Stability Evaluation Method and Device for Cover Plate Pressing

Abstract

The present application provides a method and device for evaluating the stability of cover plate pressing, which relates to the technical field of cover plate pressing. The method includes: loading a calibrated test workpiece onto a target fixture and fixing it with a cover plate; controlling the machine tool to start and perform machining; reading monitoring data in real time and normalizing the contact two-dimensional standard force field into a time-series grayscale image; establishing a feature recognition result for the time-series grayscale image; synchronously obtaining the acquisition information of a triaxial accelerometer arranged on the fixture and using the micro-vibration signal as the first verification signal; monitoring the key points of the workpiece and using the time-series displacement of the workpiece as the second verification signal; performing stability verification on the feature recognition result and outputting a stability verification evaluation result. Through the present application, the technical objectives of multi-dimensional fusion modeling and stability intelligent evaluation based on the workpiece structure characteristics and the machining dynamic process can be achieved, and the technical effects of improving the recognition accuracy of pressing stability, reducing the misjudgment rate, and enhancing the adaptability of the system to complex machining environments can be achieved.

Description

Technical Field

[0001] This application relates to the technical field of cover plate pressing, and particularly to a method and device for evaluating the stability of cover plate pressing. Background Art

[0002] In the process of precision manufacturing and high-reliability assembly, the stable pressing state of workpieces during the processing or assembly stage is directly related to the quality and performance of the final product. Especially in the machining process of complex structural parts or non-standard customized parts, if there are phenomena such as uneven pressing, loosening or instantaneous slipping in the pressing state, it is extremely easy to cause machining errors, surface defects and even equipment damage. Therefore, effective and real-time evaluation of the stability of cover plate pressing has become an important link to ensure machining reliability.

[0003] Currently, in the machining of complex parts in the prior art, the structural characteristics of workpieces are closely related to the force distribution. Traditional pressing evaluation methods cannot accurately model the force changes in different geometric regions, and also lack the ability to construct differential thresholds starting from the workpiece body structure. At the same time, most methods fail to achieve multi-source information fusion and only rely on single-channel feedback signals, resulting in large fluctuations and high misjudgment rates in the stability determination results in a dynamic machining environment, and it is difficult to meet the requirements of high-precision manufacturing scenarios for robustness and real-time performance.

[0004] In summary, there are technical problems in the prior art that due to the lack of refined force modeling and differential determination mechanisms based on workpiece structural characteristics, it is impossible to accurately reflect the true force state when facing different geometric regions, further affecting the accuracy and reliability of the pressing stability evaluation, and thus it is difficult to meet the requirements of real-time performance and robustness in high-precision machining scenarios. Summary of the Invention

[0005] The purpose of this application is to provide a method and device for evaluating the stability of cover plate pressing, so as to solve the technical problems in the prior art that due to the lack of refined force modeling and differential determination mechanisms based on workpiece structural characteristics, it is impossible to accurately reflect the true force state when facing different geometric regions, further affecting the accuracy and reliability of the pressing stability evaluation, and thus it is difficult to meet the requirements of real-time performance and robustness in high-precision machining scenarios.

[0006] In view of the above problems, this application provides a method and device for evaluating the stability of cover plate pressing.

[0007] In a first aspect, the present application provides a method for evaluating the stability of cover plate pressing, which is implemented by a device for evaluating the stability of cover plate pressing, including: after setting a sensor array on the contact surface of the calibration test workpiece, loading the calibration test workpiece onto the target fixture and fixing it with the cover plate; after confirming the completion of the cover plate assembly, activating the sensor array and controlling the machine tool to start and perform machining; reading the monitoring data of the sensor array in real time, establishing a contact two-dimensional standard force field, and normalizing the contact two-dimensional standard force field into a time-series grayscale image; identifying the characteristics of the center offset of the hot zone, the boundary pressing jump characteristic, the local maximum clustering characteristic, the integrity characteristic of the grayscale connected domain, and the symmetry imbalance characteristic of the time-series grayscale image, and establishing a characteristic recognition result; synchronously acquiring the acquisition information of the triaxial accelerometer set on the fixture to establish a micro-vibration signal, and using the micro-vibration signal as the first verification signal; using a laser displacement sensor to monitor the key points of the workpiece to establish a time-series displacement of the workpiece, and using the time-series displacement of the workpiece as the second verification signal; performing stability verification on the characteristic recognition result through the first verification signal and the second verification signal, and outputting a stability verification evaluation result.

[0008] In a second aspect, the present application further provides a device for evaluating the stability of cover plate pressing, which is used to execute the method for evaluating the stability of cover plate pressing as described in the first aspect, including: a fixing module, which is used to load the calibration test workpiece onto the target fixture and fix it with the cover plate after setting a sensor array on the contact surface of the calibration test workpiece; a processing module, which is used to activate the sensor array and control the machine tool to start and perform machining after confirming the completion of the cover plate assembly; a reading module, which is used to read the monitoring data of the sensor array in real time, establish a contact two-dimensional standard force field, and normalize the contact two-dimensional standard force field into a time-series grayscale image; a feature recognition module, which is used to identify the characteristics of the center offset of the hot zone, the boundary pressing jump characteristic, the local maximum clustering characteristic, the integrity characteristic of the grayscale connected domain, and the symmetry imbalance characteristic of the time-series grayscale image, and establish a characteristic recognition result; a signal establishment module, which is used to synchronously acquire the acquisition information of the triaxial accelerometer set on the fixture to establish a micro-vibration signal, and use the micro-vibration signal as the first verification signal; a monitoring module, which is used to use a laser displacement sensor to monitor the key points of the workpiece to establish a time-series displacement of the workpiece, and use the time-series displacement of the workpiece as the second verification signal; a verification module, which is used to perform stability verification on the characteristic recognition result through the first verification signal and the second verification signal, and output a stability verification evaluation result.

[0009] The technical solution provided in the present application has at least the following technical effects or advantages: by achieving the technical goal of multi-dimensional fusion modeling and stability intelligent evaluation based on the workpiece structure characteristics and the machining dynamic process, the technical effects of improving the recognition accuracy of pressing stability, reducing the misjudgment rate, and enhancing the adaptability of the system to complex machining environments are achieved.

