A bearing edge correction system based on machine vision

Abstract

The present invention discloses a bearing edge correction system based on machine vision, which relates to the field of machine vision technology. The system accurately collects and processes bearing images and edge correction data through the machine vision and sensor technology of the correction data acquisition module to generate a standardized data set. The error calculation and priority allocation module accurately calculates the bearing edge correction error Ei and allocates the correction priority Pi to ensure that key tasks are processed first. The multi-bearing synchronous correction scheduling module introduces an intelligent scheduling algorithm to optimize task scheduling. The correction optimization adjustment module calculates the correction time Tfix based on the correction path optimization formula, adjusts the correction path and strength, and ensures accuracy through a secondary evaluation of the residual error Eres. The feedback and adaptive adjustment module dynamically adjusts through an adaptive formula when an error occurs to further improve the correction effect. The overall evaluation and optimization module evaluates the system performance by calculating the correction efficiency Etotal.

Description

Technical Field

[0001] The invention relates to the technical field of machine vision, and in particular to a bearing edge correction system based on machine vision. Background Art

[0002] In modern industrial production, machine vision-based detection and correction technology can analyze and correct workpieces in real time through high-precision image acquisition and processing technology, greatly improving production efficiency and quality. In specific applications, the edge correction of bearings is a typical scenario, involving the detection of bearing surface flatness and edge correction. The machine vision-based bearing edge correction system can effectively replace traditional manual or low-precision mechanical methods and improve correction accuracy and consistency.

[0003] At present, in the process of bearing edge correction, traditional correction methods have many defects. First, traditional manual correction relies on the experience of the operator, which is not only inefficient, but also has accuracy greatly affected by human factors. Secondly, although some semi-automatic correction systems have improved the correction efficiency to a certain extent, due to the inability to monitor and dynamically adjust the correction process in real time, the correction error is still large and cannot meet the high-precision requirements. In addition, in the environment of multi-station correction, the synchronous correction scheduling of multiple bearings is difficult to achieve intelligent optimization, which often leads to problems such as waste of resources, uneven correction time, and low production efficiency. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a bearing edge correction system based on machine vision, which solves the problems mentioned in the background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: including a correction data acquisition module, an error calculation and priority allocation module, a multi-bearing synchronous correction scheduling module, a correction optimization adjustment module, a feedback and adaptive adjustment module and an overall evaluation and optimization module;

[0006] The modified data acquisition module collects the image data and edge-smoothing data of the bearing through machine vision and an integrated sensor group, and pre-processes to obtain a standard image data set and a standard edge-smoothing data set, and then performs feature extraction on the standard image data set to obtain a bearing surface feature set;

[0007] The error calculation and priority allocation module is used to construct an error calculation formula, extract the bearing surface feature set and input it into the error calculation formula, calculate and output the whole edge error value Ei, and construct the bearing correction priority formula based on the whole edge error value Ei, calculate and output the correction priority Pi, and determine the priority correction order;

[0008] The multi-bearing synchronous correction scheduling module is used to construct an intelligent scheduling allocation algorithm formula. The edge error value Ei and the correction priority Pi are combined with the station idle state Wfree, and input into the intelligent scheduling allocation algorithm formula to calculate and output the scheduling score D i，j , and preset the scheduling threshold F1 and the obtained scheduling score D i，j Conduct preliminary comparative assessment to determine the dispatch situation;

[0009] The correction optimization adjustment module constructs a correction path optimization formula at each bearing edge correction station, calculates and outputs the correction path time Tfix, optimizes the motion trajectory and force of the correction tool for bearing edge correction, and then performs a secondary execution of the error calculation formula after correction optimization to calculate and output the corrected residual error value Eres, and then performs a secondary comparative evaluation between the preset edge correction threshold F2 and the obtained residual error value Eres;

[0010] The feedback and adaptive adjustment module performs a secondary correction feedback formula based on the secondary comparison evaluation results to calculate and output the adaptive feedback adjustment value Fadj, performs secondary correction on the residual bearing trimming after correction optimization, and adjusts the strategy in real time;

[0011] The overall evaluation and optimization module is used to construct the correction efficiency algorithm formula and calculate the output correction efficiency Etotal after the secondary correction is completed. The correction efficiency threshold F3 is preset and compared with the correction efficiency Etotal for three times to analyze the error changes of the bearing edge correction and determine the optimization effect of scheduling and correction.

[0012] Preferably, the correction data acquisition module includes a data acquisition unit and a feature extraction unit;

[0013] The data acquisition unit includes an image acquisition unit, an environment acquisition unit and a preprocessing unit;

[0014] The image acquisition unit acquires image data of the integral bearing through a machine vision camera;

[0015] The environment acquisition unit is used to install temperature sensors, humidity sensors, workstation status sensors and IMU sensors in the bearing trimming correction environment to collect trimming data in real time. The trimming data includes temperature Tenv, humidity Henv, workstation idle state Wfree and bearing rotation angle error Jd.

[0016] The preprocessing unit is used to perform standardization processing on the collected image data and edge-corrected data to obtain a standard image data set and a standard edge-corrected data set;

[0017] The standard image data set normalizes the image data and automatically scales the image data to the same size using a graphic editing tool. The bearing image is then captured using a high-precision camera. The actual edge coordinates (Xi, Yi) are obtained by combining edge detection algorithms, sub-pixel precision technology, camera calibration parameters, and pixel-to-physical coordinate conversion.

