Intelligent calibration method and system of full-automatic placement machine

Abstract

The invention discloses an intelligent calibration system of a full-automatic placement machine. The intelligent calibration system is used for improving the lamination accuracy of the full-automatic placement machine. The method comprises the following steps: collecting real-time image data of a to-be-tested placement machine during operation, wherein the real-time image data comprises real-time position information of a placement head in the to-be-tested placement machine and real-time position information of a target point location; calculating an initial deviation value of the mounting head and the target point location according to the real-time image data; generating a correction parameter based on the initial deviation value and historical verification data; correcting the current motion path of the mounting head by utilizing a Bezier curve algorithm and combining the correction parameters to obtain an adjusted motion path, and obtaining a plurality of target calibration points from the adjusted motion path; monitoring the path deviation of the mounting head in the adjustment motion path by using the plurality of target calibration points; and updating the adjusted motion path according to the path deviation to obtain a target motion path.

Description

Technical Field

[0001] This application relates to the technical field of path planning for pick-and-place machines, and particularly to an intelligent calibration system for a full-automatic pick-and-place machine. Background Art

[0002] A full-automatic pick-and-place machine is a device used in the electronics manufacturing industry, mainly for accurately placing surface-mounted components onto printed circuit boards. In modern electronics manufacturing, the research and application of full-automatic pick-and-place machines play a crucial role, directly determining the production efficiency and product quality levels. With the advancement of intelligent manufacturing, pick-and-place technology is not only a key link in achieving high-precision assembly but also a core driving force for promoting industrial upgrading.

[0003] During the actual operation of a full-automatic pick-and-place machine, due to various reasons such as mechanical wear, environmental factors, and electrical interference, the actual movement path of the pick-and-place head often deviates from the ideal path, resulting in inaccurate placement positions, thus affecting product quality and production efficiency. Traditional calibration methods usually rely on manual experience and regular mechanical adjustments. This method is not only inefficient but also difficult to adapt to complex and changing production environments and cannot calibrate the movement path of the pick-and-place head in real time and accurately. Summary of the Invention

[0004] This application provides an intelligent calibration system for a full-automatic pick-and-place machine to improve the fitting accuracy of the full-automatic pick-and-place machine.

[0005] In a first aspect of this application, an intelligent calibration method for a full-automatic pick-and-place machine is provided, including: Collecting real-time image data during the operation of the pick-and-place machine to be measured, where the real-time image data includes the real-time position information of the pick-and-place head and the real-time position information of the target point in the pick-and-place machine to be measured; Calculating an initial deviation value between the pick-and-place head and the target point based on the real-time image data; Generating a correction parameter based on the initial deviation value and historical calibration data; Using the Bezier curve algorithm and combining the correction parameter to correct the current movement path of the pick-and-place head to obtain an adjusted movement path, and obtaining several target calibration points from the adjusted movement path; Monitoring the path deviation of the pick-and-place head in the adjusted movement path using the several target calibration points; Updating the adjusted movement path according to the path deviation to obtain a target movement path.

[0006] Optionally, the generating a correction parameter based on the initial deviation value and historical calibration data includes: Initializing the Kalman filter algorithm using historical calibration data to obtain an initial prediction model; Use the initial deviation value as an observation value to update the initial prediction model, and obtain a target prediction model; Predict the motion trend and deviation change range of the placement head according to the current state estimation value output by the target prediction model; Calculate a correction parameter according to the motion trend and the deviation change range.

[0007] Optionally, correcting the current motion path of the placement head by using the Bezier curve algorithm and in combination with the correction parameter to obtain an adjusted motion path, includes: Obtain the current motion path of the placement head, where the current motion path is a preset motion path; Generate an initial Bezier curve according to the current motion path; Adjust the control points of the initial Bezier curve based on the correction parameter to obtain a corrected Bezier curve; Determine the adjusted motion path according to the corrected Bezier curve.

[0008] Optionally, obtaining a plurality of target calibration points from the adjusted motion path, includes: Taking the starting point of the adjusted motion path as a starting point, and sequentially calculating the distances between adjacent path points on the adjusted motion path; When the cumulative distance reaches a preset spacing, determine the current path point as a candidate calibration point; Judge whether the candidate calibration point is at a preset position; If so, determine the candidate calibration point as a target calibration point.

