Visual guidance automobile door cover assembling and adjusting system and method

Abstract

The invention relates to a visual guidance automobile door cover assembling and adjusting system and method. The system comprises a feeding unit, a hinge adjusting unit, a vehicle door assembling and adjusting unit, a quality monitoring unit and a controller, when a to-be-assembled door cover needs to be assembled and adjusted, the controller controls the feeding unit to convey the to-be-assembled door cover to a centering table and controls the hinge adjusting unit to obtain surface difference measurement data of a hinge on the centering table, the compensation amount is calculated, and the door cover is assembled and adjusted. Adjusting the hinge according to the compensation amount so as to enable the error of the hinge relative to the door cover to be within a first preset range, and then controlling the door cover assembling and adjusting unit to perform optimal fitting matching on the door cover subjected to hinge adjustment and the vehicle body so as to enable the clearance surface difference measurement value of each measurement point between the door cover and the vehicle body to be within a second preset range; and finally, the quality monitoring unit is controlled to suck the door cover to be flat, so that the surface difference of the measuring points at the door handle of the door cover is within a third preset range, in this way, automatic and intelligent assembling and adjusting of the automobile door cover are achieved, and the precision and flexibility of the automobile door cover in the assembling process are improved.

Description

Technical Field

[0001] This application relates to the technical field of automobile manufacturing, and particularly to a vision-guided automobile door and hood assembly and adjustment system and method. Background Art

[0002] With the rapid development of automobile manufacturing technology, the modern automobile manufacturing industry is facing the dual pressures of increasingly fierce market competition and diverse user demands, which makes the demand for host factories to adopt automatic assembly and adjustment systems on the assembly line more urgent.

[0003] Traditional manual assembly has serious quality fluctuation problems, which not only pose challenges to the product quality stability in automobile manufacturing, but also may lead to damaged market reputation and high after-sales costs. For example, during the assembly process, the gap and surface difference dimensions between the door and hood and the body may be inconsistent due to different manual technical levels, which not only affects the overall aesthetics of the body appearance, but also may affect the sealing and safety performance of the door and hood.

[0004] To solve the problems existing in the above-mentioned manual assembly, an automatic assembly and adjustment system was introduced to reduce labor costs, reduce human errors and rework, and improve production efficiency and the reliability of the assembly line. In the existing technology, the door and hood assembly in automobile manufacturing generally uses the 321 coordinate system establishment method for door and hood adjustment. By using a 3D vision sensor to take pictures of the door and hood and the body, identifying and extracting features such as corner points, edges, and surfaces, establishing coordinate systems for the body and the door and hood respectively, and performing matching, the robot system is guided to achieve door and hood assembly. The principle of the 321 coordinate system establishment method: first, the Z-axis coordinates of at least three points are used to fit the surface to limit the translation of the object in the Z direction and the rotation along the X and Y directions, then the Y-axis coordinates of at least two points are used to fit a straight line to limit the translation of the object in the Y direction and the rotation along the Z direction, and finally, the X-axis coordinate of at least one point is used to finally determine the translation of the object in the X direction, so as to finally determine the positioning method of the position of an object. However, this regular matching and adjustment method based on the 321 coordinate system establishment is difficult to adapt to the incoming material consistency error of the body and the door and hood, resulting in poor adjustment effects.

[0005] On the other hand, there are also installation errors in the door and hood hinges. In this case, even if the door and hood are adjusted well, after the hinge bolts are tightened, the position of the door and hood will change again, and it is difficult to achieve a good matching state. In the current processing methods of the existing technology, only the adjustment of the automobile door and hood is realized, which will lead to large errors in the door and hood during the assembly process. Summary of the Invention

[0006] Based on this, in view of the above technical problems, it is necessary to provide a vision-guided automobile door and hood assembly and adjustment system. In some embodiments, the system includes:

[0007] The feeding unit is used to convey the door cover to be assembled to the centering table;

[0008] The hinge adjustment unit is used to adjust the position of the hinge relative to the door cover on the centering table. It includes a hinge adjustment mechanism, a first bolt tightening robot, a first image acquisition module, and a first data processing module. The first image acquisition module includes multiple first 3D line scan cameras and multiple first 3D area array cameras; the first 3D line scan cameras are arranged on the centering table, the first 3D area array cameras are arranged on the first bolt tightening robot, and the first data processing module is electrically connected to the first image acquisition module;