[0010] The above description is only an overview of the technical solution of this application. In order to understand the technical means of this application more clearly, it can be implemented in accordance with the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the following specific embodiments of this application are specifically given. It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of this application, nor is it used to limit the scope of this application. Other features of this application will become easily understood through the following description. Brief Description of the Drawings

[0011] In order to more clearly illustrate the technical solutions in this application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only exemplary. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0012] Figure 1 It is a schematic flow chart of the method for evaluating the stability of the cover plate pressing in this application;

[0013] Figure 2 It is a schematic structural diagram of the device for evaluating the stability of the cover plate pressing in this application.

[0014] Description of the reference numerals: fixing module 11, processing module 12, reading module 13, feature recognition module 14, signal establishment module 15, monitoring module 16, verification module 17. Detailed Embodiments

[0015] By providing a method and device for evaluating the stability of cover plate pressing, this application solves the technical problems in the prior art that due to the lack of refined force modeling and differential determination mechanism based on the structural characteristics of the workpiece, it is impossible to accurately reflect the real stress state when facing different geometric regions, further affecting the accuracy and reliability of the evaluation of the pressing stability, and thus it is difficult to meet the requirements of real-time and robustness in high-precision machining scenarios. It realizes the technical goal of multi-dimensional fusion modeling and stability intelligent evaluation based on the structural characteristics of the workpiece and the dynamic process of machining, and achieves the technical effects of improving the recognition accuracy of the pressing stability, reducing the misjudgment rate, and enhancing the adaptability of the system to complex machining environments.

[0016] Next, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited by the example embodiments described herein. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application. Additionally, it should be noted that for the convenience of description, only the parts related to the present application are shown in the drawings rather than all of them.

[0017] Embodiment 1. Please refer to the appended Figure 1 , the present application provides a method for evaluating the stability of cover plate pressing, which is applied to a device for evaluating the stability of cover plate pressing, and specifically includes the following steps:

[0018] S1: After setting up a sensor array on the contact surface of the calibration test workpiece, load the calibration test workpiece into the target fixture and fix it with a cover plate.

[0019] Specifically, the purpose of setting up a sensor array on the contact surface of the calibration workpiece is to obtain distributed pressure information during the pressing process. The sensor array refers to multiple sensors arranged in a certain rule on the surface where the workpiece contacts the fixture. Each sensor can independently collect the pressure data at its location, thus forming a pressure distribution map on the entire surface. For example, if 20×10 pressure sensors are arranged on a steel plate with a length and width of 200 mm×100 mm, real-time pressure data of 100 measurement points can be obtained at an interval of 10 mm. The calibration test workpiece is a workpiece with a standard structure and known material properties, which is used to verify the accuracy and consistency of the system. Load the calibration test workpiece into the target fixture. The target fixture is a device used to position and fix the workpiece, and its structure is designed according to the geometric shape of the workpiece to ensure the stability of the workpiece during the assembly and processing. After loading, fix it with a cover plate. The cover plate is a rigid structure, and its function is to apply a pre-pressure from above to press the workpiece tightly on the fixture to prevent the workpiece from shifting or vibrating during subsequent processing. In this way, a test environment including the complete pressing contact process can be established. The spatial pressure distribution information provided by the sensor array, combined with the overall pressing effect of the cover plate, can collect data under simulated real processing conditions, laying a foundation for subsequent image conversion, feature recognition, and stability analysis.

[0020] S2: After confirming that the cover plate assembly is completed, activate the sensor array and control the machine tool to start and perform machining.

[0021] Specifically, after confirming that the cover plate assembly is completed, it means that the preparatory action for pressing the workpiece has ended, and the contact state between the workpiece and the fixture is stable. At this time, the entire structure is ready for data acquisition and processing. If the cover plate is not properly tightened, the data collected subsequently will be distorted due to workpiece displacement or loosening. Immediately afterwards, activating the sensor array means turning on multiple pressure sensors arranged on the contact surface of the workpiece, enabling them to start working and collect pressure data in real time. The sensors are strain-type or resistive pressure sensors, which can convert pressure changes into electrical signals, thereby recording the force conditions at each measurement point. Activating the sensors changes from the standby state to the working state, preparing to receive the dynamic pressure changes generated during the machining process, and ensuring that data is recorded synchronously from the moment the workpiece starts to be stressed.

[0022] Subsequently, controlling the machine tool to start means making the CNC machine tool start its machining task through the numerical control system or manual control. The machine tool is a device used for operations such as cutting, drilling, and milling of metals or other materials. Starting the machine tool represents that the entire system enters the actual machining stage. During the operation of the CNC machine tool, various mechanical loads will be applied to the workpiece, such as cutting force, feed force, etc. The transmission of external forces will be fed back to the sensors through the fixture and the cover plate, thereby affecting the collected pressure data. Executing machining means that the machine tool performs specific cutting or other operations on the workpiece according to a preset program. The force on the workpiece is constantly changing, and the sensor array can record the dynamic pressure distribution map formed by these changes in real time.

[0023] S3: Read the monitoring data of the sensor array in real time, establish a two-dimensional scalar force field of contact, and normalize the two-dimensional scalar force field of contact into a temporal grayscale image.

[0024] Specifically, reading the monitoring data of the sensor array in real time means continuously collecting data from multiple sensors arranged on the contact surface between the workpiece and the fixture during the machining process. Each sensor records the pressure value at a measurement point, and high-time-resolution mechanical information can be obtained through periodic sampling.

[0025] Subsequently, combine the pressure values of all measurement points to establish a two-dimensional scalar force field of contact. Two-dimensional means that the data is mapped onto a plane corresponding to the length and width of the workpiece contact surface. The scalar force field means that each coordinate point corresponds to a pressure value, without including direction information, only representing the magnitude. By pairing the spatial coordinates of each sensor with its corresponding pressure value, a force field similar to a heat map can be generated on the plane.

[0026] Next, the two-dimensional scalar force field is normalized into a temporal grayscale image. Normalization means converting the original pressure values of different magnitudes into a standard range to represent the grayscale levels of the image pixels. The higher the pressure value, the greater the grayscale value, that is, the closer the image is to white; the lower the pressure value, the smaller the grayscale, and the closer the image is to black. The conversion is repeated at consecutive time points to form a sequence of images evolving over time, namely the temporal grayscale image, which can intuitively reflect the trend of pressure change over time.