[0009] Optionally, calculating the initial deviation value between the placement head and the target point according to the real-time image data, includes: Use an edge detection algorithm to extract feature points from the real-time image data to obtain the feature points corresponding to the placement head and the target point; Calculate the coordinate difference between the feature points corresponding to the placement head and the target point in the same coordinate system to obtain an initial deviation value.

[0010] Optionally, after monitoring the path deviation of the placement head in the adjusted motion path by using the plurality of target calibration points, the method further includes: Judge whether the path deviation is greater than a preset threshold; If so, determine that the placement machine to be tested is in a fault state and issue an alarm prompt; If not, update the adjusted motion path according to the path deviation to obtain a target motion path.

[0011] The second aspect of the present application provides an intelligent calibration system for a fully automatic mounter, including: An acquisition unit, configured to acquire real-time image data of the mounter to be measured during operation, where the real-time image data includes real-time position information of a placement head in the mounter to be measured and real-time position information of a target point; A calculation unit, configured to calculate an initial deviation value between the placement head and the target point based on the real-time image data; A generation unit, configured to generate a correction parameter based on the initial deviation value and historical calibration data; A correction unit, configured to correct the current movement path of the placement head by using the Bezier curve algorithm in combination with the correction parameter to obtain an adjusted movement path, and obtain a plurality of target calibration points from the adjusted movement path; A monitoring unit, configured to monitor the path deviation of the placement head in the adjusted movement path by using the plurality of target calibration points; An update unit, configured to update the adjusted movement path according to the path deviation to obtain a target movement path.

[0012] Optionally, the generation unit is specifically configured to: Initialize the Kalman filter algorithm by using historical calibration data to obtain an initial prediction model; Update the initial prediction model by using the initial deviation value as an observation value to obtain a target prediction model; Predict the movement trend and deviation change range of the placement head according to the current state estimation value output by the target prediction model; Calculate a correction parameter according to the movement trend and the deviation change range.

[0013] Optionally, the correction unit is specifically configured to: Obtain the current movement path of the placement head, where the current movement path is a preset movement path; Generate an initial Bezier curve according to the current movement path; Adjust control points of the initial Bezier curve based on the correction parameter to obtain a corrected Bezier curve; Determine an adjusted movement path according to the corrected Bezier curve.

[0014] Optionally, the correction unit is further specifically configured to: Take the initial point of the adjusted movement path as a starting point, and sequentially calculate distances between adjacent path points on the adjusted movement path; When the cumulative distance reaches a preset spacing, determine the current path point as a candidate calibration point; Judge whether the candidate calibration point is at a preset position; If so, determine the candidate calibration point as the target calibration point.

[0015] As can be seen from the above technical solutions, the present application has the following effects: First, collect the real-time image data of the to-be-tested mounter during operation. The real-time image data includes the real-time position information of the placement head and the real-time position information of the target point in the to-be-tested mounter. Then, calculate the initial deviation value between the placement head and the target point according to the real-time image data. Next, generate correction parameters based on the initial deviation value and historical calibration data. Further, use the Bezier curve algorithm and combine the correction parameters to correct the current movement path of the placement head to obtain an adjusted movement path, and obtain several target calibration points from the adjusted movement path. Then, further use the several target calibration points to monitor the path deviation of the placement head in the adjusted movement path. Finally, update the adjusted movement path according to the path deviation to obtain the target movement path. In this way, the initial deviation value between the placement head and the target point to be placed can be obtained by image analysis, and correction parameters are generated based on the initial deviation value and combined with historical calibration data to preliminarily correct the current movement path of the placement head. And determine the target calibration points for subsequent monitoring from the adjusted movement path obtained after correction to realize the real-time update of the adjusted movement path. Instead of manually adjusting the movement path, the path adjustment efficiency of the placement head can be improved, and then the fitting accuracy of the fully automatic mounter can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of an embodiment of an intelligent calibration method for a fully automatic mounter in the present application; Figure 2-1 、 Figure 2-2 and Figure 2-3 It is a schematic diagram of another embodiment of an intelligent calibration method for a fully automatic mounter in the present application; Figure 3 It is a schematic diagram of an embodiment of an intelligent calibration system for a fully automatic mounter in the present application; Figure 4 It is a schematic diagram of another embodiment of an intelligent calibration system for a fully automatic mounter in the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The present application provides an intelligent calibration system for a fully automatic mounter to improve the fitting accuracy of the fully automatic mounter.