[0018] The standard image dataset includes bearing illumination reflectance images;

[0019] The standard edge-corrected data set is obtained by cleaning and denoising the collected edge-corrected data, and then normalizing it. At the same time, the change rate of temperature Tenv and humidity Hev is calculated to obtain the temperature change rate △Tenv and the humidity change rate △Henv;

[0020] The temperature change rate △Tenv and the humidity change rate △Henv are calculated and obtained by the following algorithm formulas respectively;

[0021] ;

[0022] Where, Tenv(t1) represents the temperature value at the previous time point, Tenv(t2) represents the temperature value at the next time point, Hev(t1) represents the humidity value at the previous time point, and Hev(t2) represents the humidity value at the next time point;

[0023] The feature extraction unit performs feature extraction to obtain a bearing surface feature set based on the standard image data set obtained after preprocessing. The bearing surface feature set includes a bearing surface roughness feature Rs and a bearing material uniformity feature Mu.

[0024] The bearing surface roughness feature Rs is processed by edge detection image processing algorithm on the light reflection image captured in the standard image data set, the brightness value of each pixel in the light reflection image is extracted, the reflectivity distribution of the brightness value of each pixel is calculated, and then the reflectivity distribution and the known incident angle of the light source are calculated and obtained;

[0025] The uniformity characteristic Mu of the bearing material is obtained by comparing the spectral data of different areas in the light reflection image and calculating the deviation between the spectral curve of each area and the photovoltaic curve straight line of the standard bearing material.

[0026] Preferably, the error calculation and priority allocation module includes an error calculation unit and a priority calculation unit;

[0027] The error calculation unit is used to construct an error calculation formula, and to extract the actual edge coordinates (Xi, Yi), the bearing surface feature set, and the standard edge correction data set in real time and input them into the error calculation formula to calculate and output the edge correction error value Ei;

[0028] The edge error value Ei is calculated and output by the following error calculation formula;

[0029] ;

[0030] Wherein, Ei represents the integral error value of the i-th integral bearing, Xideal represents the i-th coordinate point on the ideal horizontal axis, Yideal represents the i-th coordinate point on the ideal vertical axis, k1 represents the adjustment coefficient of the roughness feature of the bearing surface, and k2 represents the adjustment coefficient of the bearing rotation angle error.

[0031] Preferably, the priority calculation unit constructs a priority algorithm formula, extracts the bearing surface feature set and the standard edge correction data set in real time, and calculates and outputs the correction priority Pi;

[0032] The corrected priority Pi is calculated and output by the following priority algorithm formula;

[0033] ;

[0034] Where Pi represents the correction priority of the i-th bearing.

[0035] Preferably, the multi-bearing synchronous correction scheduling module includes a scheduling allocation analysis unit and a scheduling allocation evaluation unit;

[0036] The scheduling and allocation analysis unit is used to construct an intelligent scheduling and allocation algorithm formula. The side error values ​​Ei and correction priorities Pi of all side bearings are input into the intelligent scheduling and allocation algorithm formula to calculate and output the scheduling score D. i，j , score each full-edge bearing and correction station;

[0037] Scheduling score D i，j Through the following intelligent scheduling allocation algorithm formula;

[0038] ;

[0039] Where D i，j represents the scheduling score for assigning the i-th bearing to the j-th correction station, Wfee represents the idle state of the station, 1 represents idle, 0 represents occupied, π represents the circumference of a circle, which is 3.14, Vmax represents the upper limit of the speed of the correction tool, and Emax represents the set upper limit of the correctable error. represents the length of the corrected path after optimization, Indicates the correction speed, Represents the weight value of the integral edge error value Ei, Indicates the weight value of the bearing surface roughness feature Rs;

[0040] The scheduling allocation evaluation unit sets the scheduling threshold F1 based on the historical data set and production target pre-acquired by the system, and then compares it with the obtained scheduling score Di，j Conduct a preliminary comparative assessment to determine the scheduling situation. The specific assessment contents are as follows;

[0041] When the scheduling score is D i，j > scheduling threshold F1, it is marked as the first priority, indicating that the correction priority Pi of the bearing is higher, the correction time estimate Test is shorter, or the edge error value Ei is relatively small, indicating that the correction process will not take too much time, indicating that the bearing needs to be corrected as soon as possible, and the i-th bearing is immediately assigned to the j-th workstation;

[0042] When the scheduling score is D i，j ≤ scheduling threshold F1, it is marked as the second priority and no workstation is assigned. At the same time, the i-th bearing is arranged to the first priority and sorted. When the first priority bearing is corrected, the second priority bearing is arranged to the workstation according to the sorting.

[0043] Preferably, the correction optimization adjustment module includes a correction path optimization unit and a correction error residual analysis unit;

[0044] The correction path optimization unit is used to construct the correction path optimization formula at each edge correction station, and then calculate the time Tfix of the output correction path according to the bearing rotation angle error Jd, and optimize the motion trajectory and force of the correction tool;

[0045] The time Tfix of the corrected path is calculated and obtained by the corrected path optimization formula;

[0046] ;

[0047] Where N represents the number of error points that need to be corrected, n represents the index of the error point, which is used to traverse N error points that need to be corrected, and Jd n Indicates the bearing rotation angle error Jd at the nth error point during the current bearing correction process, Jd n，ideal represents the ideal rotation angle of the ith error point, k3 represents the rotation correction coefficient, and fi represents the correction force of the bearing edge correction tool, which is automatically adjusted to cope with the unevenness of the bearing material.

[0048] Preferably, the correction error residual analysis unit is used to execute the error calculation formula for the second time to calculate and output the corrected residual error value Eres after the path correction optimization, and preset the edge correction threshold F3 based on the bearing edge error standard, and perform a second comparative evaluation with the obtained residual error value Eres to analyze the error of the bearing edge after the path correction optimization. The specific evaluation content is as follows;

[0049] When the residual error value Eres> the edge correction threshold F2, it means that there is an error in the bearing edge correction after the path correction optimization, and the secondary correction is performed at this time;

[0050] When the residual error value Eres ≤ the edge correction threshold F2, it means that after the path correction optimization, there is no error in the bearing edge correction, and the processing is completed.