[0027] S4: Identify the characteristics of the center offset of the hot zone, the boundary holding jump characteristic, the local maximum clustering characteristic, the integrity characteristic of the grayscale connected domain, and the symmetry imbalance characteristic in the temporal grayscale image, and establish the feature recognition result.

[0028] Specifically, identify the regions with higher brightness in each frame of the image, which represent the parts where the pressure is concentrated or the holding is tight. By calculating the centroid position and comparing it with the ideal geometric center, the degree of offset is obtained. The center offset of the hot zone reflects whether the applied pressure is uniform. If the center deviates greatly, it indicates that there is pressure skew or unilateral force. Subsequently, identify the boundary holding jump characteristic by analyzing the grayscale change gradient in the edge region of the image. Within the set boundary region width range, extract the grayscale values of this region for each frame of the image, and calculate the change rate of the grayscale mean value between adjacent regions. If there is a sudden change, it indicates that the edge force is discontinuous or the pressure jumps. Next, identify the local maximum clustering characteristic, mainly detecting whether multiple independent high-grayscale points in the image are concentratedly distributed. The local maximum refers to the position where the pixel has the highest grayscale value in its surrounding neighborhood. If the maxima are highly concentrated, it indicates local overload at the pressure application point, which may cause local material deformation. Then, after binarizing the image, analyze whether the white regions representing the presence of pressure are coherent to perform the integrity characteristic recognition of the grayscale connected domain and evaluate whether the overall force is continuous and complete. If there are many broken or fractured connected domains, it indicates that there are breakpoints on the holding surface or the holding area is interrupted. Finally, identify the symmetry imbalance characteristic. For each frame of the image, perform horizontal and vertical flips respectively to obtain the mirror image, and calculate the absolute difference of the grayscale at each pixel position between the mirror image and the original image. If the difference is large, it indicates that the grayscale distribution of the image is asymmetric left and right or up and down, that is, the holding is uneven.

[0029] S5: Synchronously obtain the acquisition information of the triaxial accelerometer set on the fixture, establish a micro-vibration signal, and use the micro-vibration signal as the first verification signal.

[0030] Specifically, during the machining process, the data output by the triaxial accelerometer mounted on the fixture is read simultaneously. The triaxial accelerometer is a sensor capable of detecting acceleration changes in three directions, namely the X-axis, Y-axis, and Z-axis, and is used to monitor the dynamic response or minute vibrations of the mechanical structure in three dimensions. The collected information refers to the acceleration values recorded by the sensor at each moment, which is used to capture high-frequency mechanical changes. Based on the collected data, a micro-vibration signal is established, and the variation law of the acceleration value over time is extracted, thereby forming a time-series signal representing the minute vibrations of the fixture during operation. The micro-vibration signal contains low-amplitude, high-frequency vibration changes, reflecting the dynamic stability during the machining process. For example, during the machining process, the fixture exhibits periodic vibrations with an amplitude of 0.02 meters per second squared and a frequency of 100 hertz in the Z-axis direction. The micro-vibration signal is used as the primary basis for determining the stability of the machining process to verify whether there are abnormalities in the clamping state or the force on the workpiece. If the intensity of the micro-vibration significantly increases during a certain period, it may indicate unstable clamping force or abnormal force transmission between the machine tool tool and the workpiece.

[0031] S6: Use a laser displacement sensor to monitor key points of the workpiece, establish the workpiece time-series displacement, and use the workpiece time-series displacement as the second verification signal.

[0032] Specifically, through the laser displacement sensor installed in the machining area, the position changes of some representative parts on the workpiece surface during the machining process are measured in real time. The laser displacement sensor is a high-precision non-contact measurement device that can record the distance between the target surface and the sensor with a resolution at the micron or even nanometer level. The key points of the workpiece refer to the areas on the workpiece where the force or deformation is more obvious and where displacement changes are likely to occur, such as the corners, center hole positions, or structural mutation points, etc. Monitoring these positions can more directly reflect the actual state of force and stability. The time series formed by arranging the position data continuously measured by the laser displacement sensor in chronological order is used to establish the workpiece time-series displacement, which is used to describe the displacement trend of the key points during the entire machining process and reflects the minute movement changes of the workpiece due to reasons such as machine tool movement, uneven force, or unstable clamping. Using the workpiece time-series displacement as the second verification signal, it serves as the second layer of reference basis for the analysis of the clamping stability during the machining process. By comparing the displacement data, it can be determined whether the workpiece has abnormal movement due to uneven pressure or unreasonable structural support.

[0033] S7: Perform stability verification on the feature recognition result through the first verification signal and the second verification signal, and output the stability verification evaluation result.

[0034] Specifically, two signal channels with different but complementary sources are used to detect the authenticity and consistency of the feature recognition results obtained from image analysis. The first verification signal refers to the micro-vibration signal collected by the triaxial accelerometer on the fixture, which can reflect the small mechanical vibrations generated due to uneven force or structural looseness during the machining or clamping process. The second verification signal is the displacement information of the key points of the workpiece measured by the laser displacement sensor, which is used to judge the deformation degree or displacement amplitude of the workpiece during the machining or clamping process. The feature recognition result is the discrimination information of the clamping or stress state obtained through the image processing process. By comparing the two signals with the image feature results, it can be determined whether the state of image recognition truly reflects the mechanical changes in the physical process, and the stability verification evaluation result is output to characterize whether the clamping state under the current working condition is reliable. If the first verification signal shows that the vibration is very small, and the second verification signal indicates that the displacement of the key points is also within the normal range, and the image features also show that the clamping force distribution is uniform, then the determination of "stable" can be output. On the contrary, if any one of them shows an abnormality, such as severe micro-vibration or sudden increase in displacement change, and there is a difference from the results such as image grayscale or symmetry, it is determined as "unstable" or "needs adjustment".

[0035] Furthermore, this application also includes: obtaining the zero-point grayscale image corresponding to the time sequence zero point; calculating the grayscale mean value of the zero-point grayscale image, and using the grayscale mean value to perform traversal deviation comparison on the zero-point grayscale image to generate a zero-point pressure deviation map; performing feature extraction on the calibrated test workpiece, and establishing a zero-point feature recognition result using the workpiece feature extraction result and the zero-point pressure deviation map.