[0018] The intelligent calibration method for the fully automatic mounter described in the present application is implemented by being executed on a system, a server, or other terminals with logical analysis and processing capabilities. The specific execution entity is not limited in this embodiment.

[0019] Please refer to Figure 1As shown in the figure, an embodiment of the intelligent calibration method for a full-automatic mounter in this application includes: 101. Collect real-time image data of the mounter to be measured during operation. This real-time image data includes the real-time position information of the placement head in the mounter to be measured and the real-time position information of the target point. In this embodiment, the target point is the area on the printed circuit board where components are to be placed. In the working area of the mounter to be measured, multiple industrial cameras are installed according to its layout and vision requirements. For example, for a large full-automatic mounter, industrial cameras can be installed respectively above the center, on the side, and around the target point of the full-automatic placement head to ensure capturing images from different angles and avoiding visual blind spots. When the mounter to be measured starts to operate, trigger the industrial cameras to perform image acquisition, so that the industrial cameras continuously capture images of the placement head and the target point at a set frame rate and transmit the images to the system in real time to obtain the real-time image data of the mounter to be measured during operation.

[0020] 102. Calculate the initial deviation value between the placement head and the target point according to the real-time image data. In this embodiment, after receiving the real-time image data, the system can first perform preprocessing on it. For example, use Gaussian filtering to remove noise in the real-time image data to enhance the clarity of the real-time image data; adopt histogram equalization method to adjust the contrast of the real-time image data to make the features in the real-time image data more obvious. Then, use edge detection algorithm or object detection algorithm based on deep learning to extract the position coordinates of the placement head and the target point from the preprocessed real-time image data. Among them, for the placement head, its center coordinates can be extracted; for the target point, the coordinates of its center point or specific identification point can be extracted. Then establish a unified rectangular coordinate system in the working space of the mounter to be measured, and convert the position coordinates of the placement head and the target point to this rectangular coordinate system. The origin of this rectangular coordinate system can be selected as a certain fixed reference point of the mounter, such as a corner of the machine base. Finally, calculate the coordinate difference between the placement head and the target point in this rectangular coordinate system to obtain the initial deviation value.

[0021] 103. Generate correction parameters based on the initial deviation value and historical calibration data. In this embodiment, obtain a large amount of historical calibration data accumulated during the past operation of the mounter to be measured. This part of historical calibration data includes the motion parameters, deviation values of the placement head, and corresponding calibration measures and results under different production conditions such as different circuit board types, electronic component specifications, environmental temperature and humidity. The obtained historical calibration data can be cleaned and preprocessed to remove outliers and incorrect data to improve the accuracy and reliability of the historical calibration data. Among them, the Kalman filtering algorithm can be used to calculate the correction parameters by combining the initial deviation value and the historical calibration data, which will be described in specific subsequent embodiments.

[0022] 104. Use the Bezier curve algorithm and combine it with correction parameters to correct the current movement path of the placement head, obtain an adjusted movement path, and obtain several target calibration points from the adjusted movement path. In this embodiment, the Bezier curve algorithm is a mathematical method for defining the shape of a curve through control points. For the correction of the movement path of the placement head, the current movement path of the placement head can be regarded as an initial Bezier curve, and the correction parameter is used as the offset of the control point. By adjusting the position of the control point, the shape of the Bezier curve is changed, thereby obtaining an adjusted movement path. Then, according to the preset calibration point selection principle and constraint conditions, calibration points are determined from the adjusted movement path. For example, consider the overall accuracy index of the placement machine to be measured and the requirements for the component placement position accuracy in a specific placement task; or consider the movement characteristics of the placement head such as speed, acceleration, and path curvature; or comprehensively consider the characteristics of the hardware equipment.