[0051] Preferably, the feedback and adaptive module is triggered when there is an error in the bearing edge correction after the path correction optimization, and calculates and outputs the adaptive feedback adjustment value Fadj by constructing a secondary correction feedback formula to dynamically adjust the correction force and path;

[0052] The adaptive feedback adjustment value Fadj is calculated and outputted by the following quadratic correction feedback formula;

[0053] ;

[0054] In the formula, Represents the adaptation coefficient of the residual error value Eres, Represents the environmental adaptation coefficient; used to control the adjustment of the correction strength.

[0055] Preferably, the overall evaluation and optimization module includes an overall correction efficiency analysis unit and an overall correction efficiency evaluation unit;

[0056] The overall correction efficiency analysis unit is used to construct the correction efficiency algorithm formula and calculate the output correction efficiency Etotal after the secondary correction is completed, and analyze the overall correction efficiency of all bearings;

[0057] The corrected efficiency Etotal is calculated by the following corrected efficiency algorithm formula;

[0058] ;

[0059] Where, I represents the total number of full-edge bearings, Ei fin represents the error value of the i-th bearing after correction, and Ttotal represents the total running time of the system.

[0060] Preferably, the overall correction efficiency evaluation unit performs three comparative evaluations based on the correction efficiency threshold F3 preset by the user and the obtained correction efficiency Etotal to measure the overall correction effect of the system. The specific evaluation contents are as follows;

[0061] When the correction efficiency Etotal ≤ the correction efficiency threshold F3, it means that the bearing edge correction effect is normal and the correction is completed;

[0062] When the correction efficiency Etotal>correction efficiency threshold F3, it indicates that the bearing edge correction effect is abnormal. At this time, the iterative correction optimization adjustment module is performed until the bearing edge correction effect is normal, and the iteration is terminated.

[0063] The present invention provides a bearing edge correction system based on machine vision. It has the following beneficial effects:

[0064] (1) The system significantly improves the correction accuracy and consistency of the system by correcting the data acquisition module and the error calculation and priority allocation module. The system uses high-precision cameras and edge detection algorithms, combined with temperature sensors, humidity sensors, IMU sensors, etc., to collect real-time image data and edge correction data of the bearing, ensuring accurate detection of the surface features of the bearing. By calculating the edge correction error value Ei and intelligently allocating the correction priority Pi, the system can prioritize the correction of bearings with large errors to ensure that each bearing reaches the ideal state. Compared with traditional manual correction methods, this system has a high degree of automation, can eliminate human operation errors, ensure more consistent correction results, and greatly improve the reliability of production quality.

[0065] (2) The system uses a multi-bearing synchronous correction scheduling module, which greatly improves production efficiency and resource utilization. The scheduling allocation analysis unit can integrate the edge correction error value Ei, the correction priority Pi and the idle state of the workstation Wfree to construct an intelligent scheduling allocation algorithm formula to ensure that multiple bearings can be corrected synchronously. The scheduling allocation evaluation unit can be combined with the scheduling score D i,j By comparing the workstations, the correction tasks are arranged to idle workstations first, thus avoiding the waste of correction workstation resources. The priority division of scheduling scores also ensures that high-priority bearings can be corrected in time, while low-priority bearings are arranged to be corrected when the workstations are idle, thus optimizing the entire production process, shortening the correction cycle, and significantly improving the operating efficiency of the production line.

[0066] (3) The system can realize intelligent feedback and adaptive optimization through the correction optimization adjustment module and the feedback and adaptive adjustment module. The system constructs a correction path optimization formula at each correction station, calculates the correction path time Tfix, and performs dynamic correction optimization based on the bearing rotation angle error Jd. After the correction is completed, the system will perform a secondary evaluation of the residual error Eres. If the error exceeds the edge correction threshold F2, the secondary correction will be automatically started. The module dynamically adjusts the correction force and path through the adaptive feedback adjustment value Fadj to ensure that the system can automatically adjust the correction strategy according to environmental changes after each correction to improve the correction accuracy. In addition, the overall evaluation and optimization module continuously monitors and optimizes the system's correction efficiency Etotal through the correction efficiency algorithm formula to ensure that each correction task achieves the best effect, avoids correction failure or insufficient accuracy, and improves the overall production quality and the intelligence level of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 The present invention is a flow chart of a bearing edge correction system based on machine vision. DETAILED DESCRIPTION

[0068] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0069] Example 1

[0070] See also Figure 1 The present invention provides a bearing edge correction system based on machine vision. To achieve the above purpose, the present invention is implemented through the following technical solutions: including a correction data acquisition module, an error calculation and priority allocation module, a multi-bearing synchronous correction scheduling module, a correction optimization adjustment module, a feedback and adaptive adjustment module and an overall evaluation and optimization module;

[0071] The modified data acquisition module collects the image data and edge-smoothing data of the bearing through machine vision and an integrated sensor group, and pre-processes to obtain a standard image data set and a standard edge-smoothing data set, and then performs feature extraction on the standard image data set to obtain a bearing surface feature set;

[0072] The error calculation and priority allocation module is used to construct the error calculation formula, extract the bearing surface feature set and input it into the error calculation formula, calculate and output the whole edge error value Ei, and construct the bearing correction priority formula based on the whole edge error value Ei, calculate and output the correction priority Pi, and determine the priority correction order;

[0073] The multi-bearing synchronous correction scheduling module is used to construct an intelligent scheduling allocation algorithm formula. The edge error value Ei and the correction priority Pi are combined with the station idle state Wfree, and input into the intelligent scheduling allocation algorithm formula to calculate and output the scheduling score D i，j , and preset the scheduling threshold F1 and the obtained scheduling score D i，j Conduct preliminary comparative assessment to determine the dispatch situation;

