Automatic deviation correction control method and system for shaft box cleaning and polishing station

Abstract

The application discloses an automatic deviation rectification control method and system for shaft box cleaning and polishing stations, and relates to the technical field of industrial visual detection and automatic control. The method comprises the following steps: identifying the workpiece model and loading corresponding standard template parameters through an RFID reader; collecting shaft box feature area images through a visual detection unit; extracting reference surface contour lines and positioning hole center coordinates through edge detection and Hough circle transformation; matching the extracted features with the standard template to calculate translation deviation and rotation deviation; converting the deviation into an actuator compensation correction amount through coordinate transformation; and driving the actuator to complete deviation rectification through closed-loop iterative control. The application solves the problems of secondary positioning cumulative error, tool tolerance superposition and poor model switching adaptability in the multi-station circulation of shaft boxes, and the deviation rectification accuracy is better than ±0.2mm and ±0.1°, and the single deviation rectification time is less than 3 seconds.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial visual inspection and automatic control, and particularly relates to an automatic deviation correction control method and system for a shaft box cleaning and polishing station. BACKGROUND

[0002] The shaft box is a key component of the bogie of a rail transit vehicle, and the machining precision thereof directly affects the safety and stability of the vehicle in operation. In the production and machining process of the shaft box, cleaning and polishing are two important processes. The cleaning process is used to remove oil stains, cuttings and oxide scales on the surface of a blank or semi-finished product, and the polishing process is used to perform precision machining on the key surface of the shaft box to meet the assembly precision requirements. Modern rail transit vehicle manufacturing enterprises usually adopt an automatic production line to realize multi-station flow machining of the shaft box. The workpiece is transmitted between cleaning stations, polishing stations and other machining stations through a tooling tray.

[0003] In the prior art, a workpiece surface detection method based on visual inspection and related devices are disclosed in Chinese patent application No. CN116071348A. The method detects the workpiece polishing quality by respectively constructing a side surface polishing detection model and an outer curved surface polishing detection model, and adjusts the parameters of the workpiece polishing equipment according to the detection results. The prior art mainly focuses on the visual inspection of polishing quality and the adjustment of equipment parameters, and uses a deep learning model to identify and locate polishing defects. However, the prior art has the following deficiencies: first, the method focuses on quality detection after polishing and does not involve positioning and deviation correction of the workpiece during the flow process between stations, and cannot solve the cumulative error caused by secondary positioning; second, the method uses a single workpiece template image for feature point matching and does not consider the feature differences of different types of workpieces. When the production line needs to produce multiple types of workpieces, the template needs to be manually switched; third, the detection results of the method are used to adjust the force parameters of the polishing equipment, and do not form a closed-loop control with the station positioning system, and cannot realize active pose correction before machining.

[0004] In actual production, the shaft box body faces the following technical problems when flowing between multiple stations. First, the problem of cumulative error of secondary positioning. When the workpiece is transferred from one station to the next, due to the limited positioning accuracy of the transmission mechanism and the gap between the tooling and the workpiece, each transfer introduces a new positioning error. These errors gradually accumulate during the multi-station flow process, eventually leading to a significant deviation of the machining trajectory from the actual position of the workpiece. Second, the problem of tolerance superposition of tooling fixtures. Different batches of tooling trays have manufacturing tolerances, and tooling trays of the same batch will also deform after long-term use. These tolerance superpositions will cause differences in the positioning reference of the workpiece on different tooling, and in turn cause quality problems of uneven polishing. Third, the problem of poor adaptability of traditional mechanical positioning methods. Existing mechanical limit positioning methods are designed for specific models of workpieces. When the production line needs to switch to producing different models of shaft box bodies, manual replacement of positioning tooling or adjustment of the limiting mechanism is required, which is low in efficiency and prone to human error.

[0005] Therefore, there is an urgent need for a method that can automatically detect the pose deviation of the workpiece at the station entrance, calculate the compensation amount in real time, and drive the actuator to complete the correction action, to solve the above technical problems and improve the positioning accuracy and production efficiency of the multi-station machining of the shaft box body. SUMMARY

[0006] The present application aims to provide an automatic correction control method and system for shaft box body cleaning and polishing stations to solve the technical problems of secondary positioning cumulative error, tooling fixture tolerance superposition, and poor adaptability of traditional mechanical positioning methods during the flow of shaft box bodies between cleaning and polishing stations.

[0007] To achieve the above object, the application provides an automatic deviation correction control method for an axle box body cleaning and polishing station, comprising: a model identification step, in which an RFID reader arranged at the entrance of the station reads the RFID tag data on a tool tray, determines the workpiece model information of a current axle box body according to the RFID tag data, and loads the standard template parameters corresponding to the workpiece model information from a multi-model template library; an image acquisition step, in which a visual detection unit arranged at the entrance of the cleaning station or the entrance of the polishing station vertically downwardly shoots the characteristic area of the axle box body on the tool tray to acquire the characteristic area image; a feature extraction step, in which the edge detection processing is performed on the characteristic area image to extract the reference surface profile line of the axle box body, and the Hough circle transformation processing is performed on the characteristic area image to identify the positioning hole of the axle box body and acquire the positioning hole center coordinates; a deviation calculation step, in which the reference surface profile line and the positioning hole center coordinates are matched and calculated with the standard template parameters to determine the X-direction translation deviation, the Y-direction translation deviation and the rotation deviation around the Z axis of the current axle box body relative to the standard pose; a coordinate transformation step, in which the translation deviation and the rotation deviation are converted into the compensation correction amount of the end of a servo platform or a robot according to the transformation relationship between the coordinate system of the current station and the coordinate system of the servo platform or the base coordinate system of the robot; and a closed-loop deviation correction step, in which the servo platform or the end of the robot is driven to perform the deviation correction action according to the compensation correction amount, the real-time position feedback data of the actuator is acquired, the residual deviation between the current pose and the target pose is calculated, and when the residual deviation is greater than a preset deviation correction accuracy threshold, the residual deviation is fed back to the coordinate transformation step for iterative deviation correction until the residual deviation is less than or equal to the preset deviation correction accuracy threshold.