[0023] 105. Use several target calibration points to monitor the path deviation of the placement head in the adjusted movement path. In this embodiment, sensor devices such as laser displacement sensors and encoders are arranged near the placement head and the target calibration points. The laser displacement sensor is used to measure the distance and position between the placement head and the target calibration point in real time, and the encoder is used to measure the movement speed and angle of the placement head. The sensors collect data according to the set sampling frequency and transmit the data to the system. When the placement head moves to the target calibration point, the data collected by the sensors is read to obtain the actual position information of the placement head. The difference between the actual position information and the theoretical position information of the target calibration point is calculated, and the path deviation is calculated based on the position differences between each target calibration point and the placement head. In another implementable manner, real-time image acquisition of the placement head and the target calibration points can also be performed through an industrial camera, and the collected images are analyzed to determine the path deviation of the placement head in the adjusted movement path.

[0024] 106. Update the adjusted movement path according to the path deviation to obtain the target movement path.

[0025] After obtaining the path deviation, it is judged whether the path deviation is within the allowable deviation range. If not, adjustment amounts such as displacement and speed are calculated based on the path deviation, and these adjustment amounts are applied to the adjusted movement path to update the adjusted movement path, and the updated adjusted movement path is determined as the target movement path. When the placement head moves according to the updated target movement path, the path deviation is continuously monitored to verify the effect of the path update.

[0026] In this embodiment, in this way, the initial deviation value between the placement head and the target point to be placed can be obtained by image analysis. Based on this initial deviation value and combined with historical calibration data, correction parameters are generated to preliminarily correct the current movement path of the placement head. And the target calibration point for subsequent monitoring is determined from the adjusted movement path obtained after correction to achieve real-time update of the adjusted movement path. And there is no need to manually adjust the movement path, so the path adjustment efficiency of the placement head can be improved, and thus the fitting accuracy of the fully automatic mounter can be improved.

[0027] Please refer to Figure 2-1 、 Figure 2-2 、and Figure 2-3 As shown, another embodiment of the intelligent calibration method of the fully automatic mounter in this application includes: 201. Collect the real-time image data of the mounter to be measured during operation. The real-time image data includes the real-time position information of the placement head and the real-time position information of the target point in the mounter to be measured; Step 201 in this embodiment is similar to step 101 in the embodiment shown above Figure 1 and will not be elaborated here.

[0028] 202. Use the edge detection algorithm to extract feature points from the real-time image data to obtain the feature points corresponding to the placement head and the target point; 203. Calculate the coordinate difference between the feature points corresponding to the placement head and the target point in the same coordinate system to obtain the initial deviation value; Optionally, in this embodiment, the edge detection algorithm is an image processing technology that can detect the position of the object edge in the image. Common edge detection algorithms include the Sobel operator, Prewitt operator, and Canny operator, etc. Among them, since the Canny operator has better edge localization accuracy and anti-noise ability, and it can accurately detect the real edge in the image through multi-stage processing such as Gaussian smoothing, gradient calculation, non-maximum suppression, and double-threshold processing, this embodiment can use the Canny operator as the edge detection algorithm. In the process of extracting feature points using the edge detection algorithm, first perform Gaussian smoothing processing on the real-time image data to remove the noise in the real-time image data and improve the accuracy of edge detection; then calculate the gradient amplitude and direction of the real-time image data to determine the change rate and direction information of the pixels in the real-time image data; then perform non-maximum suppression on the gradient amplitude, only retaining the local maximum value to refine the edge; then determine the final edge image through double-threshold processing and connecting edges; finally, extract the feature points of the placement head and the target point from the edge image, such as corner points, end points, etc.

[0029] After extracting the feature points of the placement head and the target points, match the feature points of the placement head with those of the target points to find their corresponding relationships. Descriptor-based matching methods such as Scale-Invariant Feature Transform (SIFT), Speeded-Up Robust Features (SURF), etc. can be used to achieve accurate matching of feature points under different lighting, scale, and rotation conditions. Establish a unified rectangular coordinate system in the working space of the placement machine to be measured. According to the matched feature points, determine the spatial coordinate values of the placement head and the target points in this rectangular coordinate system, and calculate the difference between the spatial coordinate value of the placement head and that of the target points. This difference includes the coordinate differences of the placement head and the target points on the X-axis, Y-axis, and θ-axis to obtain the initial deviation values of the placement head relative to the target points in the horizontal, vertical, and rotational directions.

