A repair welding system and repair welding method for a vehicle frame

Abstract

This application discloses a vehicle frame repair welding system and method, relating to the fields of visual inspection and industrial control technology. By combining a vision system to identify defect information of welds on the vehicle frame, it can avoid the problems of inconsistent standards and low efficiency caused by manual experience judgment. In addition, during the scanning of each weld, the absolute value of the difference between the lens's focusing distance and the preset focal length is less than a preset first difference threshold, which can ensure the clarity of the weld image and the integrity of the defect details, thereby improving the recognition accuracy. Combined with a repair welding robot, it can achieve fully automated repair welding, ensuring both repair welding efficiency and consistent repair welding quality.

Description

Technical Field

[0001] This application relates to the fields of visual inspection and industrial control technology, and in particular to a chassis repair welding system and repair welding method. Background Technology

[0002] As the core load-bearing component of a car, the welding quality of the chassis directly determines the structural strength and driving safety of the entire vehicle. Defects such as microcracks, porosity, and incomplete welds at the weld seams will continue to expand under the vibration and load of long-term driving, eventually leading to serious safety accidents such as chassis fracture. Therefore, after the chassis welding is completed, the weld seams must be inspected, and any defects detected must be repaired by welding.

[0003] In the traditional approach, weld quality is inspected and repaired manually based on the experience of skilled workers. However, this method relies heavily on the experience of skilled workers, and different workers have different criteria for judging defects such as microcracks and porosity. Some small defects that are difficult to hide are easily missed. The control of penetration depth and weld formation during repair welding is entirely based on feel, resulting in inconsistent repair welding quality and making it difficult to guarantee the welding consistency of the same batch of frames. Summary of the Invention

[0004] This application provides a frame repair welding system and method, which can improve the efficiency of repair welding and ensure the consistency of frame welding quality.

[0005] In a first aspect, this application provides a chassis welding repair system, comprising: The upper-level device is used to acquire a three-dimensional model of the entire frame, determine the regional position and shape classification of each weld seam based on the three-dimensional model, and generate a scanning path based on the regional position and shape classification of each weld seam. The scanning path includes robot position parameters and robot posture parameters corresponding to multiple scanning points. The scanning path ensures that the absolute value of the difference between the lens focusing distance and the preset focal length is less than a preset first difference threshold during the scanning of each weld seam by the vision system. The robot position parameters and robot posture parameters are parameters of the robot used to drive the movement of the vision system. A vision inspection and control module is used to control the robot that drives the vision system to perform weld seam scanning based on the scanning path. The vision system is used to acquire weld images at different locations along a scanning path under the control of a robot that drives the vision system's movement. Based on each weld image, the system identifies the corresponding weld defect information and transmits the defect information to the upper-level device. The defect information includes the weld defect type and the weld defect location. The upper-level device is also used to receive defect information of each weld seam transmitted by the vision system, generate corresponding repair welding control programs based on the defect information of each weld seam, and send them to the repair welding control module. The welding repair control module is used to receive the welding repair control logic for each weld from the upper-level device, and drive the welding repair robot to complete the welding repair operation of the frame based on the welding repair control program.

[0006] In one alternative embodiment of the first aspect, the upper-layer device is specifically used for: Extract the region location and point set data of each weld from the three-dimensional model; Based on the coordinates of each point in the point set data of each weld, the curvature of each point is calculated; The morphological classification of each weld is determined based on the curvature of each point.

[0007] In one alternative embodiment of the first aspect, the upper-layer device is specifically used for: Multiple center points of each weld area are extracted and connected to form a scanning baseline; Based on the length of the scanning baseline, the scanning baseline is divided into at least one line segment, such that the length of all the resulting line segments is within a preset length range, and the number of the resulting line segments is minimized under the condition of satisfying the length constraint. The intersection of adjacent line segments is used as the visual focus, and the position and attitude parameters of the robot used to drive the movement of the vision system are determined based on the coordinates of each visual focus.