[0074] The correction optimization adjustment module constructs a correction path optimization formula at each bearing edge correction station, calculates and outputs the correction path time Tfix, optimizes the motion trajectory and force of the correction tool for bearing edge correction, and then performs a secondary execution of the error calculation formula after correction optimization to calculate and output the corrected residual error value Eres, and then performs a secondary comparative evaluation between the preset edge correction threshold F2 and the obtained residual error value Eres;

[0075] The feedback and adaptive adjustment module performs a secondary correction feedback formula based on the secondary evaluation results to calculate and output the adaptive feedback adjustment value Fadj, performs secondary correction on the residual bearing trimming after correction optimization, and adjusts the strategy in real time;

[0076] The overall evaluation and optimization module is used to construct the correction efficiency algorithm formula and calculate the output correction efficiency Etotal after the secondary correction is completed. The correction efficiency threshold F3 is preset and compared with the correction efficiency Etotal for three times to analyze the error changes of the bearing edge correction and determine the optimization effect of scheduling and correction.

[0077] In this embodiment, the system uses the correction data acquisition module with the help of machine vision and sensor technology to accurately collect and process the image and trimming data of the bearing, generate a standardized data set, and provide a reliable basis for subsequent correction. Then, the error calculation and priority allocation module accurately calculates the trimming error Ei of the bearing and allocates the correction priority Pi by constructing the error calculation formula and priority algorithm, ensuring that key tasks are given priority. The multi-bearing synchronous correction scheduling module introduces an intelligent scheduling algorithm, combines the station status and error data, optimizes the task scheduling, and improves the synchronization and resource utilization of multi-station correction. In the specific correction process, the correction optimization adjustment module calculates the correction time Tfix based on the correction path optimization formula, adjusts the correction path and strength, and ensures the correction accuracy through the secondary evaluation of the residual error Eres. When an error occurs, the feedback and adaptive adjustment module dynamically corrects the path and strength through the adaptive adjustment formula Fadj to further improve the correction effect. Finally, the overall evaluation and optimization module comprehensively analyzes the quality and efficiency of the correction task through the calculation and multiple evaluation of the correction efficiency Etotal, and continuously optimizes the system performance. Compared with the prior art, the present invention has made significant progress in correction accuracy, efficiency and automation. Traditional technologies often rely on manual or low-intelligence systems, resulting in large correction errors and uneven resource utilization. However, this system reduces correction time, improves accuracy and consistency, and realizes a highly automated bearing edge correction process through intelligent scheduling and adaptive feedback. Ultimately, the present invention effectively improves production efficiency, reduces resource waste, and ensures high-quality correction results.

[0078] Example 2

[0079] This embodiment is explained in Example 1, please refer to Figure 1 ,Specifically: the modified data acquisition module includes a data acquisition unit and a feature extraction unit;

[0080] The data acquisition unit includes an image acquisition unit, an environment acquisition unit and a preprocessing unit;

[0081] The image acquisition unit acquires image data of the integral bearing through a machine vision camera;

[0082] The environment acquisition unit is used to install temperature sensors, humidity sensors, workstation status sensors and IMU sensors in the bearing trimming correction environment to collect trimming data in real time. The trimming data includes temperature Tenv, humidity Henv, workstation idle state Wfree and bearing rotation angle error Jd.

[0083] The preprocessing unit is used to perform standardization processing on the collected image data and edge-corrected data to obtain a standard image data set and a standard edge-corrected data set;

[0084] The standard image data set normalizes the image data and automatically scales the image data to the same size using a graphic editing tool. The bearing image is then captured using a high-precision camera. The actual edge coordinates (Xi, Yi) are obtained by combining edge detection algorithms, sub-pixel precision technology, camera calibration parameters, and pixel-to-physical coordinate conversion.

[0085] The standard image dataset includes bearing illumination reflectance images;

[0086] The standard edge-corrected data set is obtained by cleaning and denoising the collected edge-corrected data, and then normalizing it. At the same time, the change rate of temperature Tenv and humidity Hev is calculated to obtain the temperature change rate △Tenv and the humidity change rate △Henv;

[0087] The temperature change rate △Tenv and the humidity change rate △Henv are calculated and obtained by the following algorithm formulas respectively;

[0088] ;

[0089] Where, Tenv(t1) represents the temperature value at the previous time point, Tenv(t2) represents the temperature value at the next time point, Hev(t1) represents the humidity value at the previous time point, and Hev(t2) represents the humidity value at the next time point;

[0090] The feature extraction unit performs feature extraction to obtain a bearing surface feature set based on the standard image data set obtained after preprocessing. The bearing surface feature set includes a bearing surface roughness feature Rs and a bearing material uniformity feature Mu.

[0091] The bearing surface roughness feature Rs is processed by edge detection image processing algorithm on the light reflection image captured in the standard image data set, the brightness value of each pixel in the light reflection image is extracted, the reflectivity distribution of the brightness value of each pixel is calculated, and then the reflectivity distribution and the known incident angle of the light source are calculated and obtained;

[0092] The uniformity characteristic Mu of the bearing material is obtained by comparing the spectral data of different areas in the light reflection image and calculating the deviation between the spectral curve of each area and the photovoltaic curve straight line of the standard bearing material.