[0030] 204. Initialize the Kalman filter algorithm using historical calibration data to obtain an initial prediction model; 205. Update the initial prediction model using the initial deviation value as the observation value to obtain a target prediction model; 206. Predict the movement trend and deviation change range of the placement head according to the current state estimation value output by the target prediction model; 207. Calculate the correction parameters according to the movement trend and deviation change range; Optionally, in this embodiment, first, according to the historical calibration data and the physical model of the placement head, parameters required for the Kalman filter algorithm such as the system state transition matrix, the observation matrix, the process noise covariance matrix, and the observation noise covariance matrix are determined. Then, the state estimate value and the error covariance matrix of the Kalman filter algorithm are initialized to obtain an initial prediction model. Among them, the initial state estimate value can be set as the actual position or the target position of the placement head at the previous moment, and the error covariance matrix can be set according to experience or the statistical characteristics of historical data. Next, the initial deviation value is input into the initial prediction model as an observation value. The predicted state is transformed to the same coordinate system as the observation value through the observation matrix, and the difference between the predicted value and the observation value is calculated. The Kalman gain is calculated based on the observation noise covariance matrix and the predicted error covariance matrix, and the predicted value is corrected using the Kalman gain to obtain the current state estimate value. At the same time, the error covariance matrix is updated to obtain the target prediction model. Based on this current state estimate value, the movement trend of the placement head in the next period of time is predicted, such as predicting the position change, speed change, and angle change of the placement head in the next few control cycles. At the same time, the error covariance matrix is used to estimate the deviation change range. It should be noted that the diagonal elements of the error covariance matrix reflect the variances of the estimated values of each state variable. By taking the square root of the variance and combining a certain confidence level, the change range of the deviation value in each direction can be obtained. Finally, according to the predicted movement trend and deviation change range, and in combination with the accuracy requirements of the placement process and the control characteristics of the equipment, the correction parameters for correcting the initial deviation value of the placement head are calculated.

[0031] 208. Obtain the current movement path of the placement head, where the current movement path is a preset movement path; 209. Generate an initial Bézier curve according to the current movement path; 210. Adjust the control points of the initial Bézier curve based on the correction parameters to obtain a corrected Bézier curve; 211. Determine the adjusted movement path according to the corrected Bézier curve; Optionally, in this embodiment, before the placement machine to be measured starts working, the current movement path of the placement head to be measured is preset. The preset current movement path is determined according to the requirements of the placement task, such as the positions of the components to be placed, the placement sequence, and other factors. During the generation of the initial Bézier curve, key points need to be extracted from the current movement path first. These key points usually include the starting point, the ending point of the current movement path, and the turning points on the path. For example, when the placement head moves from a component feeder to the placement position on the circuit board, if there is a turn in the path, the point at the turn is a key point. These key points determine the general shape and trend of the Bézier curve. After determining the key points, based on the geometric relationship of the key points, for example, for two adjacent key points, a point with a certain ratio on the line connecting them can be taken as a control point, or the position of the control point can be adjusted according to factors such as the curvature of the path. Finally, the initial Bézier curve is generated based on this part of the control points.

[0032] After generating the initial Bézier curve, the original coordinates of each control point are operated with the corresponding correction parameters to adjust the control points, and a corrected Bézier curve is generated based on the adjusted control points. The corrected Bézier curve is a continuous curve, while the placement head requires discrete motion instructions to achieve motion control. Therefore, the corrected Bézier curve needs to be discretized. Among them, a series of sampling points can be selected on the corrected Bézier curve, and the spacing of these sampling points can be determined according to the requirements of placement accuracy and motion control. For each sampling point, corresponding motion instructions are generated according to information such as its position and the tangent direction of the curve. The motion instructions include parameters such as the moving speed, acceleration, and moving direction of the placement head. For example, if the distance between sampling points is large, a higher moving speed can be set; if the tangent direction of the curve changes greatly, the acceleration can be adjusted to ensure the smooth movement of the placement head. Combining the motion instructions corresponding to all sampling points in sequence, the corrected adjusted motion path can be obtained.