[0036] Specifically, during the workpiece machining process, select a moment as the reference starting point, that is, the time point when the pressure has not started or just after clamping but before the machining action is started, and use the clamped grayscale image collected at this moment as the zero-point image. The grayscale image means that each pixel point expresses the stress intensity through the grayscale value, and the grayscale value usually varies between 0 and 255. The higher the value, the greater the pressure.

[0037] Subsequently, calculate the grayscale mean value of the zero-point grayscale image, that is, average the grayscale values of all pixel points in the image to reflect the basic stress level of the entire image at the zero point. Then, traverse each pixel point in the image, compare its grayscale value with the grayscale mean value, and obtain the deviation value, that is, the difference between the actual grayscale of a certain pixel and the overall mean value. Finally, generate a new image called the zero-point pressure deviation map, which shows the areas with uneven stress distribution in the zero-point state. For example, if the pixels on the left side of the image are 20 lower than the average value and the right side is 15 higher, it means that there is a left-right imbalance in the clamping structure at the zero starting point.

[0038] Next, perform feature extraction on the calibrated test workpiece, that is, identify the structural features of the workpiece in terms of geometric structure, shape contour, symmetry, etc. The extracted information includes boundary coordinates, center point position, and key structure distribution. The workpiece features provide a basis for interpreting the meaning of each region in the image. For example, whether a region with a concentrated gray-scale deviation exactly falls on the weak structure of the workpiece. Finally, combine and analyze these workpiece features with the zero-pressure deviation map to establish a set of zero-point feature recognition results for judging whether there are structural abnormalities or uneven clamping in the initial state of the workpiece during the initial clamping. Table 1 shows the record of the most recent zero-point clamping feature recognition.

[0039] Table 1: Record of the Most Recent Zero-Point Clamping Feature Recognition

[0040] Pixel coordinates (X, Y) Zero gray value Average gray value Deviation value = gray value - average value Match the characteristic area of the workpiece Zero pressure deviation state (10,15) 123 130 -7 Yes Slightly on the low side (22,33) 142 130 +12 No Ignore (45,18) 129 130 -1 Yes Normal (50,50) 117 130 -13 Yes Significantly on the low side (64,72) 134 130 +4 Yes Normal (78,29) 111 130 -19 Yes Seriously on the low side

[0041] Furthermore, this application also includes: performing control information interaction of the machine tool system to obtain processing control information; obtaining the tool feed rate mapped to the tool path according to the processing control information, compensating for the influence of the tool feed rate based on the workpiece feature extraction result and the tool path, and establishing the actual feed rate; using the actual feed rate to perform machining fitting of the calibrated test workpiece fixed by the pressure plate to generate a time-series compensation force field; normalizing the time-series compensation force field and then correcting the time-series gray-scale image, and establishing a feature recognition result based on the corrected time-series gray-scale image.

[0042] Specifically, read the processing parameters in real time from the machine tool control unit through the numerical control system or industrial bus, such as processing instructions, start signals, spindle speed, feed rate, etc. The control information includes the movement path, speed, and feed rate of each tool cut that the tool should execute during the processing. Based on the control information, the tool feed rate corresponding to the actual tool path can be further extracted. The tool path is the path that the tool moves in space, such as moving 50 millimeters along a straight line or cutting along an arc with a radius of 20 millimeters. The feed rate refers to the distance that the tool advances along a given direction each time it cuts, which may be a fixed value such as 0.1 millimeter each time. However, the structure of the workpiece is not always solid. For example, there are cavities or grooves at certain positions. Even if the tool performs a feed action, no material is actually cut. Therefore, it is necessary to compensate for the influence on the original feed rate by combining the workpiece feature information extracted previously, and then calculate the actual feed rate, that is, the position and amplitude of the actual cutting force applied by the tool in the effective material area.

[0043] After obtaining the true feed rate, the force state of the calibrated test workpiece fixed by the pressing plate during the machining process can be further fitted to establish a compensation force field corresponding to the machining time sequence. The compensation force field is a two-dimensional distribution map that changes with time and reflects the actual clamping force generated in different regions under the action of the tool. The compensation force field takes into account the relationship between the machining path, the feed rate, and the workpiece structure, and can more realistically restore the force changes during the clamping process.

[0044] Finally, the compensation force field is normalized and transformed into a form with a unified dimension and numerical range for comparison and correction with the original time-sequence grayscale image. The normalized compensation force field is used as a correction factor to superimpose or adjust the grayscale values of each pixel in the original grayscale image, generating a set of grayscale image sequences corrected by the actual machining effects. Using the corrected images for subsequent image feature recognition can improve the judgment accuracy and reduce misjudgments caused by differences in machining paths or material structures.

[0045] Furthermore, this application also includes: establishing force-similar position points at different time nodes using the true feed rate, the cover plate structure, and the workpiece feature extraction results; performing comparison and verification of the time-sequence grayscale images through the force-similar position points to establish additional verification features; and compensating the stability verification evaluation results according to the additional verification features.

[0046] Specifically, establishing force-similar position points at different time nodes using the true feed rate, the cover plate structure, and the workpiece feature extraction results means that during the machining process, by combining the actual feed depth and speed of the tool at a certain moment, i.e., the true feed rate, the geometric feature information of the workpiece, and the structural distribution of the pressing cover plate, the machining states at different time points are compared to identify position points with similar force conditions. The true feed rate refers to the actual material cutting depth considering the influence of air cutting, feed speed changes, etc.; the cover plate structure refers to the physical structure shape, rigidity distribution, etc. of the pressing cover plate; and the workpiece feature extraction results refer to the surface, internal, or edge feature information of the workpiece obtained through image processing or modeling. Through joint analysis, position points with similar force states can be found at different time points, although the machining times are different, due to similar structures, feeds, and material distributions.

[0047] Compare the grayscale values of these similar force - applied position points at different times to confirm whether the same or similar sensing responses are given under the same stress conditions. The sequential grayscale images reflect the time - variation of the pressure or force - applied state monitored by the sensor. By comparing the changes in the grayscale values of the force - similar points on the image, it can be determined whether the system is stable and consistent. The comparison results are used to establish additional verification features, that is, a class of additional information features, to test the consistency or repeatability of the system response. For example, during two processes in different regions but with similar forces, if the trends of the grayscale changes in the images are highly consistent, the additional verification feature shows a high degree of consistency.