[0008] Preferably, the preset deviation correction accuracy threshold comprises a position accuracy threshold and an angle accuracy threshold, the value range of the position accuracy threshold is 0.1mm to 0.3mm, and the value range of the angle accuracy threshold is 0.05° to 0.15°.

[0009] Preferably, the visual detection unit comprises an industrial camera and a ring-shaped light source, the resolution of the industrial camera is not less than 5 million pixels, and the color temperature range of the ring-shaped light source is 5500K to 6500K.

[0010] To achieve the above object, the application further provides an automatic deviation correction control system for a shaft box body cleaning and polishing station, comprising: a model identification unit, configured to read RFID tag data on a tool tray through an RFID reader deployed at an entrance of the station, determine workpiece model information of a current shaft box body according to the RFID tag data, and load standard template parameters corresponding to the workpiece model information from a multi-model template library; a visual detection unit, configured to vertically downwardly shoot a feature area of the shaft box body on the tool at an entrance of a cleaning station or an entrance of a polishing station, and acquire a feature area image; a feature extraction unit, configured to perform edge detection processing on the feature area image to extract a reference surface contour line of the shaft box body, and perform Hough circle transformation processing on the feature area image to identify a positioning hole of the shaft box body and acquire a positioning hole center coordinate; a template matching unit, configured to match and calculate the reference surface contour line and the positioning hole center coordinate with the standard template parameters, and determine a translation deviation and a rotation deviation of the current shaft box body relative to a standard pose; a coordinate transformation unit, configured to convert the translation deviation and the rotation deviation into a compensation correction amount according to a transformation relationship between a coordinate system of a current station and a servo platform coordinate system or a robot base coordinate system; and a closed-loop control unit, configured to drive an actuator to perform a deviation correction action according to the compensation correction amount, acquire real-time position feedback data of the actuator, calculate a residual deviation, and feed back the residual deviation to the coordinate transformation unit for iterative deviation correction when the residual deviation is greater than a preset deviation correction accuracy threshold.

[0011] The application has the following beneficial effects:

[0012] Firstly, by respectively deploying visual detection units at entrances of the cleaning station and the polishing station, the pose deviation of the workpiece can be detected in real time before the workpiece enters a machining area, and automatic deviation correction is completed through closed-loop control, so that the influence of secondary positioning cumulative error on machining accuracy is eliminated, and the deviation correction accuracy is better than ±0.2 mm and ±0.1°.

[0013] Secondly, by constructing a multi-model template library and using RFID to automatically identify the current workpiece model, the system can automatically load corresponding template parameters when the workpiece arrives, so that mixed production of workpieces of multiple models can be realized without manual intervention, and the flexibility and model changing efficiency of the production line are improved.

[0014] Thirdly, by forming closed-loop control of the visual detection result and a servo platform or a robot actuator, complete deviation correction closed loop of detection, calculation, execution and feedback is realized, the time consumption of single deviation correction is less than 3 seconds, and the consistency of machining references of stations and the stability of machining quality are ensured. BRIEF DESCRIPTION OF DRAWINGS

[0015] Fig. 1 is a flowchart of the automatic deviation correction control method for the shaft box body cleaning and polishing station provided by the application;

[0016] Fig. 2is a framework diagram of an automatic deviation correction control system of an axle box body cleaning and polishing station provided by an embodiment of the present application. DETAILED DESCRIPTION

[0017] Reference will be made to the drawings in conjunction with the embodiments of the present application, and the technical solutions in the embodiments of the present application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application. Figs. 1-2 Reference will be made to the drawings in conjunction with the embodiments of the present application, and the technical solutions in the embodiments of the present application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0018] Fig. 1 In the embodiments of the present application, an automatic deviation correction control method of an axle box body cleaning and polishing station is provided, which is applied to an automatic production line of rail transit vehicle axle box bodies, and is used to solve the positioning deviation problem of the axle box body when flowing between a cleaning station and a polishing station. In the embodiments, the production line includes the cleaning station and the polishing station arranged in sequence, and the axle box body is carried by a tool tray and transmitted between the stations by a conveying mechanism. The steps of the method will be described in detail below.

[0019] Step S1: model identification step

[0020] In an embodiment of the present application, the model identification step reads the RFID tag data on the tool tray through the RFID reader and writer arranged at the station entrance, determines the workpiece model information of the current axle box body according to the RFID tag data, and loads the standard template parameters corresponding to the workpiece model information from the multi-model template library.