[0033] 212. Taking the starting point of the adjusted motion path as the starting point, calculate the distances between adjacent path points on the adjusted motion path in sequence; 213. When the cumulative distance reaches the preset spacing, determine the current path point as the candidate calibration point; 214. Determine whether the candidate calibration point is at the preset position. If so, execute step 215; 215. Determine the candidate calibration point as the target calibration point; Optionally, in this embodiment, first set a fixed preset spacing. Starting from the initial point of the adjusted motion path as the starting point, measure the distances between adjacent points along the adjusted motion path in sequence and accumulate them one by one. When the cumulative distance reaches the preset spacing for any path point, set the current path point as a candidate calibration point. Then, starting from this candidate calibration point, continue to accumulate distances to find the next candidate calibration point. Then, identify the preset positions during the mounting process, such as the pick-up point, the placement point, the turning points in the path, and the positions where the speed or acceleration changes suddenly, etc. When the candidate calibration point is at a preset position, the candidate calibration point can be determined as the target calibration point.

[0034] 216. Use several target calibration points to monitor the path deviation of the placement head in the adjusted motion path; Step 216 in this embodiment is similar to Figure 1 Step 105 in the embodiment shown above, and details will not be repeated here.

[0035] 217. Determine whether the path deviation is greater than the preset threshold. If so, execute step 218; if not, execute step 219; 218. Determine that the placement machine to be tested is in a faulty state and issue an alarm prompt; 219. Update the adjusted motion path according to the path deviation to obtain the target motion path.

[0036] Optionally, in this embodiment, the preset threshold can be set according to the actual placement accuracy requirements and the equipment characteristics of the placement machine to be tested. When the path deviation is greater than the preset threshold, it means that the path deviation of the placement head exceeds the adjustable range, and the placement machine to be tested may have a fault problem at this time. Therefore, an alarm prompt can be issued to remind the staff to handle the fault in time. When the path deviation is less than or equal to the preset threshold, it means that the path deviation of the placement head is within the adjustable range. At this time, the adjusted motion path can be updated according to the path deviation to obtain a target motion path with higher accuracy.

[0037] Please refer to Figure 3 As shown, an embodiment of the intelligent calibration system of the full-automatic placement machine in this application includes: An acquisition unit 301, configured to acquire real-time image data during the operation of the placement machine to be tested. The real-time image data includes the real-time position information of the placement head and the real-time position information of the target point in the placement machine to be tested; A calculation unit 302, configured to calculate the initial deviation value between the placement head and the target point according to the real-time image data; A generation unit 303, configured to generate a correction parameter based on the initial deviation value and historical calibration data; A correction unit 304 is configured to correct the current movement path of the placement head by using the Bezier curve algorithm and in combination with correction parameters to obtain an adjusted movement path, and obtain a plurality of target calibration points from the adjusted movement path; A monitoring unit 305 is configured to monitor the path deviation of the placement head in the adjusted movement path by using a plurality of target calibration points; An updating unit 306 is configured to update the adjusted movement path according to the path deviation to obtain a target movement path.

[0038] In this embodiment, the acquisition unit 301 acquires real-time image data during the operation of the to-be-tested mounter. The real-time image data includes the real-time position information of the placement head and the real-time position information of the target point in the to-be-tested mounter. The calculation unit 302 calculates the initial deviation value between the placement head and the target point according to the real-time image data. The generation unit 303 generates correction parameters based on the initial deviation value and historical calibration data. The correction unit 304 corrects the current movement path of the placement head by using the Bezier curve algorithm and in combination with the correction parameters to obtain an adjusted movement path, and obtains a plurality of target calibration points from the adjusted movement path. The monitoring unit 305 monitors the path deviation of the placement head in the adjusted movement path by using a plurality of target calibration points. The updating unit 306 updates the adjusted movement path according to the path deviation to obtain a target movement path. In this way, the initial deviation value between the placement head and the target point to be placed can be obtained by using image analysis. Based on the initial deviation value and in combination with historical calibration data, correction parameters are generated to preliminarily correct the current movement path of the placement head. And the target calibration points for subsequent monitoring are determined from the adjusted movement path obtained after correction to realize the real-time update of the adjusted movement path. Instead of manually adjusting the movement path, the path adjustment efficiency of the placement head can be improved, and further the fitting accuracy of the full-automatic mounter can be improved.