[0048] Compensating the stability verification evaluation result according to the additional verification feature means taking the newly extracted verification feature as a correction factor and adding it to the original stability evaluation system to improve its accuracy. If it is considered that the holding pressure in a certain stage is unstable in the preliminary evaluation, but the additional verification feature shows that the force response in this stage is highly consistent with other normal stages, the original judgment can be corrected and compensated. For example, if the original deviation value is 0.08 and the additional verification feature shows that the similarity is more than 90%, the deviation value can be corrected to 0.04 to avoid false alarms.

[0049] Furthermore, this application also includes: sorting the sequential grayscale images in chronological order into frame images; calculating the centroid of the grayscale image and the geometric center of the grayscale image for each frame image, performing the deviation comparison between the centroid of the grayscale image and the geometric center of the grayscale image, and establishing the hot - zone center offset feature; setting the regional width value of the boundary region, extracting the grayscale data of the boundary region of each frame image within the regional width value; calculating the grayscale mean for each boundary region based on the grayscale data to generate a grayscale mean set mapped to the boundary region; and using the grayscale mean set to perform the neighborhood gradient calculation to establish the boundary holding jump feature.

[0050] Specifically, the two - dimensional grayscale images continuously collected during the processing are arranged in chronological order into a series of frames, reflecting the pressure distribution state of the workpiece contact surface at different time points. Each frame represents an instantaneous two - dimensional force field state. Through sorting, the evolution process of the pressure field over time can be intuitively captured. For example, at the beginning of the processing, the pressure distribution is sparse, while at the peak cutting stage, the grayscale is dense, providing a logical basis for the independent analysis of each frame image and the identification of dynamic trends.

[0051] For each frame of the image, calculate the centroid of the grayscale image and the geometric center of the grayscale image. The centroid of the grayscale image refers to the weighted average position of all pixel grayscale values in the image, representing the position of the main pressure concentration area; the geometric center is the geometric center point on the image plane, which is determined only by the image shape without considering the grayscale distribution. The deviation distance between the two center points can be used as a measure of the hot zone offset to determine whether the pressure center deviates from the structural center. The deviation ratio quantifies the coordinate difference between the two. If the offset continuously increases over time, it may indicate uneven stress on the workpiece or displacement is occurring.

[0052] Set the regional width value of the boundary area, which is used to expand a fixed-width range inward from the four sides of the image. Then, take a 10-pixel-wide area from the image edge towards the center as the object of boundary analysis. Extract the grayscale values for each frame of the image in the boundary area, and calculate the grayscale mean value of this area to obtain a set of average value data corresponding to each boundary area, representing the overall stress level of the edge area at that time point, forming a grayscale mean value set that can change with time, reflecting the change trend of pressure in the boundary area during the entire processing.

[0053] Using the grayscale mean value set, further calculate the grayscale change rate between adjacent time points, that is, perform neighborhood gradient calculation. The gradient refers to the change rate of grayscale values. By analyzing the change rate, the sudden change points of the boundary stress can be captured. Establishing the boundary clamping jump feature is to identify the positions and times of abnormal changes, which is used to reflect whether the workpiece has loosened or the local stress is unstable in the edge area.

[0054] Furthermore, this application also includes: obtaining the stable clamping pressure of the calibrated test workpiece, converting the stable clamping pressure into the corresponding target grayscale value; after configuring the local window, constructing the local neighborhood of the pixel points in each frame of the image; calculating the local mean and local standard deviation of the local neighborhood, and using the local mean, local standard deviation, and the target grayscale value for fusion calculation to establish a binary segmentation threshold; performing binary processing on each frame of the image based on the binary segmentation threshold to establish a binary image; performing image evaluation on the binary image through the connectivity index to establish the integrity feature of the grayscale connected domain.

[0055] Specifically, after the workpiece is clamped, the stable and uniform standard pressure value applied by the clamping system to the workpiece in a static state is obtained as the stable clamping pressure of the calibrated test workpiece, which is used as the benchmark for judging the normal clamping state. Using the pressure-grayscale mapping relationship of the sensor, the stable clamping pressure is mapped to a specific grayscale level in the image, which is used for comparison and judgment in the subsequent image processing process to help determine which areas belong to the normal clamping state.

[0056] Subsequently, a smaller rectangular area, such as a 5-pixel × 5-pixel square area, is selected in the entire image to establish a local neighborhood around each pixel point. The construction of the local neighborhood helps analyze the image characteristics around each pixel, without being interfered by the global features of the entire image, thereby improving the processing fineness and sensitivity, and is especially suitable for detecting local abnormal changes.

[0057] Next, in each local neighborhood, the local mean and the local standard deviation are calculated. The local mean represents the average level of pixel grayscale within the neighborhood, and the local standard deviation represents the degree of grayscale dispersion, which is the magnitude of the fluctuation. Combining the obtained target grayscale value, a binary segmentation threshold for the pixels in this neighborhood is established through a fusion calculation formula as a judgment criterion to classify the current pixel into two categories: "normal pressing" or "abnormal pressing". For example, if the target grayscale value is 180, the local mean is 160, and the standard deviation is 20, the threshold obtained after fusion may be 170, and pixels lower than 170 will be regarded as abnormal areas.

[0058] Then, each frame of the image is processed using the binary segmentation threshold to perform binarization, that is, each pixel in the grayscale image is converted into one of two states: black or white. Black usually represents the abnormal pressing area, and white represents the normal pressing area, forming a binary image, which is the basis for subsequent connectivity analysis. The advantage of binarization is that it significantly reduces the image complexity, making the structural feature analysis clearer and the calculation efficiency higher.

[0059] Finally, an image evaluation is performed on the binary image, and a connectivity index is used to measure the continuity and integrity of the white area in the image. Connectivity refers to whether pixels can be connected into a continuous area through adjacent pixels, such as analyzing the number and shape of its connected components. If the white area is highly coherent and concentrated, it indicates that the pressing is overall uniform and stable; on the contrary, if there are a large number of scattered white isolated blocks, there may be problems such as local pressing looseness and abnormal fixture contact, thereby establishing the integrity feature of the grayscale connected domain.