[0021] Specifically, when the tool tray carrying the axle box body reaches the cleaning station entrance or the polishing station entrance, the RFID reader and writer installed at the station entrance automatically senses and reads the RFID electronic tag fixed on the side of the tool tray. The RFID electronic tag stores information such as the workpiece model code, batch number and tool tray number of the current axle box body. In the embodiments, the RFID reader and writer uses ultra-high frequency RFID technology, the working frequency is 920-925 MHz, the reading distance can reach 500-1000 mm, and the non-contact data reading can be completed during the movement of the tool tray, and the reading time is less than 100 ms.

[0022] ​The system retrieves and loads the corresponding standard template parameters from a pre-established multi-model template library according to the read workpiece model code. The multi-model template library is a structured parameter database that stores the standard pose parameters of all producible model shaft housings on the production line. In a preferred embodiment of the present application, the standard template parameters include reference surface contour line standard coordinate sequence, positioning hole standard center coordinates, positioning hole standard radius, and feature point relative position relationship matrix. The reference surface contour line standard coordinate sequence records the pixel coordinates of the shaft housing reference surface boundary in the standard pose, stored in the form of a discrete point sequence with a point spacing of no more than 2 pixels. The positioning hole standard center coordinates record the pixel coordinates of the center of each positioning hole in the standard pose, and a shaft housing usually has 2 to 4 positioning holes. The positioning hole standard radius records the theoretical radius value of each positioning hole, which is used for parameter constraint in subsequent Hough circle transformation. The feature point relative position relationship matrix records the relative distance and angle relationship between each feature point, which is used for matching verification and anomaly detection.

[0023] The core idea of the model adaptive template loading algorithm proposed in the present application is to establish a mapping relationship between the RFID model code and the template parameter set, and to realize automatic template switching when the workpiece arrives. Let the current read RFID model code be , and the template set stored in the multi-model template library be , where each template contains a model code and a corresponding parameter set . The template loading process first retrieves the template that satisfies in the template library, and if the retrieval is successful, the parameter set is loaded into the system working memory for subsequent steps, and if the retrieval fails, an unknown model alarm is triggered. In this embodiment, the template library supports storing template parameters of not less than 50 different model shaft housings, and the template retrieval and loading time is less than 50 ms.

[0024] Through the RFID-driven multi-model template dynamic loading mechanism, the present application solves the problem of manual adjustment when switching models in the traditional mechanical positioning method, and realizes the mixed-line automatic production of multi-model workpieces. When the production plan scheduling system arranges the alternate production of different model shaft housings, the system can automatically identify and switch the corresponding positioning parameters without stopping the line for model change, significantly improving the production efficiency.

[0025] Step S2: image acquisition step

[0026] In an embodiment of the present application, the image acquisition step is performed by vertically downwardly shooting the feature area of the shaft housing on the tooling through a visual detection unit deployed at the entrance of the cleaning station or the entrance of the grinding station to obtain the feature area image.

[0027] The hardware components of the visual inspection unit include an industrial camera, a lens, a ring light source, and a mounting bracket. In the present embodiment, the industrial camera uses a planar CMOS sensor with a resolution of 2592x1944 pixels (about 5 million pixels), a frame rate of 15 fps, and a pixel size of 2.2 μm x 2.2 μm. The lens uses a fixed-focus industrial lens with a focal length of 16 mm, an aperture of F2.8, and a working distance of 400 mm to 600 mm. Under this configuration, the actual physical size corresponding to a single pixel is about 0.1 mm, which can meet the detection requirements of a ±0.2 mm correction accuracy. The ring light source uses a white LED array with a color temperature range of 5500 K to 6500 K. Through ring-shaped uniform illumination, the specular reflection interference on the surface of the workpiece can be eliminated, and the contrast of the edge features can be enhanced. The mounting bracket uses a gantry structure to fix the camera and the light source above the work station entrance, ensuring that the camera optical axis is perpendicular to the tool tray plane.

[0028] When the conveying mechanism transports the tool tray to the preset shooting position, the position sensor mounted on the conveying mechanism sends a trigger signal, and the visual inspection unit starts image acquisition according to the trigger signal. The original image obtained by the camera acquisition is an 8-bit grayscale image with an image size of 2592x1944 pixels. In the present embodiment, the feature region refers to the local region of the shaft box body containing the reference surface boundary and the positioning hole. The position of this region in the image is determined according to the workpiece model and the camera installation position in advance. The system crops the feature region image from the original image according to the feature region position information in the standard template parameters loaded in step S1. The cropped image size is usually 1024x1024 pixels to 1600x1200 pixels.

[0029] In a preferred embodiment of the present application, to improve the image quality and the stability of subsequent feature extraction, the image acquisition step further includes an image preprocessing process. The image preprocessing process first performs histogram equalization on the original image to enhance the overall contrast of the image. Then, a bilateral filter is used for noise reduction processing. The bilateral filter can smooth noise while preserving edge information. The filter window size is set to 9x9 pixels, the spatial domain standard deviation is set to 75, and the grayscale domain standard deviation is set to 75. Finally, the filtered image is subjected to grayscale normalization to map the grayscale value to the standard range of 0 to 255.

[0030] Step S3: Feature extraction step

[0031] In an embodiment of the present application, the feature extraction step performs edge detection processing on the feature region image to extract the reference surface profile of the shaft box body, and performs Hough circle transformation processing on the feature region image to identify the positioning hole of the shaft box body and obtain the positioning hole center coordinates.