[0009] The door cover assembly and adjustment unit is used for the automatic assembly and adjustment of the door cover and the vehicle body. It includes a door cover adjustment robot, a second image acquisition device, a second bolt tightening robot, and a second data processing module. The second image acquisition module includes multiple second 3D line scan cameras and multiple second 3D area array cameras; the second 3D line scan cameras are arranged at the end of the door cover adjustment robot, the second 3D area array cameras are arranged at the end of the second bolt tightening robot, and the second data processing module is electrically connected to the second image acquisition module;

[0010] The quality monitoring unit is used for automatically flattening the door cover and monitoring the dimensions of each measuring point;

[0011] The controller, when the door cover to be assembled needs to be assembled and adjusted, controls the feeding unit to convey the door cover to be assembled to the centering table, and controls the hinge adjustment unit to obtain the surface difference measurement data of the hinge on the centering table, calculate the compensation amount, and adjust the hinge according to the compensation amount so that the error between the hinge and the door cover is within a first preset range. Then, it controls the door cover assembly and adjustment unit to perform the best fit matching between the door cover after hinge adjustment and the vehicle body so that the gap surface difference measurement values of each measuring point between the door cover and the vehicle body are within a second preset range. Finally, it controls the quality monitoring unit to flatten the door cover so that the surface difference of the measuring point at the door handle of the door cover is within a third preset range.

[0012] In some of the embodiments, the feeding unit includes a feeding robot, a third image acquisition module, a ranging module, and a third data processing module; the third image acquisition module and the ranging module are both arranged at the end of the feeding robot, and the third data processing module is electrically connected to the third image acquisition module and the laser ranging module respectively;

[0013] The ranging module is used to obtain the first distance between the feeding robot and the door cover. The feeding robot moves to the first preset position according to the first distance. The third image acquisition module obtains the door cover image of the door cover at the first preset position and determines the first point cloud data according to the door cover image. The third data processing module determines the door cover deviation correction amount according to the first point cloud data. The controller controls the feeding robot to perform grasping according to the door cover deviation correction amount and move to the centering table.

[0014] In some of the embodiments, the hinge adjustment mechanism includes a tightening unit and a moving mechanism with a supporting block. The first 3D line scan camera measures the second distance between the supporting block and the hinge. The controller controls the moving mechanism to move to make the supporting block and the hinge in contact, and clamps the hinge through the cylinder fixture on the centering table. The first 3D area array camera is used to measure the first surface difference between the door cover and the hinge. The second data processing module is used to obtain the first surface difference and determine the compensation amount of the hinge according to the second surface difference of the surface of the body thread hole measured in advance. The controller controls the moving mechanism to adjust the hinge according to the compensation amount so that the error of the hinge is within the first preset range.

[0015] In some of the embodiments, the door cover adjustment robot includes an assembly fixture, and the bolt tightening robot includes a tightening gun. The second 3D line scan camera is arranged on the assembly fixture and is used to measure the gap surface difference between the door cover and the body. The second 3D area array camera is arranged on the tightening gun and is used to measure the center coordinates and normal vectors of the body thread hole or the hinge stud, and guide the tightening gun to perform deviation correction.

[0016] In some of the embodiments, the controller controls the first bolt tightening robot to move the first 3D area array camera to the second preset position and obtain the first center coordinates of the hinge hole. The controller controls the door cover adjustment robot to move the door cover after hinge adjustment to the third preset position and perform centering according to the first center coordinates and the thread hole of the body.

[0017] In some of the embodiments, the second image acquisition module is used to acquire the second point cloud data of the door cover and the body. The second data processing module is used to determine the characteristic values of the door cover and the body according to the second point cloud data, and determine the pose offset value between the door cover and the body according to the characteristic values. The controller controls the door cover adjustment robot to adjust the door cover according to the pose offset value so that the gap surface difference between the door cover and the body is within the second preset range.

[0018] In some of these embodiments, the quality monitoring unit includes a flattening device that adsorbs at the door handle of the door cover. The second image acquisition module is used to obtain the third point cloud data at the door handle. The second data processing module is used to determine the surface difference of the measurement points at the door handle according to the third point cloud data and determine the adjustment amount of the door cover according to the surface difference of the measurement points at the door handle. The controller controls the door cover adjustment robot to adjust the door cover according to the adjustment amount so that the surface difference of the measurement points at the door handle of the door cover is within a third preset range.