[0060] Furthermore, this application also includes: after performing connected domain segmentation on the binary image, obtaining a connected domain quantity index; obtaining the total connected area of all connected domains, and establishing a maximum connected ratio index based on the total connected area; using the connected domain quantity index and the connected ratio index as the connectivity index to complete the image evaluation.

[0061] Specifically, identify the continuous white pixel regions in the image that has undergone black-and-white binarization processing, and count how many connected regions there are in total in the image. The concept of a connected component means a group of same-color pixel blocks in the image that can be continuously connected in the up-down, left-right, or diagonal directions. White represents the normal pressing area, so each connected component represents a locally normally stressed area. The fewer the number of connected components, the more concentrated and complete the pressing area is usually; while a larger number may indicate scattered pressing, uneven stress, or abnormal phenomena.

[0062] Next, accumulate the pixel counts of all connected white pixel blocks to obtain a total area value. Then, divide the total connected area by the total pixel area of the entire image to get a ratio, called the maximum connected ratio index, which is used to describe the proportion of the pressing area in the image. If the total number of pixels in the image is 100,000 pixels and the total white area is 60,000 pixels, then the maximum connected ratio is 0.6. The larger the ratio, the more areas are in a normally stressed state; on the contrary, the smaller the ratio, the lower the coverage rate of the pressing area, indicating that there may be a large number of uncompacted or loose parts.

[0063] Finally, use these two results, the connected component number index and the maximum connected ratio index, together as the connectivity index to complete the evaluation of the structural integrity of the entire image. The connectivity index is a comprehensive index that reflects whether the pressing area is concentrated and whether the stressed area is extensive. For example, if an image has 1 connected component and a connected ratio of 0.9, it means that almost the entire pressing surface is stressed uniformly; while another image with 15 connected components and a connected ratio of only 0.4 indicates the existence of multiple scattered stressed areas and a large area not being effectively pressed.

[0064] Furthermore, this application also includes: horizontally and vertically flipping each frame of the image, calculating the absolute difference in grayscale between the original image and the mirror image; generating horizontal and vertical influence factors using the workpiece feature extraction results of the calibrated test workpiece; calculating the symmetry imbalance feature using the influence factors and the absolute difference in grayscale.

[0065] Specifically, horizontally and vertically flip each frame of the image, and perform mirror transformation on the original grayscale image along the horizontal and vertical directions respectively. Horizontal flipping will move the pixels on the left side of the image to the right side and the right side to the left side; vertical flipping will invert the image up and down, moving the upper pixels to the bottom and the bottom pixels to the top, for analyzing the symmetry of the image itself in the spatial structure and judging whether the grayscale distributions on both sides or up and down of the image are consistent. Subsequently, calculate the absolute difference in grayscale between the flipped mirror image and the original image at the corresponding pixel positions, that is, subtract each pixel and then take its absolute value, which represents the difference size on the axis of symmetry of the image. The smaller the difference, the more symmetric the image is.

[0066] Next, to improve the analysis accuracy, the results of workpiece feature extraction are introduced to generate horizontal and vertical influence factors. Workpiece feature extraction refers to identifying key structural features from the calibrated test workpiece, such as contour, center point, boundary range, etc., which can reveal the physical symmetry characteristics of the workpiece. The influence factor is a set of weight parameters calculated from the features, used to describe the importance or deviation degree of the symmetric structure of the workpiece in the horizontal and vertical directions. Combining the influence factor with the absolute difference in grayscale, through methods such as weighting or normalization, the symmetry imbalance feature is calculated, which reflects the left-right or up-down imbalance degree of the image during the force application or pressing process, and is used to measure the symmetry stability of the pressing structure. For example, if the overall left-right mirror grayscale difference is less than 10 and the workpiece has strong left-right symmetry, the symmetry imbalance value may be less than 0.1, indicating balanced force application; but if the difference is as high as over 30 and the influence factor weight is high, the symmetry imbalance value will increase significantly, suggesting uneven structure pressing, which may affect the processing accuracy.

[0067] Furthermore, this application also includes: extracting the grayscale values of feature points changing over time from the sequential grayscale images to establish a three-dimensional matrix; extracting typical grayscale trajectories based on the three-dimensional matrix, and using the typical grayscale trajectories for evolutionary morphology clustering analysis to establish dynamic trajectory features; and supplementing and adding the dynamic trajectory features to the feature recognition results.

[0068] Specifically, pixel points at specific positions are selected from multiple consecutive grayscale images, and the grayscale changes of the pixel points in different time frames are tracked. The grayscale value is the numerical value representing the brightness in the image, where 0 represents black and 255 represents white. After summarizing the grayscale changes of the feature points at different time points, a three-dimensional matrix with spatial position and time as the three-dimensional coordinate axes is formed to describe the brightness change trajectory during the workpiece pressing process.

[0069] Then, based on the three-dimensional matrix, typical grayscale trajectories can be extracted, that is, identifying representative grayscale change patterns from numerous change trajectories. For example, in the stable pressing area, the grayscale values of some pixels may continuously remain around 220, while in the unstable area, they may fluctuate significantly between 180 and 230. By performing evolutionary morphology clustering analysis on the trajectories, that is, classifying the trajectories into several categories according to the change trend and morphology, such as stable rising type, severe oscillation type, or periodic fluctuation type, dynamic trajectory features are established, which can then reflect the possible stable areas, unstable areas, or potential abnormal areas during the cover pressing process.

[0070] Finally, the extracted dynamic trajectory features are added to the existing feature recognition results as a dynamic supplement to two-dimensional features such as hot zone offset and connectivity integrity in the grayscale image, making the evaluation result of the entire system more comprehensive, including both the instantaneous features of the pressing force in the spatial distribution and the trend features evolving over time.

[0071] In summary, the stability evaluation method for cover plate pressing provided by the present application has the following technical effects: By achieving the technical goal of multi-dimensional fusion modeling and stability intelligent evaluation based on the workpiece structure characteristics and the machining dynamic process, the technical effects of improving the recognition accuracy of pressing stability, reducing the misjudgment rate, and enhancing the adaptability of the system to complex machining environments are achieved.

