impact on the measurement results. Once the convergence condition is met, the final converged full trajectory compensation vector sequence is stored as the operating error compensation parameter of the coordinate measuring machine. This parameter is then used to call upon subsequent detection trajectories with the same spatial distribution to perform geometric error compensation control.

[0032] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0033] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for intelligent geometric error identification and adaptive compensation based on coordinate measuring machine (CMM) detection, characterized in that, include: S1: Set the detection trajectory, control the probe to perform forward and reverse scanning along the detection trajectory, collect the actual spatial coordinates and displacement commands of each sampling point in the two scanning directions, and form the original scanning data; S2: Based on the original scanning data, differential processing is performed on the corresponding sampling points of forward and reverse scanning to obtain a differential error sequence. According to the speed change rate of the probe during the scanning process along the detection trajectory, the dynamic offset component in the differential error sequence is identified and removed to obtain a net difference component sequence. S3: Divide the detection trajectory into several spatial segments according to spatial location, calculate the coordinate deviation between the actual spatial coordinates of the sampling point and the command coordinates corresponding to its displacement command in each spatial segment, and perform directional consistency correction on the coordinate deviation between forward and reverse scanning based on the net difference component sequence, and extract the geometric error from the corrected coordinate deviation. S4: Discretize the detection trajectory into several continuous micro-displacement units according to a preset step size. Based on the geometric error of the sampling points associated with the position of each micro-displacement unit in the spatial segment to which it belongs, determine the compensation vector corresponding to each micro-displacement unit. Arrange the compensation vectors in the order of the detection trajectory to obtain the full trajectory compensation vector sequence. S5: After smoothing the full trajectory compensation vector sequence, load it into the coordinate measuring machine, perform geometric error compensation on the detected trajectory and retest, obtain the residual error sequence, and update the full trajectory compensation vector sequence according to the residual error sequence. Repeat the process until convergence to achieve adaptive iterative compensation.

2. The intelligent geometric error identification and adaptive compensation method as described in claim 1, characterized in that, The acquisition of the net difference component sequence includes: Based on the original scan data, the actual spatial coordinates of the corresponding sampling points of the forward and reverse scans are differentially calculated to obtain the differential error sequence. Based on the rate of change of the probe's speed along the detection trajectory, the scanning process is divided into different segments, including acceleration segment, constant speed segment and deceleration segment; Within each segment, the differential error sequence is segmented and statistically analyzed. The correlation between the differential error of a sampling point and its corresponding rate of change is calculated. Based on the correlation, the differential error sequence is filtered for velocity correlation. When the correlation of a sampling point exceeds a preset threshold, the corresponding differential error is determined to be a dynamic offset component. The dynamic offset component is removed from the differential error sequence to obtain the net difference component sequence.

3. The intelligent geometric error identification and adaptive compensation method as described in claim 2, characterized in that, The method for calculating the correlation degree is as follows: The correlation degree is calculated according to the following formula: ； Where i is the sampling point index, Let be the correlation between the differential error at the i-th sampling point and the rate of change of velocity. This represents the trend of the differential error sequence within the segment containing the i-th sampling point. Let be the rate of change of velocity corresponding to the i-th sampling point. The variation amplitude of the differential error within the segment containing the i-th sampling point. For speed weighting index, This is the trend smoothing coefficient. The coupling coefficient is denoted as ; wherein the trend quantity is obtained by linearly fitting the differential error sequence within the segment, and the fluctuation amplitude is obtained by calculating the standard deviation of the differential error within the segment. The settings are based on the dynamic characteristics of the coordinate measuring machine and the rate of change of the probe scanning speed.

4. The intelligent geometric error identification and adaptive compensation method as described in claim 3, characterized in that, The extraction of the geometric error includes: The detection trajectory is divided into several spatial segments according to spatial location, and the coordinate deviation between the actual spatial coordinates of the sampling point and the command coordinates corresponding to its displacement command is calculated in each spatial segment. Based on the net difference component sequence, the coordinate deviation between the forward and reverse scans is corrected for directional consistency to obtain the coordinate deviation after directional correction. Within each spatial segment, the coordinate deviation after direction correction is divided into components according to spatial direction, and the geometric error is extracted, including: Tangential error component, wherein the tangential error component is the projection component of the coordinate deviation after direction correction on the tangential direction of the detection trajectory; The attitude error component is a trend component obtained by low-frequency filtering or fitting of the sequence formed by the projection components of the coordinate deviation after direction correction onto the normal direction of the detection trajectory. The local distortion error component is the residual component obtained by removing the tangential error component and the attitude error component from the coordinate deviation after direction correction.

5. The intelligent geometric error identification and adaptive compensation method as described in claim 4, characterized in that, The acquisition of the full trajectory compensation vector sequence includes: The detection trajectory is discretized into several continuous micro-displacement units according to a preset step size; For each micro-displacement unit, the spatial segment to which it belongs on the detection trajectory is determined, and the geometric error of the sampling point associated with the spatial position of the micro-displacement unit within the spatial segment is called to determine the compensation vector corresponding to the micro-displacement unit. The compensation vector includes tangential error compensation vector components, attitude error compensation vector components, and local distortion error compensation vector components. Wherein, when the position of the micro-displacement unit coincides with that of a sampling point, the geometric error of that sampling point is the compensation vector; When the small displacement unit is located between adjacent sampling points, the geometric error of the adjacent sampling points is interpolated to obtain the compensation vector; When the micro-displacement unit covers multiple sampling points, the geometric error of the multiple sampling points is weighted averaged or fitted to obtain the compensation vector. The compensation vectors corresponding to each small displacement unit are arranged in the order of the detected trajectory to obtain the full trajectory compensation vector sequence.

6. The intelligent geometric error identification and adaptive compensation method as described in claim 5, characterized in that, Smoothing the full trajectory compensation vector sequence includes: Based on the compensation vectors corresponding to adjacent small displacement elements, the rate of change of each compensation vector component is calculated. Based on the resolution of the displacement command of the coordinate measuring machine and the sampling fluctuation characteristics of the compensation vector components on adjacent micro-displacement units, a continuity constraint is established, which limits the maximum allowable variation of the compensation vector components between adjacent micro-displacement units. Under the continuity constraint, filtering or fitting processing is performed on each compensation vector component sequence in the full trajectory compensation vector sequence to obtain a smoothed full trajectory compensation vector sequence.

7. The intelligent geometric error identification and adaptive compensation method as described in claim 6, characterized in that, The adaptive iterative compensation includes: The smoothed full trajectory compensation vector sequence is loaded into the coordinate measuring machine, geometric error compensation is performed on the detection trajectory and the retest is carried out to obtain the residual error sequence. Based on the spatial distribution of the residual error sequence on the detection trajectory, the compensation vector of the corresponding small displacement unit is corrected to generate an updated full trajectory compensation vector sequence. The geometric error compensation and retest are performed repeatedly until the compensation meets the convergence condition and the iteration is terminated. The convergence condition is that the maximum value of the residual error sequence is less than a preset limit. The final converged full trajectory compensation vector sequence is stored as the operating error compensation parameters of the coordinate measuring machine, and is used for subsequent measurements of the same detection trajectory.

Images