[0008] In an optional embodiment of the first aspect, determining the position and attitude parameters of the robot used to drive the motion of the vision system at the scanning point based on the coordinates of each visual focus includes: Calculate the normal vector of the visual focus based on its coordinates; Based on the normal vector, coordinates, and preset focal length of the visual focus, the robot position parameters and robot posture parameters at the corresponding scanning points are generated.

[0009] In one alternative embodiment of the first aspect, the morphological classification includes straight welds and curved welds.

[0010] In one alternative embodiment of the first aspect, the upper-layer device is specifically used for: For straight welds, the length of the resulting segment is set to be within a first length range; For the scanning baseline of the curved weld, the scanning baseline is divided into smooth segments and / or non-smooth segments according to the curvature of the points on the scanning baseline. The smooth segments are further segmented according to their lengths, and / or the non-smooth segments are further segmented according to their lengths. The lengths of the segments obtained from the smooth segments are within a second length range, and the lengths of the segments obtained from the non-smooth segments are within a third length range. The first length range has the largest interval, the second length range has the second largest interval, and the third length range has the smallest interval. Furthermore, the maximum value of the first length range is greater than the maximum value of the second length range, which in turn is greater than the maximum value of the third length range.

[0011] In one alternative embodiment of the first aspect, the upper-layer device divides the smooth segments and non-smooth segments in the following manner: A continuous segment on the scanning baseline of a curved weld where the absolute value of the curvature difference between k consecutive points is less than a preset threshold is defined as a smooth segment. A point is defined as an inflection point when the absolute value of the curvature difference between it and the two points before and after it is greater than or equal to a preset threshold. A continuous segment consisting of an inflection point and z points before and after it is defined as a non-smooth segment, where k and z are both integers greater than or equal to 1.

[0012] Secondly, this application provides a method for repairing and welding a vehicle frame, applied to the vehicle frame repair and welding system provided in the first aspect. The method for repairing and welding the vehicle frame includes: The vision system is driven to acquire weld images at different locations along the scanning path. The scanning path ensures that the absolute value of the difference between the lens's focusing distance and the preset focal length is less than a preset first difference threshold during the scanning process of each weld on the vehicle frame. Defect information of the corresponding weld is identified based on the images of each weld. A corresponding repair welding control program is generated based on the defect information of each weld. The welding robot is driven by a welding control program to complete the welding repair work on the chassis.

[0013] The frame repair welding system and welding method provided in this application have the following beneficial effects: By combining a vision system to identify defect information in welds on the chassis, the problems of inconsistent standards and low efficiency caused by human experience-based judgment can be avoided. In addition, during the scanning of each weld, the absolute value of the difference between the lens's focusing distance and the preset focal length is less than a preset first difference threshold, which can ensure the clarity of the weld image and the integrity of the defect details, thereby improving the recognition accuracy. Combined with a welding repair robot, fully automated welding repair can be achieved, ensuring both welding repair efficiency and consistent welding quality. Attached Figure Description

[0014] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0015] Figure 1 A structural diagram of an embodiment of the welding system for the vehicle frame provided in this application; Figure 2 A schematic diagram illustrating the division of straight weld seams provided in this application; Figure 3 A schematic diagram illustrating the division of curved surface weld seams provided in this application; Figure 4 This is a flowchart illustrating an embodiment of the vehicle frame repair welding method provided in this application. Detailed Implementation

[0016] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0017] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0018] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0019] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0020] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0021] The welding repair system of this application embodiment will be described in detail below with reference to the accompanying drawings.

[0022] Please see Figure 1The welding repair system proposed in this application includes an upper-level device 1, a vision inspection and control module 2, a vision system 3, and a welding repair control module 4. The upper-level device 1 can be a host computer with data processing capabilities, responsible for core calculations such as 3D model analysis, weld feature extraction, and scanning path planning, and generating corresponding analysis results and control commands. The vision inspection and control module 2 drives the robot (i.e., the vision robot) scanned by the vision system 3 to move along the planned path to scan the weld. The welding repair control module 4 receives the welding repair control program issued by the upper-level device and drives the welding repair robot to complete the welding repair operation of the defective weld.