[0039] Please refer to Figure 4 As shown in the figure, another embodiment of the intelligent calibration system of the full-automatic mounter in the present application includes: An acquisition unit 401 is configured to acquire real-time image data during the operation of the to-be-tested mounter. The real-time image data includes the real-time position information of the placement head and the real-time position information of the target point in the to-be-tested mounter; A calculation unit 402 is specifically configured to extract feature points from the real-time image data by using an edge detection algorithm to obtain the feature points corresponding to the placement head and the target point; calculate the coordinate difference between the feature points corresponding to the placement head and the target point in the same coordinate system to obtain an initial deviation value; A generating unit 403, specifically configured to initialize a Kalman filtering algorithm by using historical calibration data to obtain an initial prediction model; update the initial prediction model by using the initial deviation value as an observation value to obtain a target prediction model; predict the movement trend and deviation change range of the placement head according to the current state estimation value output by the target prediction model; calculate a correction parameter according to the movement trend and deviation change range; A correcting unit 404, specifically configured to obtain the current movement path of the placement head, where the current movement path is a preset movement path; generate an initial Bézier curve according to the current movement path; adjust control points of the initial Bézier curve based on the correction parameter to obtain a corrected Bézier curve; determine an adjusted movement path according to the corrected Bézier curve; take the initial point of the adjusted movement path as a starting point, and sequentially calculate distances between adjacent path points on the adjusted movement path; when the cumulative distance reaches a preset spacing, determine the current path point as a candidate calibration point; determine whether the candidate calibration point is at a preset position; if so, determine the candidate calibration point as a target calibration point; A monitoring unit 405, configured to monitor the path deviation of the placement head in the adjusted movement path by using a plurality of target calibration points; A judging unit 406, configured to judge whether the path deviation is greater than a preset threshold; An alarming unit 407, configured to determine that the to-be-tested mounter is in a fault state and issue an alarm prompt when the path deviation is greater than the preset threshold; An updating unit 408, specifically configured to update the adjusted movement path according to the path deviation to obtain a target movement path when the path deviation is less than or equal to the preset threshold.

[0040] In this embodiment, the functions of each unit are similar to those of steps 201 to 219 in the foregoing Figure 2-1 、 Figure 2-2 and Figure 2-3 illustrated embodiments, and details are not described herein again.

[0041] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described system, device, and unit can refer to the corresponding processes in the foregoing method embodiments, and details are not described herein again.

[0042] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces, and the indirect coupling or communication connection of the device or unit can be in an electrical, mechanical, or other form.

[0043] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0044] In addition, in each embodiment of the present application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0045] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

Claims

1. An intelligent calibration method for a fully automatic mounter, characterized in that, Comprising: Collecting real-time image data of the to-be-tested mounter during operation, where the real-time image data includes the real-time position information of the placement head and the real-time position information of the target point in the to-be-tested mounter; Calculating an initial deviation value between the placement head and the target point according to the real-time image data; Generating a correction parameter based on the initial deviation value and historical calibration data; Using the Bezier curve algorithm and combining with the correction parameter to correct the current movement path of the placement head, obtaining an adjusted movement path, and obtaining a plurality of target calibration points from the adjusted movement path; Monitoring the path deviation of the placement head in the adjusted movement path by using the plurality of target calibration points; Updating the adjusted movement path according to the path deviation to obtain a target movement path.

2. The intelligent calibration method of the full-automatic mounter according to claim 1, characterized in that, The generating a correction parameter based on the initial deviation value and historical calibration data includes: Initializing the Kalman filter algorithm by using historical calibration data to obtain an initial prediction model; Updating the initial prediction model by using the initial deviation value as an observation value to obtain a target prediction model; Predicting the movement trend and deviation change range of the placement head according to the current state estimation value output by the target prediction model; Calculating a correction parameter according to the movement trend and the deviation change range.