[0072] Embodiment 2. Based on the same inventive concept as the stability evaluation method for cover plate pressing in the foregoing embodiment, the present application also provides a stability evaluation device for cover plate pressing. Please refer to the attached Figure 2 , including: a fixing module 11, configured to load a calibration test workpiece onto a target fixture and fix it with a cover plate after setting a sensor array on the contact surface of the calibration test workpiece; a machining module 12, configured to activate the sensor array and control the machine tool to start and perform machining after confirming that the cover plate assembly is completed; a reading module 13, configured to read the monitoring data of the sensor array in real time, establish a contact two-dimensional calibration force field, and normalize the contact two-dimensional calibration force field into a time-series grayscale image; a feature recognition module 14, configured to recognize features such as the center offset of the hot zone, the boundary pressing jump feature, the local maximum clustering feature, the integrity feature of the grayscale connected domain, and the symmetry imbalance feature of the time-series grayscale image, and establish a feature recognition result; a signal establishing module 15, configured to synchronously obtain the acquisition information of a triaxial accelerometer arranged on the fixture, establish a micro-vibration signal, and use the micro-vibration signal as a first verification signal; a monitoring module 16, configured to monitor key points of the workpiece using a laser displacement sensor, establish a workpiece time-series displacement, and use the workpiece time-series displacement as a second verification signal; a verification module 17, configured to perform stability verification on the feature recognition result through the first verification signal and the second verification signal, and output a stability verification evaluation result.

[0073] Further, the stability evaluation device for cover plate pressing is further configured to: obtain a zero-point grayscale image corresponding to the time series zero point; calculate the grayscale mean value of the zero-point grayscale image, and use the grayscale mean value to perform a traversal deviation comparison of the zero-point grayscale image to generate a zero-point pressure deviation map; perform feature extraction of the calibration test workpiece, and establish a zero-point feature recognition result using the workpiece feature extraction result and the zero-point pressure deviation map.

[0074] Furthermore, the stability evaluation device for cover plate pressing is also used for: performing control information interaction of the machine tool system to obtain processing control information; obtaining the tool feed rate mapped with the tool path according to the processing control information, performing influence compensation of the tool feed rate based on the workpiece feature extraction result and the tool path, and establishing the actual feed rate; using the actual feed rate to perform machining fitting of the calibrated test workpiece fixed by the pressed plate to generate a time-sequential compensation force field; normalizing the time-sequential compensation force field, correcting the time-sequential gray-scale image, and establishing a feature recognition result based on the corrected time-sequential gray-scale image.

[0075] Furthermore, the stability evaluation device for cover plate pressing is also used for: establishing stress similarity position points at different time nodes by using the actual feed rate, the cover plate structure, and the workpiece feature extraction result; performing comparison and verification of the time-sequential gray-scale image through the stress similarity position points to establish additional verification features; compensating the stability verification evaluation result according to the additional verification features.

[0076] Furthermore, the stability evaluation device for cover plate pressing is also used for: sorting the time-sequential gray-scale images into frame images in chronological order; calculating the centroid of the gray-scale image and the geometric center of the gray-scale image for each frame image, performing deviation comparison of the centroid of the gray-scale image and the geometric center of the gray-scale image, and establishing a hot zone center offset feature; setting the region width value of the boundary region, and extracting the gray-scale data of the boundary region of each frame image within the region width value; calculating the gray-scale mean value of each boundary region based on the gray-scale data to generate a gray-scale mean value set mapped with the boundary region; performing neighborhood gradient calculation by using the gray-scale mean value set to establish a boundary pressing jump feature.

[0077] Furthermore, the stability evaluation device for cover plate pressing is also used for: obtaining the stable pressing pressure of the calibrated test workpiece, and converting the stable pressing pressure into the corresponding target gray-scale value; configuring a local window and constructing a local neighborhood of the pixel points within each frame image; calculating the local mean value and the local standard deviation of the local neighborhood, and performing fusion calculation by using the local mean value, the local standard deviation, and the target gray-scale value to establish a binary segmentation threshold; performing binary processing on each frame image based on the binary segmentation threshold to establish a binary image; performing image evaluation of the binary image through a connectivity index to establish a gray-scale connected domain integrity feature.

[0078] Furthermore, the stability evaluation device for cover plate pressing is also used for: after performing connected domain segmentation on the binary image, obtaining a connected domain quantity index; obtaining the total connected area of all connected domains, and establishing a maximum connected ratio index according to the total connected area; using the connected domain quantity index and the connected ratio index as the connectivity index to complete image evaluation.

[0079] Further, the stability evaluation device for cover plate pressing is further configured to: horizontally and vertically flip each frame of image, and calculate the absolute difference in grayscale between the original image and the mirror image; generate horizontal and vertical influence factors by using the workpiece feature extraction results of the calibrated test workpiece; calculate the symmetry imbalance feature by using the influence factors and the absolute difference in grayscale.

[0080] Further, the stability evaluation device for cover plate pressing is further configured to: extract the grayscale values of feature points changing with time by using the sequential grayscale images, and establish a three-dimensional matrix; extract typical grayscale trajectories based on the three-dimensional matrix, and perform evolutionary morphology clustering analysis by using the typical grayscale trajectories to establish dynamic trajectory features; supplement and add the dynamic trajectory features to the feature recognition results.

[0081] The various embodiments in this specification are described in a progressive manner, and the key points of each embodiment are the differences from other embodiments. The stability evaluation method and specific examples for cover plate pressing in the foregoing Embodiment 1 are equally applicable to the stability evaluation device for cover plate pressing in this embodiment. Through the foregoing detailed description of the stability evaluation method for cover plate pressing, those skilled in the art can clearly know the stability evaluation device for cover plate pressing in this embodiment. Therefore, for the sake of brevity of the specification, it will not be described in detail herein.

[0082] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0083] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is also intended to include these changes and modifications.