[0023] The vision robot and the welding repair robot can be the same robot. In this case, the vision detection control module 2 and the welding repair control module 4 can be the same microcontroller unit in the robot, which executes scanning and welding repair commands respectively by switching working modes. Alternatively, they can be different microcontroller units, each independently responsible for the corresponding motion control.

[0024] The vision robot and the welding repair robot can also be different robots. The vision detection and control module 2 is the microcontroller unit in the vision robot, and the welding repair control module 4 is the microcontroller unit in the welding repair robot. The two robots can work together through the upper device 1.

[0025] The vision system 3 may include an RGBD camera or an RGB camera to acquire images of the weld seam and provide data support for defect identification.

[0026] The following is a detailed description of each part of the system.

[0027] The upper-level device is used to acquire a three-dimensional model of the entire frame, determine the regional location and shape classification of each weld seam based on the three-dimensional model, and generate a scanning path for the vision system based on the regional location and shape classification of each weld seam. The scanning path includes robot position parameters and robot posture parameters corresponding to multiple scanning points. The scanning path ensures that the absolute value of the difference between the lens focusing distance and the preset focal length is less than a preset first difference threshold during the scanning of each weld seam. The robot position parameters and robot posture parameters are parameters of the robot used to drive the movement of the vision system.

[0028] The vision inspection and control module is used to perform weld seam scanning on a robot that drives the vision system based on the scanning path control.

[0029] A vision system is used to acquire weld images at different locations along a scanning path under the control of a robot that drives the vision system's movement.

[0030] The weld location refers to a rough range of the weld's location, which is larger than the fine location. Weld morphology is classified into straight welds and curved welds.

[0031] Optionally, the upper-level device can extract the regional location and point set data of each weld from the three-dimensional model; calculate the curvature of each point based on the coordinates of each point in the point set data of each weld; and determine the morphological classification of each weld based on the curvature of each point.

[0032] The overall 3D model of the chassis serves as the original 3D design model for the chassis product. This model records the structural dimensions of each metal component of the chassis, the assembly positional relationships between components, and connection boundary information such as lap joints, butt joints, and corner joints. The upper-level device identifies weld areas and determines weld morphology through model analysis and geometric calculations, as follows: The upper-level device traverses the assembly boundaries of all components within the original design model, and filters out target areas that meet the welding process requirements. These areas include two categories: one is the overlapping surfaces, butt joint edges, and corner joint junctions between various components of the frame, which are the core load-bearing parts of the frame; the other is the welding candidate areas pre-marked in the design, which, combined with the frame assembly process documents, locks in the precise range and finally extracts the independent area positions corresponding to each weld.

[0033] Then, for each identified weld area, the upper device extracts the global discrete point set data within the area along the extension direction of the weld, and further extracts the coordinates of all discrete points on the outer contour of the weld to form a complete point dataset for a single weld, containing the three-dimensional coordinates of each point, providing basic data for subsequent curvature calculation.

[0034] Next, for each discrete point on the outer contour of a single weld, a curve fitting is performed on multiple adjacent points before and after that point to calculate the curvature value of that point, in order to judge the curvature of the entire weld contour: if the curvature values ​​of discrete points on all sides of the weld contour are less than the preset minimum curvature threshold (approaching 0, which can be set according to the process settings), and the absolute value of the curvature difference of multiple consecutive points is less than the preset minimum curvature threshold, and the contour trajectory has no obvious bending change, then the weld is determined to be a straight weld; if the curvature value of any discrete point on the outer contour of the weld is greater than or equal to the preset minimum curvature threshold, and the contour trajectory exhibits arc, bend, or curve, then the weld is determined to be a curved weld, and further distinctions are made between smooth and non-smooth segments within the curved weld.