[0093] In this embodiment, the system uses an image acquisition unit to collect image data of the bearing with high precision through a machine vision camera, and combines edge detection algorithms and sub-pixel precision technology to accurately obtain the actual edge coordinates Xi, Yi of the bearing, providing an accurate data basis for subsequent correction. The environment acquisition unit uses temperature sensors, humidity sensors, station status sensors and IMU sensors to collect temperature and humidity, station idle state Wfree and rotation angle error Jd in the correction environment in real time, ensuring that the correction process can adapt to changes in different environmental conditions. The preprocessing unit performs standardization processing on all collected data, cleans noise data and normalizes them to ensure the consistency and reliability of subsequent algorithm processing. Through the feature extraction unit, the system can extract the surface roughness feature Rs and material uniformity feature Mu of the bearing, which are crucial to the accuracy of bearing correction. Compared with traditional correction systems, the present invention can capture the various characteristics of the bearing more accurately and in real time through high-precision sensors and data processing algorithms, providing reliable data support for subsequent error calculation and correction decisions. Traditional correction systems often rely on manual or low-precision sensors, and cannot accurately obtain the real-time state of the bearing, resulting in insufficient correction accuracy and consistency. This system achieves more intelligent bearing correction operations through improved data collection and processing methods, greatly improving correction accuracy, consistency and overall production efficiency.

[0094] Example 3

[0095] This embodiment is explained in Example 2. Please refer to Figure 1 ,Specifically: the error calculation and priority allocation module includes an error calculation unit and a priority calculation unit;

[0096] The error calculation unit is used to construct an error calculation formula, and to extract the actual edge coordinates (Xi, Yi), the bearing surface feature set, and the standard edge correction data set in real time and input them into the error calculation formula to calculate and output the edge correction error value Ei;

[0097] The edge error value Ei is calculated and output by the following error calculation formula;

[0098] ;

[0099] Wherein, Ei represents the integral error value of the i-th integral bearing, Xideal represents the i-th coordinate point on the ideal horizontal axis, Yideal represents the i-th coordinate point on the ideal vertical axis, k1 represents the adjustment coefficient of the roughness feature of the bearing surface, and k2 represents the adjustment coefficient of the bearing rotation angle error.

[0100] The priority calculation unit constructs a priority algorithm formula, extracts the bearing surface feature set and the standard edge correction data set in real time, and calculates and outputs the correction priority Pi;

[0101] The corrected priority Pi is calculated and output by the following priority algorithm formula;

[0102] ;

[0103] Where Pi represents the correction priority of the i-th bearing.

[0104] In this embodiment, the system uses the error calculation unit to input the constructed error calculation formula based on the real-time extracted bearing actual edge coordinates (Xi, Yi), the bearing surface roughness features Rs and the rotation angle error Jd, and outputs the edge correction error value Ei to provide accurate error data for the correction process. The priority calculation unit extracts the bearing surface features and the standard edge correction data set, combines the correction priority formula, and outputs the correction priority Pi to ensure that key tasks are processed first. Compared with traditional technical means, this module significantly improves the intelligence and automation level of correction tasks. Traditional technologies usually rely on fixed-order correction operations and lack a reasonable allocation of correction priorities for each bearing, resulting in resource waste and low correction efficiency. However, this system ensures that bearings with large errors can be corrected first through intelligent error calculation and priority allocation, reducing unnecessary waiting time and resource waste, and optimizing the production process. Through this module, the system not only improves the correction accuracy and efficiency, but also significantly improves the rationality of resource allocation, realizes efficient and accurate multi-bearing synchronous correction, and greatly improves the overall operation efficiency of the production line.

[0105] Example 4

[0106] This embodiment is explained in Example 3, please refer to Figure 1 ,Specifically: the multi-bearing synchronous correction scheduling module includes a ,scheduling allocation analysis unit and a scheduling allocation evaluation unit;

[0107] The scheduling and allocation analysis unit is used to construct an intelligent scheduling and allocation algorithm formula. The side error values ​​Ei and correction priorities Pi of all side bearings are input into the intelligent scheduling and allocation algorithm formula to calculate and output the scheduling score D. i，j , score each full-edge bearing and correction station;

[0108] Scheduling score D i，j Through the following intelligent scheduling allocation algorithm formula;

[0109] ;

[0110] Where Di，j represents the scheduling score for assigning the i-th bearing to the j-th correction station, Wfee represents the idle state of the station, 1 represents idle, 0 represents occupied, π represents the circumference of a circle, which is 3.14, Vmax represents the upper limit of the speed of the correction tool, and Emax represents the set upper limit of the correctable error. represents the length of the corrected path after optimization, Indicates the correction speed, Represents the weight value of the integral edge error value Ei, Indicates the weight value of the bearing surface roughness feature Rs;

[0111] The scheduling allocation evaluation unit sets the scheduling threshold F1 based on the historical data set and production target pre-acquired by the system, and then compares it with the obtained scheduling score D i，j Conduct a preliminary comparative assessment to determine the scheduling situation. The specific assessment contents are as follows;

[0112] When the scheduling score is D i，j > scheduling threshold F1, it is marked as the first priority, indicating that the correction priority Pi of the bearing is higher, the correction time estimate Test is shorter, or the edge error value Ei is relatively small, indicating that the correction process will not take too much time, indicating that the bearing needs to be corrected as soon as possible, and the i-th bearing is immediately assigned to the j-th workstation;

[0113] When the scheduling score is D i，j ≤ scheduling threshold F1, it is marked as the second priority and no workstation is assigned. At the same time, the i-th bearing is arranged to the first priority and sorted. When the first priority bearing is corrected, the second priority bearing is arranged to the workstation according to the sorting.