[0032] The specific process of the reference surface profile line extraction is as follows. First, the Canny edge detection algorithm is used for edge extraction on the pre-processed feature region image. The Canny edge detection algorithm includes four steps of Gaussian filtering, gradient calculation, non-maximum suppression and double threshold processing. In the embodiment, the kernel size of the Gaussian filtering is set to 5x5, and the standard deviation is set to 1.4; the gradient components in the x direction and the y direction are calculated by using the Sobel operator; the non-maximum suppression is used to thin the edges along the gradient direction; the high threshold of the double threshold processing is set to 100, and the low threshold is set to 50, and the ratio of the two is about 2:1. After the Canny edge detection, the output is a binary edge image, and the edge pixel value is 255, and the non-edge pixel value is 0.

[0033] Then, the edge pixel points in the binary edge image belonging to the reference surface boundary region are subjected to straight line fitting. The determination of the reference surface boundary region is based on the preset region of interest range in the standard template parameter loaded in step S1. In the region, the least square method is used for straight line fitting of the edge pixel points, and the straight line equation parameters of the reference surface profile line are obtained. Let the edge pixel point set be , the straight line equation is , and the target of the least square method is to minimize the sum of square fitting errors:

[0034]

[0035] , wherein: is the sum of square fitting errors; is the total number of edge pixel points; is the coordinate of the i-th edge pixel point; is the slope of the straight line; is the intercept of the straight line.

[0036] By taking the partial derivatives of with respect to and and setting them to zero, the optimal slope and intercept can be obtained:

[0037]

[0038]

[0039] When the reference surface of the axle box body contains multiple boundary lines, the straight line fitting is performed on each boundary line to obtain the straight line equation parameter set of the reference surface profile line.

[0040] ​​​​The Hough circle transform algorithm is used to extract the center coordinates of the positioning holes. The Hough circle transform is a classic algorithm for detecting circular targets in images. Its basic principle is to map a circle in image space to a point in parameter space, and then determine the circle's parameters by finding accumulated peak values ​​in the parameter space. For circles in an image... Its center coordinates are , radius is In this embodiment, the gradient-based Hough circle transform method is adopted. This method first uses the gradient direction information of the edge points to determine the possible center positions of the circle, then calculates the cumulative radius value at the candidate center, and finally determines the parameters of the circle by threshold judgment.

[0041] The parameters for the Hough circle transform are set as follows: the minimum center distance is set to 1 / 8 of the image's shorter side size to avoid repeated detection of the same positioning hole; the edge detection threshold is set to 100; the center accumulation threshold is set to 50; the minimum radius is set to 0.8 times the standard radius of the positioning hole in the standard template parameters; and the maximum radius is set to 1.2 times the standard radius. By constraining the radius range, interference circles that are not positioning holes can be effectively filtered out.

[0042] After Hough circle transform, the output is a list of parameters for the detected circles, including the center coordinates for each circle. and radius In this embodiment, a typical axle box has 2 to 4 positioning holes. The system performs preliminary verification against a standard template based on the number and position distribution of the detected circles, eliminating false detections and missed detections.

[0043] Step S4: Deviation Calculation Step

[0044] In one embodiment of the present invention, the deviation calculation step matches and calculates the coordinates of the reference surface contour line and the center of the positioning hole with the standard template parameters to determine the X-direction translational deviation, Y-direction translational deviation and Z-axis rotational deviation of the current shaft box relative to the standard pose.

[0045] The core idea of ​​the multi-feature fusion deviation calculation algorithm proposed in this invention is to comprehensively utilize the features of the reference surface contour line and the positioning hole features, and calculate the workpiece's pose deviation through a two-stage coarse and fine matching strategy. Compared with methods that use only a single feature, multi-feature fusion can improve the robustness and accuracy of matching.

[0046] The first stage of deviation calculation is a coarse estimate of the rotational deviation based on the reference plane profile. Let the slope of the currently detected reference plane profile be... The slope of the reference plane contour line in the standard template is The initial rotational deviation estimate is:

[0047] ,

[0048] wherein: is the initial rotational deviation estimate, in radian; is the current detected datum profile line slope; is the datum profile line slope in the standard template; is the arctangent function.

[0049] The second stage of deviation calculation is the translational deviation calculation based on the positioning hole center coordinates. First, the current detected positioning hole center coordinates are compensated for rotation according to the initial rotational deviation estimate, transforming them to the same pose as the standard template. Let the current detected positioning hole center coordinates be , the compensated coordinates are:

[0050] ,

[0051] ,

[0052] wherein: is the compensated positioning hole center coordinates; is the rotation center coordinates, usually taking the image center or the reference origin defined in the standard template. Then, the positional difference between the compensated positioning hole center coordinates and the standard positioning hole center coordinates is calculated. Let the standard positioning hole center coordinates be

[0053] , the translational deviation corresponding to the positioning hole is: ,

[0054] ,

[0055] ,

[0056] For an axle box body with positioning holes, the comprehensive translational deviation is calculated by weighted average:

[0057] ,

[0058] ,

[0059] wherein: and are the comprehensive translational deviations in X and Y directions, respectively; is the weight coefficient of the th positioning hole, usually determined according to the detection confidence of the positioning hole or the deviation from the standard radius; is the total number of positioning holes. ​​​

[0060] The third stage of the deviation calculation is a fine correction of the rotation deviation. Based on the matching residuals of multiple feature points, an iterative closest point algorithm is used to fine correct the rotation deviation. Let the current rotation deviation estimate be , the translation deviation estimate be , the feature point set include the positioning hole center and the sampling points on the reference surface contour, the objective of the iterative optimization is to minimize the sum of squared distances between the transformed current feature points and the standard template feature points:

[0061] ,

[0062] wherein: is the matching error function; is the total number of feature points; is the homogeneous transformation function containing rotation and translation; is the current detected th feature point coordinate; is the th feature point coordinate in the standard template; is the Euclidean distance norm.