[0035] Alternatively, the 3D model may be a model generated by a preliminary global rough scan of the finished vehicle frame using an RGBD camera. During the scanning process, the lens was not subject to focal length constraints, and the focusing distance changed arbitrarily with the unevenness and curvature of the vehicle frame surface. It could only restore the overall outline of the vehicle frame, and the point accuracy and outline clarity of the weld area were low.

[0036] Furthermore, the upper-level device has two processing paths for this rough scan model: one is to accurately register it with the original chassis design model and correct any deviations, eliminating point errors and contour distortions generated during the scanning process, and then repeating the above processing steps for the design model to complete the extraction and morphological classification of the weld area; the other is to directly locate the weld features based on the images captured by the camera without relying on the original design model, using image feature recognition algorithms to directly identify and extract the corresponding area position of the weld, and then simultaneously extract the coordinates of discrete points on the outer contour of the weld, calculate the curvature of each point, and complete the morphological determination of straight welds and curved welds.

[0037] Optionally, the upper-level device is specifically used for: Multiple center points of each weld area are extracted and connected to form a scanning baseline; Based on the length of the scanning baseline, the scanning baseline is divided into at least one line segment, such that the length of all the resulting line segments is within a preset length range, and the number of the resulting line segments is minimized under the condition of satisfying the length constraint. The intersection of adjacent line segments is used as the visual focus, and the position and attitude parameters of the robot used to drive the movement of the vision system are determined based on the coordinates of each visual focus.

[0038] It should be noted that the center point of the weld area refers to the center point of each sub-region after dividing the outer contour strip area of ​​the weld into several sub-regions along the weld trajectory. For curved welds, the center points of each sub-region need to be connected according to the weld trajectory.

[0039] Optionally, for straight welds, the length of the resulting segments is set to be within a first length range; for the scanning reference line of curved welds, the scanning reference line is divided into smooth segments and / or non-smooth segments according to the curvature of the points on the scanning reference line, and segmented according to the length of the smooth segments, and / or segmented according to the length of the non-smooth segments; wherein, the length of the resulting segments of the smooth segments is within a second length range, and the length of the resulting segments of the non-smooth segments is within a third length range; The three preset lengths satisfy the following size relationship: the first length range has the largest interval, the second length range has the second largest interval, and the third length range has the smallest interval, and the maximum value of the first length range is greater than the maximum value of the second length range, which is greater than the maximum value of the third length range.

[0040] Specifically, after classifying weld morphology, the upper-level device plans differentiated scanning point layout rules for straight welds and curved welds. The layout rules are as follows: the smoother the weld morphology, the larger the point spacing; the more complex the weld morphology, the smaller the point spacing, in order to balance overall detection efficiency and defect identification accuracy. The specific layout method is as follows: For straight welds: Straight welds present as rectangular areas. The upper device first extracts the central axis of this area as the scanning baseline. Then, based on the coordinates of the endpoints of the scanning baseline, it calculates the straight length of the entire scanning baseline. The entire scanning baseline is then divided into several line segments, ensuring that the length of all resulting segments is within a preset first length range. Furthermore, under this length constraint, the number of resulting line segments is minimized. Finally, the intersection of adjacent line segments is used as the visual focus. Figure 2 As shown, the black dots represent the visual focus, and the length of the resulting line segment is L1.

[0041] Next, each visual focus is converted into the position and posture parameters of the visual robot. This conversion process is based on kinematic principles and spatial coordinate transformation, which is existing technology and will not be elaborated on in detail. Finally, the scanning points and focus planning for the entire straight weld seam are completed, ensuring that all scanning points and corresponding visual scanning focuses are evenly distributed along the central axis of the weld seam, with no overlap or gaps between points, and completely covering the entire straight weld seam area.

[0042] For curved surface welds: The upper device first extracts the center point of the outer contour area of ​​the weld, connects each center point in sequence to form a continuous scanning baseline, and then divides the entire scanning baseline into smooth segments and non-smooth segments based on the curvature of each point on the scanning baseline.