Claims

1. A method for evaluating the stability of cover plate pressing, characterized in that, Including: After setting up the sensor array on the contact surface of the calibration test workpiece, load the calibration test workpiece onto the target fixture and fix it with the cover plate; After confirming the completion of the cover plate assembly, activate the sensor array and control the machine tool to start and perform machining; Read the monitoring data of the sensor array in real time, establish a contact two-dimensional calibration force field, and normalize the contact two-dimensional calibration force field into a time-series grayscale image; Identify the hot zone center offset feature, boundary pressing jump feature, local maximum clustering feature, grayscale connected domain integrity feature, and symmetry imbalance feature of the time-series grayscale image, and establish a feature recognition result; Synchronously obtain the acquisition information of the triaxial accelerometer set on the fixture, establish a micro-vibration signal, and use the micro-vibration signal as the first verification signal; Use a laser displacement sensor to monitor the key points of the workpiece, establish a workpiece time-series displacement, and use the workpiece time-series displacement as the second verification signal; Stably verify the feature recognition result through the first verification signal and the second verification signal, and output a stability verification evaluation result; The establishment of the feature recognition result further includes: Obtain the zero-point grayscale image corresponding to the time series zero point; Calculate the grayscale mean value of the zero-point grayscale image, and use the grayscale mean value to perform a traversal deviation comparison of the zero-point grayscale image to generate a zero-point pressure deviation map; Perform feature extraction of the calibration test workpiece, and establish a zero-point feature recognition result using the workpiece feature extraction result and the zero-point pressure deviation map.

2. The stability evaluation method for cover plate pressing according to claim 1, characterized in that, Before establishing the feature recognition result, it further includes: Execute the control information interaction of the machine tool system to obtain machining control information; Obtain the tool feed corresponding to the tool path mapping according to the machining control information, and perform an influence compensation of the tool feed based on the workpiece feature extraction result and the tool path to establish a true feed; Use the true feed to perform machining fitting of the calibration test workpiece fixed by the pressing plate to generate a time-series compensation force field; After normalizing the time-series compensation force field, correct the time-series grayscale image, and establish a feature recognition result using the corrected time-series grayscale image.

3. The stability evaluation method for cover plate pressing according to claim 2, wherein The output of the stability verification evaluation result includes: Establish force-similar position points at different time nodes using the true feed, cover plate structure, and workpiece feature extraction result; Perform a comparison verification of the time-series grayscale image through the force-similar position points to establish additional verification features; Compensate the stability verification evaluation result according to the additional verification features.

4. The stability evaluation method for cover plate pressing according to claim 1, characterized in that, The identification of the hot zone center offset feature, boundary pressing jump feature, local maximum clustering feature, grayscale connected domain integrity feature, and symmetry imbalance feature of the time-series grayscale image includes: Sort the time-series grayscale image in chronological order as frame images; For each frame image, calculate the center of gravity of the grayscale image and the geometric center of the grayscale image, perform a deviation comparison between the center of gravity of the grayscale image and the geometric center of the grayscale image, and establish a hot zone center offset feature; Set the region width value of the boundary region, and extract the boundary region grayscale data of each frame image within the region width value; Calculate the grayscale mean value of each boundary region based on the grayscale data to generate a grayscale mean value set mapped to the boundary region; Perform neighborhood gradient calculation using the grayscale mean set to establish the boundary holding jump feature.

5. The stability evaluation method for cover plate pressing according to claim 1, characterized in that The recognition of the hot zone center offset feature, boundary holding jump feature, local maximum clustering feature, grayscale connected domain integrity feature, and symmetry imbalance feature for the sequential grayscale images further includes: Obtain the stable holding pressure of the calibrated test workpiece and convert the stable holding pressure into the corresponding target grayscale value; After configuring the local window, construct the local neighborhood of the pixel points within each frame of the image; Calculate the local mean and local standard deviation of the local neighborhood, and perform fusion calculation using the local mean, local standard deviation, and the target grayscale value to establish the binary segmentation threshold; Perform binary processing on each frame of the image based on the binary segmentation threshold to establish a binary image; Perform image evaluation of the binary image through the connectivity index to establish the grayscale connected domain integrity feature.

6. The stability evaluation method for cover plate pressing according to claim 5, characterized in that The performing image evaluation of the binary image through the connectivity index to establish the grayscale connected domain integrity feature includes: After performing connected domain segmentation on the binary image, obtain the connected domain quantity index; Obtain the total connected area of all connected domains, and establish the maximum connected ratio index based on the total connected area; Use the connected domain quantity index and the connected ratio index as the connectivity index to complete the image evaluation.

7. The stability evaluation method for cover plate pressing according to claim 5, wherein The recognition of the hot zone center offset feature, boundary holding jump feature, local maximum clustering feature, grayscale connected domain integrity feature, and symmetry imbalance feature for the sequential grayscale images further includes: Perform horizontal flipping and vertical flipping on each frame of the image, and calculate the absolute grayscale difference between the original image and the mirror image; Generate horizontal and vertical influence factors using the workpiece feature extraction results of the calibrated test workpiece; Calculate the symmetry imbalance feature using the influence factors and the absolute grayscale difference.

8. The stability evaluation method for cover plate pressing according to claim 1, wherein The establishment of the feature recognition result further includes: Extract the grayscale values of the feature points changing with time using the sequential grayscale images to establish a three-dimensional matrix; Extract typical grayscale trajectories based on the three-dimensional matrix, and perform evolutionary morphology clustering analysis using the typical grayscale trajectories to establish dynamic trajectory features; Supplement and add the dynamic trajectory features to the feature recognition result.

9. A stability evaluation device for pressing a cover plate, characterized in that, The steps for implementing the stability evaluation method for cover plate holding according to any one of claims 1 to 8 include: A fixing module, configured to, after setting a sensor array on the contact surface of the calibrated test workpiece, load the calibrated test workpiece onto the target fixture and fix it with the cover plate; A processing module, configured to, after confirming the completion of the cover plate assembly, activate the sensor array, control the machine tool to start, and perform machining; A reading module, configured to read the monitoring data of the sensor array in real time, establish a contact two-dimensional standard force field, and normalize the contact two-dimensional standard force field into a sequential grayscale image; A feature recognition module, configured to recognize the hot zone center offset feature, boundary holding jump feature, local maximum clustering feature, grayscale connected domain integrity feature, and symmetry imbalance feature for the sequential grayscale images, and establish a feature recognition result; A signal establishment module, configured to synchronously obtain the acquisition information of a triaxial accelerometer disposed on a fixture, establish a micro-vibration signal, and use the micro-vibration signal as a first verification signal; A monitoring module, configured to monitor key points of a workpiece by using a laser displacement sensor, establish a workpiece time-series displacement, and use the workpiece time-series displacement as a second verification signal; A verification module, configured to perform a stability verification on the feature recognition result through the first verification signal and the second verification signal, and output a stability verification evaluation result.

Images