[0019] In a second aspect, the present application also provides a vision-guided method for assembling and adjusting an automobile door cover, which uses the vision-guided system for assembling and adjusting an automobile door cover described in the first aspect. The method includes:

[0020] Transport the door cover to be assembled to the centering table;

[0021] Determine the hinge compensation amount between the door cover and the hinge, and adjust the hinge according to the compensation amount so that the error of the hinge is within a first preset range;

[0022] Determine the pose offset amount between the door cover and the vehicle body, and adjust the door cover for optimal fitting and matching according to the pose offset value so that the gap surface difference of each measurement point between the door cover and the vehicle body is within the second preset range;

[0023] Determine the surface difference of the measurement points at the door handle of the door cover, and determine the adjustment amount of the door cover according to the surface difference of the measurement points at the door handle. Adjust the door cover according to the adjustment amount so that the surface difference of the measurement points at the door handle of the door cover is within a third preset range.

[0024] In some of these embodiments, transporting the door cover to be assembled to the centering table includes:

[0025] Obtain the door cover image of the door cover at a first preset position;

[0026] Determine the first point cloud data according to the door cover image;

[0027] Determine the door cover deviation correction amount according to the first point cloud data;

[0028] Grasp the door cover according to the door cover deviation correction amount and move it to the centering table.

[0029] In some of these embodiments, determining the hinge compensation amount between the door cover and the hinge, and adjusting the hinge according to the compensation amount so that the error of the hinge is within a first preset range includes:

[0030] Obtain the first surface difference between the door cover and the hinge on the centering table;

[0031] Determine the compensation amount of the hinge according to the first surface difference and the second surface difference of the surface of the body thread hole measured in advance;

[0032] Adjust the hinge according to the compensation amount so that the error of the hinge is within the first preset range.

[0033] In some of these embodiments, the best fit matching of the door cover after adjustment with the body is performed so that the gap surface difference between the door cover and the body is within the second preset range, including:

[0034] Obtain the first central coordinates of the hinge hole at the second preset position;

[0035] Move the door cover after hinge adjustment to the third preset position and align it according to the first central coordinates and the thread hole of the body;

[0036] Collect the second point cloud data of the door cover and the body;

[0037] Determine the characteristic values of the door cover and the body according to the second point cloud data;

[0038] Determine the pose offset value between the door cover and the body according to the characteristic values;

[0039] Adjust the door cover according to the pose offset value so that the gap surface difference between the door cover and the body is within the second preset range.

[0040] In some of these embodiments, determining the pose offset value between the door cover and the body according to the characteristic values includes:

[0041] Input the characteristic values into the trained complete door cover and body fitting model to output the pose offset value between the door cover and the body. Among them, the trained complete door cover and body fitting model is generated by inputting the point cloud data of the door cover and body CAD digital models in the virtual scene into the model and training until convergence, and is used to output the pose offset amount between the door cover and the body. The fitting model is trained based on an artificial intelligence algorithm.

[0042] In some of these embodiments, the training method of the trained complete door cover and body fitting model includes:

[0043] Obtain the point cloud data of the body and door cover CAD digital models in the virtual scene and the fitting model to be trained;

[0044] Adjust the values of the door cover and the body in the virtual scene to the preset position;

[0045] Obtain the point cloud data of the door cover and the body in different poses;

[0046] Input the point cloud data into the fitting model to be trained, and obtain the trained fitting model of the door cover and the vehicle body.

[0047] In some of these embodiments, determining the surface difference of the measurement points at the door handle of the door cover, determining the adjustment amount of the door cover according to the surface difference of the measurement points at the door handle, and adjusting the door cover according to the adjustment amount so that the surface difference of the measurement points at the door handle of the door cover is within a third preset range includes:

[0048] Obtain the third point cloud data at the door handle of the door cover;

[0049] Determine the surface difference of the measurement points at the door handle according to the third point cloud data;

[0050] Determine the adjustment amount of the door cover according to the surface difference of the measurement points at the door handle;

[0051] Adjust the door cover according to the adjustment amount so that the surface difference of the measurement points at the door handle of the door cover is within a third preset range.