[0063] The above objective function is optimized by gradient descent or Gauss-Newton method, and , and are iteratively updated until convergence or the maximum number of iterations is reached. In the present embodiment, the iteration termination condition is set to be that the parameter variation of the adjacent two iterations is less than or the number of iterations exceeds 20 times.

[0064] The final output deviation includes the X-direction translation deviation , the Y-direction translation deviation and the rotation deviation around the Z-axis , which describe the pose deviation of the current shaft box relative to the standard pose in the tooling plane. It should be noted that since the camera is installed vertically downward and the shaft box is placed on the horizontal tooling tray, the present method only calculates the three degrees of freedom deviations in the plane (X translation, Y translation, Z rotation), and does not involve the deviation calculation of the height direction and the inclination angle.

[0065] Step S5: coordinate transformation step

[0066] In an embodiment of the present application, the coordinate transformation step converts the X-direction translation deviation, the Y-direction translation deviation and the rotation deviation around the Z-axis into the compensation correction amount of the servo platform or the robot end based on the transformation relationship between the coordinate system of the current station and the servo platform coordinate system or the robot base coordinate system.

[0067] In a multi-station automatic production line, the vision coordinate system where the vision detection unit is located and the servo platform coordinate system or robot base coordinate system where the deviation correction action is performed are usually not coincident, and coordinate transformation is needed to convert the deviation amount obtained by vision detection into a compensation instruction executable by the execution mechanism.

[0068] The adaptive coordinate transformation algorithm provided by the application adopts a combination of pre-calibration and online compensation. The pre-calibration stage is completed during system installation and debugging, and the homogeneous transformation matrix from the vision coordinate system to the servo platform coordinate system (or the robot base coordinate system) is determined by measuring the calibration plate. Let the vision coordinate system be , the servo platform coordinate system be , and the homogeneous transformation matrix between the two be:

[0069] ,

[0070] Among them: is the homogeneous transformation matrix from the vision coordinate system to the servo platform coordinate system; is the rotation angle of the vision coordinate system relative to the servo platform coordinate system; and are the translation components in the X direction and the Y direction, respectively.

[0071] The deviation vector in the vision coordinate system is , where and are the components of the translation deviation in the vision coordinate system calculated in step S4. Multiply the deviation vector by the homogeneous transformation matrix to obtain the deviation vector in the servo platform coordinate system:

[0072] ,

[0073] After expansion, we get:

[0074] ,

[0075] ,

[0076] Among them: and are the compensation correction amounts in the X direction and the Y direction in the servo platform coordinate system, respectively.

[0077] For the rotation deviation, since the Z-axis directions of the vision coordinate system and the servo platform coordinate system are the same (both perpendicular to the tooling plane upward), the rotation deviation around the Z-axis in the two coordinate systems is the same, that is:

[0078] ,

[0079] Among them: This is the rotation compensation correction amount in the servo platform coordinate system; The rotational deviation around the Z-axis is calculated in step S4.

[0080] The online compensation phase is used to compensate for coordinate system transformation parameter drift caused by factors such as temperature changes and mechanical wear. In a preferred embodiment of the invention, the system periodically performs online calibration using standard calibration blocks, updating the parameters of the homogeneous transformation matrix by comparing the deviation between the visually detected values ​​and known standard values. The online calibration cycle can be set to once per shift or once per week, depending on the stability of the production environment.

[0081] The output of the coordinate transformation step is the compensation correction amount in the servo platform coordinate system (or robot base coordinate system). This compensation correction is directly used to drive the actuator to perform corrective actions.

[0082] Step S6: Closed-loop correction steps

[0083] In one embodiment of the present invention, the closed-loop correction step drives the servo platform or robot end effector to perform correction actions according to the compensation correction amount, collects real-time position feedback data of the actuator, calculates the residual deviation between the current pose and the target pose, and when the residual deviation is greater than the preset correction accuracy threshold, the residual deviation is fed back to the coordinate transformation step for iterative correction until the residual deviation is less than or equal to the preset correction accuracy threshold.

[0084] The closed-loop iterative correction control algorithm proposed in this invention adopts an incremental correction strategy and an iterative convergence mechanism based on residual deviation feedback to ensure that the correction accuracy meets the preset requirements while avoiding overshoot and oscillation.

[0085] Incremental correction strategy refers to a strategy where the amount of correction performed in each step is not directly calculated using the compensation adjustment amount, but rather the product of the compensation adjustment amount and a preset proportional coefficient. Let the first step be... The compensation correction amount calculated in the next iteration is: The actual amount of correction implemented is:

[0086] ,

[0087] in: For the first The actual amount of correction performed in each iteration; The preset scaling factor ranges from 0.6 to 0.9. In this embodiment, the preset scaling factor is set to 0.8. Using a scaling factor less than 1 can avoid overshoot caused by system errors or detection noise, thus improving the stability of the correction process.