3. The intelligent calibration method of the full-automatic mounter according to claim 1, characterized in that The using the Bezier curve algorithm and combining with the correction parameter to correct the current movement path of the placement head, obtaining an adjusted movement path, includes: Obtaining the current movement path of the placement head, where the current movement path is a preset movement path; Generating an initial Bezier curve according to the current movement path; Adjusting the control points of the initial Bezier curve based on the correction parameter to obtain a corrected Bezier curve; Determining the adjusted movement path according to the corrected Bezier curve.

4. The intelligent calibration method of the full-automatic component mounter according to claim 1, wherein The obtaining a plurality of target calibration points from the adjusted movement path includes: Taking the initial point of the adjusted movement path as a starting point, and successively calculating the distances between adjacent path points on the adjusted movement path; When the cumulative distance reaches a preset spacing, determining the current path point as a candidate calibration point; Judging whether the candidate calibration point is at a preset position; If so, determining the candidate calibration point as a target calibration point.

5. The intelligent calibration method of the full-automatic component mounter according to claim 1, wherein The calculating an initial deviation value between the placement head and the target point according to the real-time image data includes: Extracting feature points from the real-time image data by using an edge detection algorithm to obtain the feature points corresponding to the placement head and the target point; Calculating the coordinate difference of the feature points corresponding to the placement head and the target point in the same coordinate system to obtain an initial deviation value.

6. The intelligent calibration method of the full-automatic mounter according to any one of claims 1 to 5, characterized in that, After the monitoring the path deviation of the placement head in the adjusted movement path by using the plurality of target calibration points, the method further includes: Judging whether the path deviation is greater than a preset threshold; If so, determining that the to-be-tested mounter is in a fault state and sending an alarm prompt; If not, updating the adjusted movement path according to the path deviation to obtain a target movement path.

7. An intelligent calibration system for a fully automatic mounter, characterized in that, Comprising: The acquisition unit is used to acquire real-time image data of the to-be-tested mounter during operation, and the real-time image data includes the real-time position information of the placement head and the real-time position information of the target point in the to-be-tested mounter; The calculation unit is used to calculate the initial deviation value between the placement head and the target point according to the real-time image data; The generation unit is used to generate correction parameters based on the initial deviation value and historical calibration data; The correction unit is used to correct the current movement path of the placement head by using the B-spline curve algorithm in combination with the correction parameters to obtain an adjusted movement path, and obtain a number of target calibration points from the adjusted movement path; The monitoring unit is used to monitor the path deviation of the placement head in the adjusted movement path by using the number of target calibration points; The update unit is used to update the adjusted movement path according to the path deviation to obtain a target movement path.

8. The intelligent calibration system of the full-automatic mounter according to claim 7, wherein, Specifically, the generation unit is used for: Initializing the Kalman filter algorithm by using historical calibration data to obtain an initial prediction model; Updating the initial prediction model by using the initial deviation value as an observation value to obtain a target prediction model; Predicting the movement trend and deviation change range of the placement head according to the current state estimation value output by the target prediction model; Calculating correction parameters according to the movement trend and the deviation change range.

9. The intelligent calibration system of the full-automatic mounter according to claim 7, characterized in that, Specifically, the correction unit is used for: Obtaining the current movement path of the placement head, and the current movement path is a preset movement path; Generating an initial B-spline curve according to the current movement path; Adjusting the control points of the initial B-spline curve based on the correction parameters to obtain a corrected B-spline curve; Determining the adjusted movement path according to the corrected B-spline curve.

10. The intelligent calibration system of the full-automatic mounter according to claim 7, characterized in that, Specifically, the correction unit is further used for: Taking the initial point of the adjusted movement path as a starting point, and sequentially calculating the distances between adjacent path points on the adjusted movement path; When the cumulative distance reaches a preset spacing, determining the current path point as a candidate calibration point; Judging whether the candidate calibration point is at a preset position; If so, determining the candidate calibration point as a target calibration point.

Images