[0052] The above-mentioned visual guidance system and method for assembling and adjusting an automotive door cover, by transporting the door cover to be assembled to the alignment table; determining the hinge compensation amount between the door cover and the hinge, and adjusting the hinge according to the compensation amount so that the error of the hinge is within a first preset range; determining the pose offset amount between the door cover and the vehicle body, and adjusting the door cover according to the pose offset value for optimal fitting and matching so that the gap surface difference of each measurement point between the door cover and the vehicle body is within a second preset range; determining the surface difference of the measurement points at the door handle of the door cover, determining the adjustment amount of the door cover according to the surface difference of the measurement points at the door handle, and adjusting the door cover according to the adjustment amount so that the surface difference of the measurement points at the door handle of the door cover is within a third preset range, performs suction and re-measurement to achieve the best matching of the automotive door cover, hinge, and vehicle body, solves the problem of large errors in the assembly process of automotive door covers in the prior art, and improves the accuracy and flexibility of automotive door covers in the assembly process. Description of the Drawings

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments of the present application or related technologies. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0054] Figure 1 It is a schematic structural diagram of a visual guidance system for assembling and adjusting an automotive door cover in an embodiment;

[0055] Figure 2 It is a schematic structural diagram of a hinge adjustment mechanism in an embodiment;

[0056] Figure 3 It is a schematic flow chart of a vision-guided method for assembling and adjusting automobile door covers in an embodiment. Specific implementation manners

[0057] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0058] In an exemplary embodiment, as Figure 1 shown, an automobile door cover assembling and adjusting system is provided, and the system includes:

[0059] A feeding unit 10 for conveying the door cover to be assembled to the centering table 60. The door cover to be assembled is generally placed in the door cover material box 70, and the feeding unit 10 takes the material from the door cover material box 70 when needed.

[0060] A hinge adjusting unit 20 for adjusting the position of the hinge relative to the door cover on the centering table 60, which includes a hinge adjusting mechanism, a first bolt tightening robot, a first image acquisition module and a first data processing module. The first image acquisition module includes multiple first 3D line scan cameras and multiple first 3D area array cameras; the first 3D line scan cameras are arranged on the centering table 60, the first 3D area array cameras are arranged on the first bolt tightening robot, and the first data processing module is electrically connected to the first image acquisition module.

[0061] A door cover assembling and adjusting unit 30 for automatically assembling and adjusting the door cover and the vehicle body 80, which includes a door cover adjusting robot, a second image acquisition device, a second bolt tightening robot and a second data processing module. The second image acquisition module includes multiple second 3D line scan cameras and multiple second 3D area array cameras; the second 3D line scan cameras are arranged at the end of the door cover adjusting robot, the second 3D area array cameras are arranged on the second bolt tightening robot, and the second data processing module is electrically connected to the second image acquisition module.

[0062] In this embodiment, the second 3D line scan camera can be fixed on the end fixture of the door cover adjusting robot, and the second 3D area array camera can be fixed on the end tightening device of the second bolt tightening robot.

[0063] A quality monitoring unit 40 for automatically flattening the door cover and monitoring the dimensions of each measuring point.

[0064] Controller 50, when the door cover to be assembled needs to be adjusted, controls the feeding unit 10 to convey the door cover to be assembled to the alignment table 60, and controls the hinge adjustment unit 20 to obtain the surface difference measurement data between the door cover and the hinge on the alignment table 60, calculate the compensation amount, and adjust the hinge according to the compensation amount so that the error between the hinge and the door cover is within the first preset range. Then, it controls the door cover adjustment unit 30 to perform the best-fit matching between the door cover after hinge adjustment and the vehicle body 80 so that the gap surface difference measurement values of each measuring point between the door cover and the vehicle body 80 are within the second preset range. Finally, it controls the quality monitoring unit 40 to flatten the door cover so that the gap surface difference at the measuring point of the door cover handle is within the third preset range. Through the above method, the best matching of the door cover, hinge, and vehicle body is achieved, solving the problem of large errors in the assembly process of the door cover in the prior art and improving the accuracy and flexibility of the door cover in the assembly process.

[0065] In some of these embodiments, the feeding unit 10 includes a feeding robot, a third image acquisition module, a ranging module, and a third data processing module; the third image acquisition module and the ranging module are both arranged at the end of the feeding robot, and the third data processing module is electrically connected to the third image acquisition module and the laser ranging module respectively; the ranging module is used to obtain the first distance between the feeding robot and the door cover, and the feeding robot moves to the first preset position according to the first distance. The third image acquisition module obtains the door cover image of the door cover at the first preset position and determines the first point cloud data according to the door cover image. The third data processing module determines the door cover deviation amount according to the first point cloud data, and the controller controls the feeding robot to grasp and move to the alignment table 60 according to the door cover deviation amount, avoiding the problem of large errors in the subsequent installation of the door cover due to offset when the door cover moves to the alignment table 60, and reducing the error in the assembly process of the door cover.