[0043] A smooth segment is a line segment with continuous curvature without abrupt changes and a gentle trajectory. The criterion for its judgment is: a continuous segment on the scanning baseline of the curved weld where the absolute value of the curvature difference between k consecutive points is less than a preset threshold is defined as a smooth segment. For example, for 10 consecutive points, the absolute value of the curvature difference between any two adjacent points is less than the preset threshold.

[0044] Non-smooth segments are those containing curvature abrupt change inflection points and surrounding line segments. They belong to key detection areas where defects are frequent and small defects are easily hidden. The judgment criteria are as follows: points whose absolute values ​​of curvature differences with the two points before and after them are both greater than or equal to a preset threshold are defined as inflection points, and continuous segments consisting of inflection points and z points before and after them are defined as non-smooth segments, where k and z are both integers greater than or equal to 1.

[0045] Subsequently, based on the endpoint coordinates of the smooth segment, the length (arc length) of the entire smooth segment is calculated. Then, the entire smooth segment is divided into several line segments, ensuring that the length of all resulting line segments falls within a preset second length range, and that the number of line segments is minimized while satisfying this length constraint. Similarly, the non-smooth segment is divided into several line segments, ensuring that the length of all resulting line segments falls within a preset third length range, and that the number of line segments is minimized while satisfying this length constraint. Figure 3 As shown, Figure 3There are two smooth segments, with lengths L2 and L4 respectively, and the non-smooth segment has a length of L3.

[0046] It should be noted that the scanning point in this application refers to the target stopping point selected by the vision system on the extension trajectory of the weld. After the vision robot moves to the preset position corresponding to the scanning point and switches to the matching preset posture, the vision system (camera) will complete one or more image scanning acquisitions of the weld area corresponding to the point, and finally obtain the weld image based on the visual focus. One or more continuous visual focuses are set up along the extension trajectory of a single weld, and multiple local weld images covering the entire weld area are generated accordingly, so as to realize the detection of the entire weld without blind spots.

[0047] In existing conventional automated inspection technologies, most adopt a uniform scanning point interval and image acquisition frequency for all types of welds, without differentiating and optimizing based on the actual weld morphology. However, for straight welds or smooth curved sections within curved welds, the weld trajectory is gentle and the defect distribution is regular. Overly dense scanning points will generate a large amount of redundant image data, significantly increasing the data processing load and reducing overall inspection efficiency. For non-smooth sections in curved welds, especially inflection points where curvature changes abruptly, the weld surface has large undulations, and small defects are easily hidden at the curvature abrupt changes. Overly sparse scanning points will lead to insufficient sampling in this area, failing to fully capture defect details, thus forming blind spots and resulting in missed detection of defects such as microcracks and porosity.

[0048] Optionally, the upper-level device is specifically used for: Calculate the normal vector of the visual focus based on its coordinates; Based on the normal vector, coordinates, and preset focal length of the visual focus, position parameters and attitude parameters corresponding to each scanning point are generated, so that at each scanning point, whether it is a straight weld or a curved weld, the absolute value of the difference between the lens's focusing distance and the preset focal length is less than the preset first difference threshold.

[0049] The preset first difference threshold is a very small value, that is, the focusing distance is close to or even equal to the preset focal length, so as to avoid image defocusing and blurring from the source and ensure the clarity of weld defect imaging.

[0050] The calculation process of the normal vector of the visual focus is as follows: select multiple consecutive discrete points adjacent to the visual focus to construct a local surface, fit the tangent plane of the local surface by the least squares method, and then solve for the vector perpendicular to the tangent plane. After direction correction, it is the normal vector of the weld surface of the visual focus, ensuring that the direction of the normal vector is accurately perpendicular to the weld surface without angular deviation.