[0114] In this embodiment, the system constructs an intelligent scheduling algorithm formula according to the trimming error value Ei and the correction priority Pi of each bearing through the scheduling allocation analysis unit, calculates the scheduling score Di,j, and accurately evaluates the matching degree of each bearing with the correction station. The system ensures that the bearing can be corrected at the optimal station and conditions by combining the idle state Wfree of the station and the performance upper limit of the correction tool. The scheduling allocation evaluation unit sets the scheduling threshold F1 based on historical data, and prioritizes the correction tasks with high priority by comparing the scheduling score with the threshold, ensuring the rationality of resource allocation and the efficiency of the correction process. This system has achieved significant improvements in scheduling efficiency and resource utilization. The traditional system lacks priority evaluation and intelligent scheduling of correction tasks, which often leads to idle stations or waste of resources and low correction efficiency. The present invention ensures that each bearing can be promptly and efficiently assigned to the most suitable correction station through an intelligent scheduling algorithm and a scoring system, reducing waiting time and waste of resources, and improving the coordination and efficiency of the entire production process. Finally, this module realizes the synchronous correction of multiple bearings, ensures the smooth progress of correction tasks, and greatly improves the operation efficiency and production capacity of the production line.

[0115] Example 5

[0116] This embodiment is explained in Example 4. Please refer to Figure 1 ,Specifically: the correction optimization and adjustment module includes a correction path optimization unit and a correction error residual analysis unit;

[0117] The correction path optimization unit is used to construct the correction path optimization formula at each edge correction station, and then calculate the time Tfix of the output correction path according to the bearing rotation angle error Jd, and optimize the motion trajectory and force of the correction tool;

[0118] The time Tfix of the corrected path is calculated and obtained by the corrected path optimization formula;

[0119] ;

[0120] Where N represents the number of error points that need to be corrected, n represents the index of the error point, which is used to traverse N error points that need to be corrected, and Jd n Indicates the bearing rotation angle error Jd at the nth error point during the current bearing correction process, Jd n，ideal represents the ideal rotation angle of the ith error point, k3 represents the rotation correction coefficient, and fi represents the correction force of the bearing edge correction tool, which is automatically adjusted to cope with the unevenness of the bearing material.

[0121] The correction error residual analysis unit is used to execute the error calculation formula for the second time to calculate and output the corrected residual error value Eres after the path correction optimization, and preset the edge correction threshold F3 based on the bearing edge correction error standard, and perform a second comparative evaluation with the obtained residual error value Eres to analyze the error of the bearing edge correction after the path correction optimization. The specific evaluation contents are as follows;

[0122] When the residual error value Eres> the edge correction threshold F2, it means that there is an error in the bearing edge correction after the path correction optimization, and the secondary correction is performed at this time;

[0123] When the residual error value Eres ≤ the edge correction threshold F2, it means that after the path correction optimization, there is no error in the bearing edge correction, and the processing is completed.

[0124] In this embodiment, the system constructs an optimization formula through the correction path optimization unit, dynamically adjusts the motion trajectory and force of the correction tool according to the rotation angle error Jd of the bearing and the uniformity of the material, calculates the correction path time Tfix, and ensures that the correction tool is corrected along the optimal path, reducing the correction time and tool wear. After the path correction is completed, the correction error residual analysis unit further calculates the residual error Eres and compares it with the preset correction threshold F2 to ensure that the final correction result meets the quality standard. If the residual error exceeds the threshold, the system will automatically perform a secondary correction to further improve the correction accuracy.

[0125] Example 6

[0126] This embodiment is explained in Example 5. Please refer to Figure 1 Specifically: the feedback and adaptive module is triggered when there is an error in the bearing edge correction after the path correction optimization. It calculates and outputs the adaptive feedback adjustment value Fadj by constructing a secondary correction feedback formula to dynamically adjust the correction force and path;

[0127] The adaptive feedback adjustment value Fadj is calculated and outputted by the following quadratic correction feedback formula;

[0128] ;

[0129] In the formula, Represents the adaptation coefficient of the residual error value Eres, Represents the environmental adaptation coefficient; used to control the adjustment of the correction strength.

[0130] In this embodiment, when the system identifies that there is an error in the bearing trimming correction, the system automatically triggers the adaptive feedback mechanism, calculates the adaptive feedback adjustment value FadjF according to the secondary correction feedback formula, and dynamically adjusts the correction force and path. This module not only takes into account the residual error, but also combines the environmental adaptive coefficient, such as changes in temperature and humidity, to ensure that the correction force and path adjustment can adapt to different environmental conditions, thereby ensuring the stability of the correction quality.

[0131] Example 7

[0132] This embodiment is explained in Example 6. Please refer to Figure 1 ,Specifically: the overall evaluation and optimization module includes an overall ,correction efficiency analysis unit and an overall correction efficiency evaluation unit;

[0133] The overall correction efficiency analysis unit is used to construct the correction efficiency algorithm formula and calculate the output correction efficiency Etotal after the secondary correction is completed, and analyze the overall correction efficiency of all bearings;

[0134] The corrected efficiency Etotal is calculated by the following corrected efficiency algorithm formula;

[0135] ;

[0136] Where, I represents the total number of full-edge bearings, Ei fin represents the error value of the i-th bearing after correction, and Ttotal represents the total running time of the system.

[0137] The overall correction efficiency evaluation unit performs three comparative evaluations based on the correction efficiency threshold F3 preset by the user and the obtained correction efficiency Etotal to measure the overall correction effect of the system. The specific evaluation contents are as follows;

[0138] When the correction efficiency Etotal ≤ the correction efficiency threshold F3, it means that the bearing edge correction effect is normal and the correction is completed;

[0139] When the correction efficiency Etotal>correction efficiency threshold F3, it indicates that the bearing edge correction effect is abnormal. At this time, the iterative correction optimization adjustment module is performed until the bearing edge correction effect is normal, and the iteration is terminated.