[0066] Reference Figure 2, in some of these embodiments, the hinge adjustment mechanism includes a tightening unit 25 and a moving mechanism 22 with a supporting block. The first 3D line scan camera 23 measures the second distance between the supporting block and the hinge 21. The controller controls the moving mechanism 22 to move until the supporting block is in contact with the hinge 21 and clamps the hinge through the cylinder fixture on the alignment table 60. The first 3D area array camera 23 is used to measure the first surface difference between the door panel 90 and the hinge. The second data processing module is used to obtain the first surface difference and determine the compensation amount of the hinge 21 based on the pre-measured second surface difference of the surface of the body thread holes. The controller controls the moving mechanism 22 and the tightening unit 25 to adjust the hinge 21 according to the compensation amount, so that the error of the hinge 21 is within the first preset range. In this embodiment, by jointly measuring the distance between the supporting block and the hinge 21, the surface difference between the door panel 90 and the hinge 21, and the manufacturing error of the body 80, calculating the compensation amount, and driving the moving mechanism 22 and the tightening unit 25 to complete the position adjustment of the hinge 21, it avoids the problem in the prior art that only the adjustment of the door panel is achieved, resulting in a large error during the assembly of the door panel, thereby improving the accuracy of the door panel during the assembly process.

[0067] It should be noted that the pre-measured second surface difference of the surface of the body thread holes can be pre-measured and stored in a storage medium offline.

[0068] In some of these embodiments, the door panel adjustment robot includes a mounting fixture, the bolt tightening robot includes a tightening gun, the second 3D line scan camera is arranged on the mounting fixture and is used to measure the clearance surface difference between the door panel 90 and the body 80, and the second 3D area array camera is arranged on the tightening gun and is used to measure the center coordinates and normal vectors of the body thread holes or hinge studs to guide the tightening gun to correct the deviation. Through dynamic measurement of the clearance surface difference and visual servo guidance, the adaptive matching and fitting of the door panel and the body are realized.

[0069] In some of these embodiments, the controller controls the first bolt tightening robot to move the first 3D area array camera to a second preset position and obtain the first center coordinates of the hinge holes. The controller controls the door panel adjustment robot to move the door panel after hinge adjustment to a third preset position and align it according to the first center coordinates and the thread holes of the body 80.

[0070] In some of these embodiments, the second image acquisition module is used to acquire the second point cloud data of the door panel and the body 80. The second data processing module is used to determine the characteristic values of the door panel and the body 80 based on the second point cloud data, and determine the pose offset value between the door panel and the body 80 according to the characteristic values. The controller controls the door panel adjustment robot to adjust the door panel according to the pose offset value, so that the clearance surface difference between the door panel and the body 80 is within the second preset range.

[0071] In some of these embodiments, the quality monitoring unit includes a flattening device that adsorbs at the door handle of the door cover. The second image acquisition module is used to obtain the third point cloud data of the measurement point at the door handle. The second data processing module is used to determine the surface difference of the measurement point at the door handle based on the third point cloud data and determine the adjustment amount of the door cover according to the surface difference of the measurement point at the door handle. The controller controls the door cover adjustment robot to adjust the door cover according to the adjustment amount so that the surface difference of the measurement point at the door handle of the door cover is within the third preset range.

[0072] In this embodiment, a flattening device is used to adsorb the door handle, and the local adjustment amount is calculated based on the third point cloud data, and the displacement of the suction cup is closed-loop controlled to achieve fine adjustment in the door handle area.

[0073] See Figure 3 , this application also provides a visual guidance method for assembling and adjusting the car door cover, which can be applied to the visual guidance system for assembling and adjusting the car door cover in the above embodiments. The steps include:

[0074] Step S302, convey the door cover to be assembled to the alignment table to ensure the standardization of the initial position of the door cover to be assembled.

[0075] Step S304, determine the hinge compensation amount between the door cover and the hinge, and adjust the door cover and the hinge according to the compensation amount so that the error between the door cover and the hinge is within the first preset range.

[0076] Based on steps S302 to S304, convey the door cover to be assembled to the alignment table to ensure the standardization of the initial position.