[0051] The process of generating the position and attitude parameters corresponding to the scanning point: Taking the coordinates of the visual focus as the reference, the distance of the preset focal length is offset along the positive direction of the normal vector of that point. The coordinates after the offset are the position parameters of the scanning point of the vision system camera. At the same time, the optical axis direction of the camera lens is adjusted to be completely consistent with the direction of the normal vector of that point. Based on this, the joint angle of the vision robot is calculated, and the corresponding attitude parameters are generated. Through the dual constraints of "position offset and attitude orientation", the focus distance is precisely controlled.

[0052] Optionally, the preset focal length can be selected from the imaging distance with the clearest camera image and the highest defect recognition accuracy among multiple batches of inspections based on statistical analysis of historical chassis weld inspection data.

[0053] The vision system (e.g., a camera) is equipped with a built-in image processing module. The image processing module is used to identify the defect information of the corresponding weld based on the image of each weld and transmit the defect information to the upper device. The types of weld defects include common frame welding defects such as bubbles, weld deviation, incomplete welds, missing welds, and microcracks. The location of the weld defect is the three-dimensional coordinate of the corresponding defect on the weld trajectory, so as to accurately locate the specific location of problems such as bubbles, weld deviation, incomplete welds, and missing welds.

[0054] Specifically, a single weld image is input into a pre-trained defect recognition neural network model. The model extracts and analyzes the texture, contour, and grayscale features of the weld image, and outputs the corresponding weld defect type and location. It should be noted that this application does not limit the specific type and structure of the neural network model; any mature and applicable weld defect recognition neural network model in the prior art can be adapted to the weld defect information recognition task of this application.

[0055] The upper-level device is also used to receive defect information of each weld seam transmitted by the vision system, generate corresponding repair welding control programs based on the defect information of each weld seam, and send them to the repair welding control module.

[0056] The welding repair control module is used to receive the welding repair control logic for each weld from the upper-level device, and drive the welding repair robot to complete the welding repair operation of the frame based on the welding repair control program.

[0057] The upper-level device receives weld defect information transmitted from the vision system, integrates, verifies, and performs coordinate system transformation on the defect information. Based on the defect type, three-dimensional location, size, and severity of each weld, it matches the corresponding frame welding process standard, generates the corresponding repair welding control program, and sends it to the repair welding control module. Specifically, the upper-level device first converts the defect position in the vision detection coordinate system into executable coordinates in the repair welding robot's base coordinate system, and then adapts differentiated repair welding parameters for different defect types: for defects such as bubbles and micropores, it matches precise deposition parameters with low current and short stroke; for defects such as incomplete welds and missing welds, it matches standard welding current and complete weld trajectory parameters; for defects such as weld misalignment, it replans the welding torch travel path and alignment parameters. The final repair welding control program includes the repair welding trajectory, welding torch posture, welding process parameters, start and stop commands, and defect repair welding priority, ensuring that the repair welding operation conforms to the frame structure strength requirements and avoids problems such as over-welding or under-welding.

[0058] The chassis welding repair system provided in this application embodiment, by combining a vision system to identify defect information of welds on the chassis, can avoid the problems of inconsistent standards and low efficiency caused by manual experience judgment. In addition, during the scanning of each weld, the absolute value of the difference between the lens's focusing distance and the preset focal length is less than a preset first difference threshold, which can ensure the clarity of the weld image and the integrity of the defect details, thereby improving the recognition accuracy. Combined with a welding repair robot, it can achieve fully automated welding repair, ensuring both welding repair efficiency and consistent welding quality.

[0059] Based on the welding repair system provided in the above embodiments of this application, this application also provides a method for welding repair of a vehicle frame, such as... Figure 4 As shown, the welding repair method for this vehicle frame includes: S11, drive the vision system to acquire weld seam images at different locations along the scanning path, wherein the scanning path ensures that the absolute value of the difference between the lens's focusing distance and the preset focal length is less than a preset first difference threshold during the scanning process of each weld seam on the vehicle frame.

[0060] S12, based on the images of each weld seam, identify the defect information of the corresponding weld seam.