[0140] In this embodiment, the system uses the overall correction efficiency analysis unit to calculate the correction efficiency Etotal using the correction efficiency algorithm formula after completing the secondary correction, comprehensively considering the corrected error value Ei of each bearing. finand the total system operating time Ttotal, providing a quantitative evaluation of the overall correction process of the system. The overall correction efficiency evaluation unit performs three comparative evaluations with the correction efficiency threshold F3 preset by the user to ensure that the correction efficiency is within an acceptable range. When Etotal exceeds the correction efficiency threshold F3, the system will automatically trigger iterative correction to further optimize the correction process until the ideal effect is achieved. Compared with traditional correction systems, the advantage of this module lies in its globality and dynamism. Traditional technologies often lack the means to evaluate global efficiency, resulting in high local correction efficiency and low overall efficiency, making it difficult to fully optimize the correction process. This system, by introducing a global efficiency evaluation mechanism, achieves dynamic optimization of the correction process, ensuring a balance between the correction results of each bearing and the overall production efficiency.

[0141] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A bearing edge correction system based on machine vision, characterized by: It includes correction data acquisition module, error calculation and priority allocation module, multi-bearing synchronous correction scheduling module, correction optimization adjustment module, feedback and adaptive adjustment module and overall evaluation and optimization module; The correction data acquisition module collects the image data and edge-straightening data of the bearing through machine vision and an integrated sensor group, and pre-processes to obtain a standard image data set and a standard edge-straightening data set, and then performs feature extraction on the standard image data set to obtain a bearing surface feature set; The error calculation and priority allocation module is used to construct an error calculation formula, extract a bearing surface feature set and input it into the error calculation formula, calculate and output the edge error value Ei, and construct a bearing correction priority formula based on the edge error value Ei, calculate and output the correction priority Pi, and determine the priority correction order; The multi-bearing synchronous correction scheduling module is used to construct an intelligent scheduling allocation algorithm formula, which combines the edge error value Ei and the correction priority Pi with the station idle state Wfree and inputs them into the intelligent scheduling allocation algorithm formula to calculate the output scheduling score D i，j , and preset the scheduling threshold F1 and the obtained scheduling score D i，j Conduct preliminary comparative assessment to determine the dispatch situation; The correction optimization adjustment module constructs a correction path optimization formula at each bearing trimming correction station, calculates and outputs the correction path time Tfix, optimizes the motion trajectory and force of the correction tool for the bearing trimming, and then performs a secondary execution of the error calculation formula after the correction optimization to calculate and output the corrected residual error value Eres, and then performs a secondary comparison evaluation between the preset trimming correction threshold F2 and the obtained residual error value Eres; The feedback and adaptive adjustment module performs a secondary correction feedback formula based on the secondary comparison evaluation result to calculate and output an adaptive feedback adjustment value Fadj, performs secondary correction on the residual bearing trimming after correction optimization, and adjusts the strategy in real time; The overall evaluation and optimization module is used to construct a correction efficiency algorithm formula and calculate the output correction efficiency Etotal after the secondary correction is completed, and to preset the correction efficiency threshold F3 and the correction efficiency Etotal for three comparative evaluations, analyze the error changes of the bearing edge correction, and judge the optimization effect of scheduling and correction.

2. According to claim 1, a machine vision-based bearing edge correction system is characterized in that: The correction data acquisition module includes a data acquisition unit and a feature extraction unit; The data acquisition unit includes an image acquisition unit, an environment acquisition unit and a preprocessing unit; The image acquisition unit acquires image data of the full-edge bearing through a machine vision camera; The environment acquisition unit is used to install a temperature sensor, a humidity sensor, a station status sensor and an IMU sensor in the bearing trimming correction environment to collect trimming data in real time. The trimming data includes temperature Tenv, humidity Henv, station idle state Wfree and bearing rotation angle error Jd; The preprocessing unit is used to perform standardization processing on the collected image data and edge-corrected data to obtain a standard image data set and a standard edge-corrected data set; The standard image data set normalizes the image data and automatically scales the image data to the same size using a graphic editing tool. The bearing image is then photographed using a high-precision camera, and the actual edge coordinates (Xi, Yi) are obtained by combining edge detection algorithms, sub-pixel precision technology, camera calibration parameters, and pixel-to-physical coordinate conversion. The standard image data set includes bearing illumination reflection images; The standard edge-corrected data set is obtained by performing data cleaning and denoising on the collected edge-corrected data, and then performing normalization processing, and at the same time calculating the change rate of temperature Tenv and humidity Hev to obtain the temperature change rate △Tenv and the humidity change rate △Henv; The feature extraction unit performs feature extraction to obtain a bearing surface feature set based on the standard image data set obtained after preprocessing, wherein the bearing surface feature set includes a bearing surface roughness feature Rs and a bearing material uniformity feature Mu; The bearing surface roughness feature Rs is obtained by processing the light reflection image captured in the standard image data set through an edge detection image processing algorithm, extracting the brightness value of each pixel in the light reflection image, calculating the reflectivity distribution of the brightness value of each pixel, and then calculating based on the reflectivity distribution and the known incident angle of the light source; The bearing material uniformity feature Mu is obtained by comparing the spectral data of different regions in the light reflection image and calculating the deviation between the spectral curve of each region and the photovoltaic curve straight line of the standard bearing material.

3. The machine vision-based bearing edge correction system according to claim 2, characterized in that: The error calculation and priority allocation module includes an error calculation unit and a priority calculation unit; The error calculation unit is used to construct an error calculation formula, and extract the actual edge coordinates (Xi, Yi), the bearing surface feature set and the standard edge correction data set in real time and input them into the error calculation formula to calculate and output the edge correction error value Ei; The edge correction error value Ei is calculated and outputted by the following error calculation formula: ； Wherein, Ei represents the integral error value of the i-th integral bearing, Xideal represents the i-th coordinate point on the ideal horizontal axis, Yideal represents the i-th coordinate point on the ideal vertical axis, k1 represents the adjustment coefficient of the roughness feature of the bearing surface, and k2 represents the adjustment coefficient of the bearing rotation angle error.