[0077] Step S306, determine the pose offset amount between the door cover and the vehicle body, and adjust the door cover according to the pose offset value for optimal fitting and matching so that the gap surface difference of each measurement point between the door cover and the vehicle body is within the second preset range;

[0078] Step S308, determine the surface difference of the measurement point at the door handle of the door cover, and determine the adjustment amount of the door cover according to the surface difference of the door handle, and adjust the door cover according to the adjustment amount so that the surface difference of the measurement point at the door handle of the door cover is within the third preset range.

[0079] Based on the above steps S302 to S304, the compensation between the door cover and the hinge is realized. Then, based on step S306, the best dimensional matching between the door cover and the vehicle body is realized, reducing the uncontrollability of traditional manual debugging, enabling the door cover, hinge, and vehicle body to be best matched and adhered, solving the problem of large errors in the assembly process of the door cover in the prior art, and improving the accuracy and flexibility of the door cover in the assembly process.

[0080] In some of these embodiments, conveying the door cover to be assembled to the centering table includes: acquiring an image of the door cover at a first preset position; determining first point cloud data based on the door cover image; determining a door cover alignment correction amount based on the first point cloud data; and grasping and moving the door cover to the centering table according to the door cover alignment correction amount.

[0081] In some of these embodiments, determining a hinge compensation amount between the door cover and the hinge and adjusting the hinge according to the compensation amount so that the error of the hinge is within a first preset range includes: acquiring a first surface difference between the door cover and the hinge on the centering table; determining the compensation amount of the hinge according to the first surface difference and a second surface difference of the surface of the body thread hole measured in advance; and adjusting the hinge according to the compensation amount so that the error of the hinge is within the first preset range.

[0082] In some of these embodiments, optimally fitting and matching the adjusted door cover with the body so that the gap surface difference between the door cover and the body is within a second preset range includes: acquiring a first center coordinate of the hinge hole at a second preset position; moving the door cover after hinge adjustment to a third preset position and centering it according to the first center coordinate and the thread hole of the body; acquiring second point cloud data of the door cover and the body; determining characteristic values of the door cover and the body according to the second point cloud data; determining a pose offset value between the door cover and the body according to the characteristic values; and adjusting the door cover according to the pose offset value so that the gap surface difference between the door cover and the body is within the second preset range.

[0083] In some of these embodiments, determining the pose offset value between the door cover and the body according to the characteristic values includes: inputting the characteristic values into a trained and complete door cover and body fitting model to output the pose offset value between the door cover and the body, where the trained and complete door cover and body fitting model is generated by inputting the point cloud data of the CAD digital models of the door cover and the body in a virtual scene into the model and training until convergence, and is used to output the pose offset amount between the door cover and the body, and the fitting model is trained based on an artificial intelligence algorithm.

[0084] Through the above process, the artificial intelligence algorithm model can achieve high-precision prediction of the pose offset amount between the door cover and the body, significantly improving the level of assembly automation.

[0085] In some of these embodiments, the training method of the trained and complete door cover and body fitting model includes: acquiring the point cloud data of the CAD digital models of the body and the door cover in a virtual scene and the fitting model to be trained; adjusting the values of the door cover and the body in the virtual scene to a preset position; acquiring the point cloud data of the door cover and the body in different poses; and inputting the point cloud data into the fitting model to be trained to obtain a trained and complete door cover and body fitting model. In the above manner, a high-precision door cover-body pose fitting model can be trained to achieve effective migration from virtual data to the actual assembly scene.

[0086] In some of these embodiments, determining the surface difference at the measurement point of the door handle of the door cover, determining the adjustment amount of the door cover according to the surface difference at the measurement point of the door handle, and adjusting the door cover according to the adjustment amount so that the surface difference at the measurement point of the door handle of the door cover is within a third preset range includes: obtaining the third point cloud data at the door handle of the door cover; determining the surface difference at the measurement point of the door handle according to the third point cloud data; determining the adjustment amount of the door cover according to the surface difference at the measurement point of the door handle; and adjusting the door cover according to the adjustment amount so that the surface difference at the measurement point of the door handle of the door cover is within a third preset range.

[0087] The technical effects produced by the method of this application are similar to those of the system and will not be elaborated here.

[0088] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this application.

[0089] The above-described embodiments only represent several implementation manners of this application. Their descriptions are relatively specific and detailed, but they should not be construed as limiting the patent scope of this application. It should be noted that for those of ordinary skill in the art, without departing from the concept of this application, several modifications and improvements can still be made, and these all belong to the protection scope of this application. Therefore, the protection scope of this application should be subject to the appended claims.