[0088] The actuator drives the tooling pallet or workpiece to adjust its position and orientation according to the correction command. In this embodiment, the actuator uses a three-axis servo platform, including an X-axis linear module, a Y-axis linear module, and a θ-axis rotary table. The X-axis and Y-axis linear modules have a travel of ±50mm, a positioning accuracy of ±0.01mm, and a maximum moving speed of 100mm / s. The θ-axis rotary table has a rotation range of ±5°, a positioning accuracy of ±0.005°, and a maximum rotation speed of 10° / s. All three axes of the servo platform are equipped with high-precision linear scales or encoders for real-time feedback of the current position.

[0089] After the correction action is completed, the system collects real-time position feedback data from the actuator and calculates the residual deviation between the current pose and the target pose. The residual deviation can be calculated in two ways: one is based on encoder feedback from the actuator, using the difference between the correction command and the actual execution amount as the residual deviation; the other is based on secondary visual detection, where images are re-acquired after the correction action, feature extraction and deviation calculation are performed, and the newly calculated deviation value is used as the residual deviation. In a preferred embodiment of the invention, the second method based on secondary visual detection is used. This method can detect error sources other than the actuator's motion error, resulting in higher correction accuracy.

[0090] Let the first The residual error calculated after the second iteration is: The preset correction accuracy threshold includes the position accuracy threshold. and angle accuracy threshold The iteration termination condition is:

[0091] ,

[0092] When the above conditions are met, the correction process ends, the system outputs a correction completion flag, and the workpiece enters the subsequent processing steps. When the above conditions are not met, the residual deviation is fed back to step S5 for coordinate transformation to obtain a new compensation correction amount, and the next iteration of correction is performed.

[0093] In this embodiment, the position accuracy threshold is set to 0.2 mm, and the angle accuracy threshold is set to 0.1°. Under this accuracy requirement, the typical time for a single correction (including image acquisition, feature extraction, deviation calculation, coordinate transformation, and action execution) is 1.5 to 2 seconds. In most cases, the accuracy requirement can be met after 1 to 2 iterations, and the total correction time is less than 3 seconds.

[0094] To prevent infinite iteration in abnormal situations, the closed-loop correction step also sets a maximum iteration limit. When the number of correction iterations exceeds the preset maximum iteration limit (5 times in this embodiment) and the residual deviation is still greater than the preset correction accuracy threshold, the system determines that the correction is abnormal, generates a correction abnormality alarm signal, and suspends the processing flow of the current station, waiting for manual intervention. Abnormal situations may be caused by severe deformation of the workpiece, damage to the tooling, or failure of the vision system, etc.

[0095] Through the above six steps, the present application realizes automatic correction control of the shaft box during the transfer process between the cleaning and grinding stations, effectively eliminates the influence of secondary positioning cumulative error and tooling tolerance superposition, and ensures the consistency of the machining reference of each station.

[0096] In another embodiment of the present application, the application effect of the method in actual production line is tested and verified. The test selects the shaft box cleaning and grinding production line of a certain rail transit vehicle manufacturing enterprise, which has one cleaning station and one grinding station, and produces three types of shaft boxes. The test sample is 100 shaft boxes of each type, a total of 300. The test indicators include correction success rate, average correction time, position correction accuracy and angle correction accuracy.

[0097] The test results show that under the mixed production conditions of the three types of shaft boxes, the correction success rate of the method is 99.7% (only one piece fails to extract features due to severe rust on the surface of the workpiece), and the average correction time is 2.3 seconds. In terms of position correction accuracy, the mean value of X direction translation deviation is 0.08 mm, and the standard deviation is 0.04 mm; the mean value of Y direction translation deviation is 0.09 mm, and the standard deviation is 0.05 mm. In terms of angle correction accuracy, the mean value of rotation deviation around Z axis is 0.05°, and the standard deviation is 0.02°. The pose deviation of all corrected workpieces is less than the preset correction accuracy threshold (position ±0.2 mm, angle ±0.1°), which meets the positioning accuracy requirements of subsequent machining processes.

[0098] Compared with the traditional mechanical positioning method, the method of the present application does not need to manually adjust the positioning tooling when the type is switched, and the type switching time is shortened from about 30 minutes to less than 1 second for the system to automatically switch the template; in terms of secondary positioning accuracy, the cumulative error of the traditional method can reach ±1 mm to ±2 mm, and the method of the present application reduces the error to within ±0.2 mm, which is about 10 times higher.

[0099] Referring to Fig. 2 , the embodiment of the present application also provides an automatic correction control system for shaft box cleaning and grinding station, which is used to execute the automatic correction control method for shaft box cleaning and grinding station described in the above method embodiment.

[0100] The automatic deviation correction control system of the shaft box body cleaning and polishing station comprises a model identification unit 1, a visual detection unit 2, a feature extraction unit 3, a template matching unit 4, a coordinate transformation unit 5 and a closed-loop control unit 6. The functions and connection relationships of the units are described below.

[0101] The model identification unit 1 is used to read the RFID tag data on the tool tray through the RFID reader deployed at the entrance of the station, determine the workpiece model information of the current shaft box body according to the RFID tag data, and load the standard template parameters corresponding to the workpiece model information from the multi-model template library. The hardware components of the model identification unit 1 include an RFID reader and an antenna. In this embodiment, the RFID reader uses ultra-high frequency RFID technology, with a working frequency of 920-925 MHz and supports the ISO18000-6C protocol. The output of the model identification unit 1 is the workpiece model information and the corresponding standard template parameters, which are sent to the template matching unit 4.