4. The machine vision-based bearing edge correction system according to claim 3 is characterized in that: The priority calculation unit constructs a priority algorithm formula, extracts the bearing surface feature set and the standard edge correction data set in real time, and calculates and outputs the correction priority Pi; The modified priority Pi is calculated and output by the following priority algorithm formula; ； Where Pi represents the correction priority of the i-th bearing.

5. The bearing edge correction system based on machine vision according to claim 2 is characterized in that: The multi-bearing synchronous correction scheduling module includes a scheduling allocation analysis unit and a scheduling allocation evaluation unit; The scheduling allocation analysis unit is used to construct an intelligent scheduling allocation algorithm formula, input the edge error value Ei and the correction priority Pi of all edge-aligned bearings into the intelligent scheduling allocation algorithm formula, and calculate the output scheduling score D i，j , score each full-edge bearing and correction station; The scheduling score D i，j Through the following intelligent scheduling allocation algorithm formula; ； Where D i，j represents the scheduling score for assigning the i-th bearing to the j-th correction station, Wfee represents the idle state of the station, 1 represents idle, 0 represents occupied, π represents the circumference of a circle, which is 3.14, Vmax represents the upper limit of the speed of the correction tool, and Emax represents the set upper limit of the correctable error. represents the optimized corrected path length, Indicates the correction speed, Represents the weight value of the integral edge error value Ei, Indicates the weight value of the bearing surface roughness feature Rs; The scheduling allocation evaluation unit sets the scheduling threshold F1 based on the historical data set and production target pre-acquired by the system, and then compares it with the obtained scheduling score D i，j Conduct a preliminary comparative assessment to determine the scheduling situation. The specific assessment contents are as follows; When the scheduling score is D i，j > scheduling threshold F1, it is marked as the first priority, and the i-th bearing is assigned to the j-th workstation; When the scheduling score is D i，j ≤ scheduling threshold F1, it is marked as the second priority and no workstation is assigned. At the same time, the i-th bearing is arranged to the first priority and sorted. When the first priority bearing is corrected, the second priority bearing is arranged to the workstation according to the sorting.

6. The machine vision-based bearing edge correction system according to claim 5, characterized in that: The correction optimization adjustment module includes a correction path optimization unit and a correction error residual analysis unit; The correction path optimization unit is used to construct a correction path optimization formula at each edge correction station, and then calculate the time Tfix of the output correction path according to the bearing rotation angle error Jd, so as to optimize the motion trajectory and strength of the correction tool; The time Tfix of the corrected path is obtained by calculating the corrected path optimization formula; ； Where N represents the number of error points that need to be corrected, n represents the index of the error point, which is used to traverse N error points that need to be corrected, and Jd n Indicates the bearing rotation angle error Jd at the nth error point during the current bearing correction process, Jd n，ideal represents the ideal rotation angle of the i-th error point, k3 represents the rotation correction coefficient, and fi represents the correction force of the bearing edge correction tool.

7. The machine vision-based bearing edge correction system according to claim 6, characterized in that: The correction error residual analysis unit is used to execute the error calculation formula for the second time to calculate and output the corrected residual error value Eres after the path correction optimization, and preset the edge correction threshold F2 based on the bearing edge error standard, and perform a second comparison evaluation with the obtained residual error value Eres to analyze the error of the bearing edge after the path correction optimization. The specific evaluation content is as follows; When the residual error value Eres> the edge correction threshold F2, it means that there is an error in the bearing edge correction after the path correction optimization, and the secondary correction is performed at this time; When the residual error value Eres ≤ the edge correction threshold F2, it means that after the path correction optimization, there is no error in the bearing edge correction, and the processing is completed.

8. The machine vision-based bearing edge correction system according to claim 7, characterized in that: The feedback and adaptive module is triggered when there is an error in the bearing edge correction after the path correction optimization, and calculates and outputs the adaptive feedback adjustment value Fadj by constructing a secondary correction feedback formula to dynamically adjust the correction force and path; The adaptive feedback adjustment value Fadj is calculated and outputted by the following quadratic correction feedback formula: ； In the formula, Represents the adaptation coefficient of the residual error value Eres, Represents the environmental adaptation coefficient; used to control the adjustment of the correction strength.

9. The bearing edge correction system based on machine vision according to claim 1, characterized in that: The overall evaluation and optimization module includes an overall correction efficiency analysis unit and an overall correction efficiency evaluation unit; The overall correction efficiency analysis unit is used to construct a correction efficiency algorithm formula and calculate the output correction efficiency Etotal after the secondary correction is completed, and analyze the overall correction efficiency of all bearings; The modified efficiency Etotal is calculated and obtained by the following modified efficiency algorithm formula; ； Where, I represents the total number of full-edge bearings, Ei fin represents the error value of the i-th bearing after correction, and Ttotal represents the total running time of the system.

10. The bearing edge correction system based on machine vision according to claim 9, characterized in that: The overall correction efficiency evaluation unit performs three comparative evaluations based on the correction efficiency threshold F3 preset by the user and the obtained correction efficiency Etotal to measure the overall correction effect of the system. The specific evaluation contents are as follows; When the correction efficiency Etotal ≤ the correction efficiency threshold F3, it means that the bearing edge correction effect is normal and the correction is completed; When the correction efficiency Etotal>correction efficiency threshold F3, it indicates that the bearing edge correction effect is abnormal. At this time, the iterative correction optimization adjustment module is performed until the bearing edge correction effect is normal, and the iteration is terminated.

Images