Claims

1. A vision-guided automobile door and hood assembly and adjustment system, characterized in that, The system includes: a loading unit for conveying the door cover to be assembled to the centering table; a hinge adjustment unit for adjusting the position of the hinge relative to the door cover on the centering table, which includes a hinge adjustment mechanism, a first bolt tightening robot, a first image acquisition module and a first data processing module. The first image acquisition module includes multiple first 3D line scan cameras and multiple first 3D area array cameras; the first 3D line scan cameras are arranged on the centering table, the first 3D area array cameras are arranged on the first bolt tightening robot, and the first data processing module is electrically connected to the first image acquisition module; a door cover assembly and adjustment unit for automatically assembling and adjusting the door cover and the vehicle body, which includes a door cover adjustment robot, a second image acquisition device, a second bolt tightening robot and a second data processing module. The second image acquisition module includes multiple second 3D line scan cameras and multiple second 3D area array cameras; the second 3D line scan cameras are arranged at the end of the door cover adjustment robot, the second 3D area array cameras are arranged at the end of the second bolt tightening robot, and the second data processing module is electrically connected to the second image acquisition module; a quality monitoring unit for automatically leveling the door cover and monitoring the dimensions of each measurement point; a controller. When the door cover to be assembled needs to be assembled and adjusted, the controller controls the loading unit to convey the door cover to be assembled to the centering table, and controls the hinge adjustment unit to obtain the surface difference measurement data of the hinge on the centering table, calculate the compensation amount, and adjust the hinge according to the compensation amount so that the error between the hinge and the door cover is within a first preset range. Then, it controls the door cover assembly and adjustment unit to perform the best fitting and matching between the door cover after hinge adjustment and the vehicle body so that the measured values of the gap surface difference at each measurement point between the door cover and the vehicle body are within a second preset range. Finally, it controls the quality monitoring unit to level the door cover so that the surface difference at the measurement point of the door handle of the door cover is within a third preset range.

2. The system according to claim 1, wherein The loading unit includes a loading robot, a third image acquisition module, a ranging module and a third data processing module; the third image acquisition module and the ranging module are both arranged at the end of the loading robot, and the third data processing module is electrically connected to the third image acquisition module and the laser ranging module respectively; The ranging module is used to obtain the first distance between the loading robot and the door cover. The loading robot moves to a first preset position according to the first distance. The third image acquisition module obtains the door cover image of the door cover at the first preset position and determines the first point cloud data according to the door cover image. The third data processing module determines the door cover deviation amount according to the first point cloud data, and the controller controls the loading robot to grab and move to the centering table according to the door cover deviation amount.

3. The system according to claim 1, wherein The hinge adjustment mechanism includes a tightening unit and a moving mechanism with a supporting block. The first 3D line-scanning camera measures the second distance between the supporting block and the hinge. The controller controls the moving mechanism to move to make the supporting block contact the hinge, and clamps the hinge through a cylinder fixture on the centering table. The first 3D area array camera is used to measure the first surface difference between the door panel and the hinge. The second data processing module is used to obtain the first surface difference and determine the compensation amount of the hinge according to the second surface difference of the surface of the body thread hole measured in advance. The controller controls the moving mechanism to adjust the hinge according to the compensation amount, so that the error of the hinge is within the first preset range.

4. The system according to claim 1, characterized in that, The door panel adjustment robot includes an assembly fixture. The bolt tightening robot includes a tightening gun. The second 3D line-scanning camera is arranged on the assembly fixture and is used to measure the gap surface difference between the door panel and the body. The second 3D area array camera is arranged on the tightening gun and is used to measure the central coordinates and normal vectors of the body thread hole or the hinge stud, and guide the tightening gun to correct the deviation.

5. The system according to claim 1, characterized in that, The controller controls the first bolt tightening robot to move the first 3D area array camera to a second preset position and obtain the first central coordinates of the hinge hole. The controller controls the door panel adjustment robot to move the door panel after hinge adjustment to a third preset position, and perform centering according to the first central coordinates and the thread hole of the body.

6. The system according to claim 1, characterized in that, The second image acquisition module is used to acquire the second point cloud data of the door panel and the body. The second data processing module is used to determine the characteristic values of the door panel and the body according to the second point cloud data, and determine the pose offset value between the door panel and the body according to the characteristic values. The controller controls the door panel adjustment robot to adjust the door panel according to the pose offset value, so that the gap surface difference between the door panel and the body is within the second preset range.