[0102] The visual detection unit 2 is used to vertically downwardly shoot the feature area of the shaft box body on the tool at the entrance of the cleaning station or the entrance of the polishing station to obtain a feature area image. As described in the foregoing method embodiment, the hardware components of the visual detection unit 2 include an industrial camera, a lens, a ring light source and a mounting bracket. The output of the visual detection unit 2 is the feature area image, which is sent to the feature extraction unit 3.

[0103] The feature extraction unit 3 is used to perform edge detection processing on the feature area image to extract the reference surface contour line of the shaft box body, and perform Hough circle transformation processing on the feature area image to identify the positioning hole of the shaft box body and obtain the positioning hole center coordinates. The feature extraction unit 3 can be implemented by an industrial computer or an embedded image processor, with built-in edge detection algorithm module and Hough circle transformation algorithm module. The input of the feature extraction unit 3 is the feature area image from the visual detection unit 2, and the output is the reference surface contour line parameters and the positioning hole center coordinates, which are sent to the template matching unit 4.

[0104] The template matching unit 4 is used to match and calculate the reference surface contour line and the positioning hole center coordinates with the standard template parameters to determine the X-direction translation deviation, the Y-direction translation deviation and the rotation deviation around the Z-axis of the current shaft box body relative to the standard pose. As described in the foregoing method embodiment, the template matching unit 4 uses a multi-feature fusion deviation calculation algorithm to accurately calculate the deviation. The input of the template matching unit 4 includes the feature parameters from the feature extraction unit 3 and the standard template parameters from the model identification unit 1, and the output is the translation deviation and the rotation deviation, which are sent to the coordinate transformation unit 5.

[0105] The coordinate transformation unit 5 is used to convert the translation deviation and the rotation deviation into compensation correction amounts according to the transformation relationship between the coordinate system of the current station and the servo platform coordinate system or the robot base coordinate system. As described in the foregoing method embodiment, the coordinate transformation unit 5 is internally provided with the pre-calibrated homogeneous transformation matrix parameters and supports online calibration update. The input of the coordinate transformation unit 5 is the deviation amount from the template matching unit 4, and the output is the compensation correction amount, which is sent to the closed-loop control unit 6. Meanwhile, the coordinate transformation unit 5 also receives the residual deviation feedback from the closed-loop control unit 6 for coordinate transformation in iterative correction.

[0106] The closed-loop control unit 6 is used to drive the actuator to perform the correction action according to the compensation correction amount, collect real-time position feedback data of the actuator, calculate the residual deviation, and feed back the residual deviation to the coordinate transformation unit 5 for iterative correction when the residual deviation is greater than a preset correction accuracy threshold. The hardware components of the closed-loop control unit 6 include a motion controller, a servo driver and an actuator. In the embodiment, the motion controller adopts a programmable logic controller or a special motion control card, which supports multi-axis linkage control; the actuator adopts a three-axis servo platform, including an X-axis linear module, a Y-axis linear module and a θ-axis rotary table. The closed-loop control unit 6 also establishes a data communication connection with the visual detection unit 2, which is used to trigger the visual secondary detection to calculate the residual deviation after the correction action is completed.

[0107] The data flow and control flow relationship between the units of the system of the present application are as follows: after the tool tray arrives at the station entrance, the model identification unit 1 first reads the RFID tag data and loads the standard template parameters; then the visual detection unit 2 collects the feature region images and sends them to the feature extraction unit 3; the feature extraction unit 3 extracts the feature parameters and sends them to the template matching unit 4; the template matching unit 4 calculates the deviation amount in combination with the standard template parameters and sends it to the coordinate transformation unit 5; the coordinate transformation unit 5 converts the deviation amount into the compensation correction amount and sends it to the closed-loop control unit 6; the closed-loop control unit 6 drives the actuator to complete the correction action and judges whether iterative correction is needed. If iterative correction is needed, the closed-loop control unit 6 triggers the visual detection unit 2 to re-collect images and feeds back the residual deviation to the coordinate transformation unit 5, forming a closed-loop control.

[0108] In a preferred embodiment of the present application, the system further includes a system controller and a human-machine interaction interface. The system controller is used to coordinate the working time sequence of the units, manage the multi-model template library, record the correction log and handle abnormal situations. The human-machine interaction interface is used to display the system running state, real-time correction data and historical statistical information, and supports parameter configuration and manual intervention operation by the operator.

[0109] The system can be deployed at the entrance of each work station of the shaft box production line which needs accurate positioning, and the deviation correction systems of multiple work stations can independently operate or be centrally managed through an upper computer or an industrial Ethernet. Through the system, the shaft box can complete automatic deviation correction before entering a cleaning work station or a grinding work station, ensuring that the machining references of the work stations are consistent, thereby improving overall machining quality and production efficiency.

[0110] The above-described embodiments only express specific implementation manners of the present application, which are described in detail, but should not be understood as limitations on the patent scope of the present application. It should be noted that, for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the protection scope of the present application.