7. The system according to claim 1, wherein The quality monitoring unit includes a flattening device. The flattening device adsorbs at the door handle of the door panel. The second image acquisition module is used to obtain the third point cloud data at the door handle. The second data processing module is used to determine the surface difference of the measurement point at the door handle according to the third point cloud data and determine the adjustment amount of the door panel according to the surface difference of the measurement point at the door handle. The controller controls the door panel adjustment robot to adjust the door panel according to the adjustment amount, so that the surface difference of the measurement point at the door handle of the door panel is within the third preset range.

8. A visual guidance method for adjusting and installing automobile door covers, which uses the automobile door cover adjusting and installing system described in any one of claims 1 to 7, and is characterized in that, The method includes: Convey the door panel to be assembled to the centering table. Determine the hinge compensation amount between the door panel and the hinge, and adjust the hinge according to the compensation amount, so that the error of the hinge is within the first preset range. Determine the pose offset amount between the door panel and the body, and adjust the door panel according to the pose offset value for optimal fitting and matching, so that the gap surface difference of each measurement point between the door panel and the body is within the second preset range. Determine the surface difference at the measurement point of the door cover handle, determine the adjustment amount of the door cover according to the surface difference at the measurement point of the handle, and adjust the door cover according to the adjustment amount so that the surface difference at the measurement point of the door cover handle is within the third preset range.

9. The method according to claim 8, characterized in that Conveying the door cover to be assembled to the centering table includes: Obtain the door cover image of the door cover at the first preset position; Determine the first point cloud data according to the door cover image; Determine the door cover deviation correction amount according to the first point cloud data; Grasp the door cover according to the door cover deviation correction amount and move it to the centering table.

10. The method according to claim 8, wherein Determine the hinge compensation amount between the door cover and the hinge, and adjust the hinge according to the compensation amount so that the error of the hinge is within the first preset range includes: Obtain the first surface difference between the door cover and the hinge on the centering table; Determine the compensation amount of the hinge according to the first surface difference and the pre-measured second surface difference of the surface of the body thread hole; Adjust the hinge according to the compensation amount so that the error of the hinge is within the first preset range.

11. The method according to claim 10, wherein Perform the best fit matching between the adjusted door cover and the body so that the gap surface difference between the door cover and the body is within the second preset range includes: Obtain the first center coordinates of the hinge hole at the second preset position; Move the door cover after hinge adjustment to the third to preset position, and perform centering according to the first center coordinates and the thread hole of the body; Collect the second point cloud data of the door cover and the body; Determine the characteristic values of the door cover and the body according to the second point cloud data; Determine the pose offset value between the door cover and the body according to the characteristic values; Adjust the door cover according to the pose offset value so that the gap surface difference between the door cover and the body is within the second preset range.

12. The method according to claim 10, characterized in that, Determine the pose offset value between the door cover and the body according to the characteristic values includes: Input the characteristic values into the trained complete door cover and body fitting model, and output the pose offset value between the door cover and the body. Among them, the trained complete door cover and body fitting model is generated by inputting the point cloud data of the door cover and body CAD digital models in the virtual scene into the model and training until convergence, and is used to output the pose offset amount between the door cover and the body. The fitting model is trained based on the artificial intelligence algorithm.

13. The method according to claim 12, wherein The training method of the trained complete door cover and body fitting model includes: Obtain the point cloud data of the body and door cover CAD digital models in the virtual scene and the fitting model to be trained; Adjust the values of the door cover and the body in the virtual scene to the preset position; Obtain the point cloud data of the door cover and the body in different poses; Input the point cloud data into the fitting model to be trained to obtain the trained complete door cover and body fitting model.

14. The method according to claim 8, wherein Determine the surface difference at the measurement point of the door cover handle, determine the adjustment amount of the door cover according to the surface difference at the measurement point of the handle, and adjust the door cover according to the adjustment amount so that the surface difference at the measurement point of the door cover handle is within the third preset range includes: Obtain the third point cloud data at the door cover handle of the door cover; Determine the surface difference at the measurement point of the handle according to the third point cloud data; Determine the adjustment amount of the door cover according to the surface difference of the measurement point at the door handle; Adjust the door cover according to the adjustment amount so that the surface difference of the measurement point at the door handle of the door cover is within a third preset range.

Images