Claims

1. The automatic deviation correction control method of the axle box body cleaning and polishing station, characterized in that, The method comprises the following steps: a model identification step, reading the RFID tag data on the tool tray through the RFID reader deployed at the entrance of the station, determining the workpiece model information of the current shaft box body according to the RFID tag data, and loading the standard template parameters corresponding to the workpiece model information from the multi-model template library; an image acquisition step, vertically downwardly shooting the feature area of the shaft box body on the tool through the visual detection unit deployed at the entrance of the cleaning station or the entrance of the polishing station, and acquiring the feature area image; a feature extraction step, performing edge detection processing on the feature area image to extract the reference surface contour line of the shaft box body, and performing Hough circle transformation processing on the feature area image to identify the positioning hole of the shaft box body and acquire the positioning hole center coordinates; wherein the edge detection processing on the feature area image comprises: performing Gaussian filter denoising processing on the feature area image, using Canny operator to perform edge detection on the denoised image to obtain an edge image, and performing straight line fitting on the edge pixel points in the edge image to obtain the reference surface contour line; a deviation calculation step, matching and calculating the reference surface contour line and the positioning hole center coordinates with the standard template parameters to determine the X-direction translation deviation, Y-direction translation deviation and Z-axis rotation deviation of the current shaft box body relative to the standard pose; wherein the matching and calculation of the reference surface contour line and the positioning hole center coordinates with the standard template parameters comprises: calculating the angle deviation between the reference surface contour line and the standard reference surface contour line as an initial rotation deviation estimate, performing rotation compensation on the current positioning hole center coordinates according to the initial rotation deviation estimate, calculating the position difference between the compensated positioning hole center coordinates and the standard positioning hole center coordinates as the translation deviation, and finely correcting the rotation deviation according to the matching residual of multiple feature points; a coordinate transformation step, converting the X-direction translation deviation, the Y-direction translation deviation and the Z-axis rotation deviation into compensation correction amounts of the servo platform or the robot end according to the transformation relationship between the coordinate system of the current station and the servo platform coordinate system or the robot base coordinate system; a closed-loop correction step, driving the servo platform or the robot end to execute the correction action according to the compensation correction amount, acquiring the real-time position feedback data of the execution mechanism, calculating the residual deviation between the current pose and the target pose, and feeding back the residual deviation to the coordinate transformation step for iterative correction when the residual deviation is greater than a preset correction accuracy threshold, until the residual deviation is less than or equal to the preset correction accuracy threshold; wherein an incremental correction strategy is adopted, and the execution amount of each correction action is the product of the current calculated compensation correction amount and a preset proportion coefficient, and the value range of the preset proportion coefficient is 0.6 to 0.

9.

2. The automatic deviation correction control method of the axle box body cleaning and polishing station according to claim 1, characterized in that, The preset correction accuracy threshold comprises a position accuracy threshold and an angle accuracy threshold, the value range of the position accuracy threshold is 0.1mm to 0.3mm, and the value range of the angle accuracy threshold is 0.05° to 0.15°.

3. The automatic deviation correction control method of the axle box body cleaning and polishing station according to claim 1, characterized in that, The visual detection unit comprises an industrial camera with a resolution of no less than 5 million pixels and a ring-shaped light source with a color temperature ranging from 5500K to 6500K.

4. The automatic deviation correction control method of the axle box body cleaning and polishing station according to claim 1, characterized in that, The standard template parameters comprise a standard coordinate sequence of a reference surface contour line, a standard center coordinate of a positioning hole, a standard radius of the positioning hole, and a feature point relative position relationship matrix.

5. The automatic deviation correction control method of the axle box body cleaning and polishing station according to claim 1, characterized in that, In the coordinate transformation step, the coordinate transformation according to the transformation relationship between the coordinate system of the current station and the servo platform coordinate system comprises: obtaining a previously calibrated homogeneous transformation matrix of the visual coordinate system to the servo platform coordinate system, and multiplying the deviation vector in the visual coordinate system by the left side of the homogeneous transformation matrix to obtain a compensation correction amount in the servo platform coordinate system.

6. The automatic deviation correction control method of the axle box body cleaning and polishing station according to claim 1, characterized in that, The closed-loop correction step further comprises: when the number of correction iterations exceeds the preset maximum number of iterations and the residual deviation is still greater than the preset correction accuracy threshold, generating a correction abnormality alarm signal and pausing the processing flow of the current station.

7. The automatic deviation correction control system of the axle box body cleaning and polishing station, used to realize the automatic deviation correction control method of the axle box body cleaning and polishing station according to any one of claims 1-6, characterized in that, Comprise: A model identification unit is configured to read RFID tag data on a tool tray through an RFID reader deployed at an entrance of a station, determine workpiece model information of a current shaft box according to the RFID tag data, and load standard template parameters corresponding to the workpiece model information from a multi-model template library; A visual detection unit is configured to vertically downwardly shoot a feature region of a shaft box on a tool at an entrance of a cleaning station or a grinding station to obtain a feature region image; A feature extraction unit is configured to perform edge detection processing on the feature region image to extract a reference surface contour line of the shaft box, and perform Hough circle transformation processing on the feature region image to identify a positioning hole of the shaft box and obtain a positioning hole center coordinate; A template matching unit is configured to match and calculate the reference surface contour line and the positioning hole center coordinate with the standard template parameters to determine X-direction translation deviation, Y-direction translation deviation, and rotation deviation around a Z axis of a current shaft box relative to a standard pose; A coordinate transformation unit is configured to convert the translation deviation and the rotation deviation into a compensation correction amount according to a transformation relationship between a coordinate system of a current station and a servo platform coordinate system or a robot base coordinate system; A closed-loop control unit is configured to drive an actuator to perform a correction action according to the compensation correction amount, collect real-time position feedback data of the actuator, calculate a residual deviation, and feed back the residual deviation to the coordinate transformation unit for iterative correction when the residual deviation is greater than a preset correction accuracy threshold.

Images