[0061] S13, Generate the corresponding repair welding control program based on the defect information of each weld.

[0062] S14, the welding robot is driven by the welding repair control program to complete the welding repair work of the frame.

[0063] It should be noted that the specific implementation methods, principles, and beneficial effects of each of the above steps are consistent with the description of the welding repair system in the foregoing embodiments.

[0064] The above are only some embodiments of this application and do not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A chassis welding repair system, characterized in that, include: The upper-level device is used to acquire a three-dimensional model of the entire frame, determine the regional position and shape classification of each weld seam based on the three-dimensional model, and generate a scanning path based on the regional position and shape classification of each weld seam. The scanning path includes robot position parameters and robot posture parameters corresponding to multiple scanning points. The scanning path ensures that the absolute value of the difference between the lens focusing distance and the preset focal length is less than a preset first difference threshold during the scanning of each weld seam by the vision system. The robot position parameters and robot posture parameters are parameters of the robot used to drive the movement of the vision system. A vision inspection and control module is used to control the robot that drives the vision system to perform weld seam scanning based on the scanning path. The vision system is used to acquire weld seam images at different locations along a scanning path under the control of a robot that drives the vision system's movement. Based on each weld seam image, the system identifies the corresponding weld seam defect information and transmits the defect information to the upper-level device. The defect information includes the weld seam defect type and the weld seam defect location. The upper-level device is also used to receive defect information of each weld seam transmitted by the vision system, generate corresponding repair welding control programs based on the defect information of each weld seam, and send them to the repair welding control module. The welding repair control module is used to receive the welding repair control logic for each weld from the upper-level device, and drive the welding repair robot to complete the welding repair operation of the frame based on the welding repair control program.

2. The chassis welding system as described in claim 1, characterized in that, The upper-level device is specifically used for: Extract the region location and point set data of each weld from the three-dimensional model; Based on the coordinates of each point in the point set data of each weld, the curvature of each point is calculated; The morphological classification of each weld is determined based on the curvature of each point.

3. The chassis welding system as described in claim 1, characterized in that, The upper-level device is specifically used for: Multiple center points of each weld area are extracted and connected to form a scanning baseline; Based on the length of the scanning baseline, the scanning baseline is divided into at least one line segment, such that the length of all the resulting line segments is within a preset length range, and the number of the resulting line segments is minimized under the condition of satisfying the length constraint. The intersection of adjacent line segments is used as the visual focus, and the position and attitude parameters of the robot used to drive the movement of the vision system are determined based on the coordinates of each visual focus.

4. The chassis welding system as described in claim 3, characterized in that, The determination of the position and attitude parameters of the robot used to drive the vision system at the scanning point based on the coordinates of each visual focus includes: Calculate the normal vector of the visual focus based on its coordinates; Based on the normal vector, coordinates, and preset focal length of the visual focus, the robot position parameters and robot posture parameters at the corresponding scanning points are generated.

5. The frame repair welding system as described in claim 3, characterized in that, The morphological classification includes straight welds and curved welds.

6. The chassis welding repair system as described in claim 5, characterized in that, The upper-level device is specifically used for: For straight welds, the length of the resulting segment is set to be within a first length range; For the scanning baseline of the curved weld, the scanning baseline is divided into smooth segments and / or non-smooth segments according to the curvature of the points on the scanning baseline. The smooth segments are further segmented according to their lengths, and / or the non-smooth segments are further segmented according to their lengths. The lengths of the segments obtained from the smooth segments are within a second length range, and the lengths of the segments obtained from the non-smooth segments are within a third length range. The first length range has the largest interval, the second length range has the second largest interval, and the third length range has the smallest interval. Furthermore, the maximum value of the first length range is greater than the maximum value of the second length range, which in turn is greater than the maximum value of the third length range.

7. The chassis welding system as described in claim 6, characterized in that, The upper-layer device divides smooth segments and non-smooth segments in the following way: A continuous segment on the scanning baseline of a curved weld where the absolute value of the curvature difference between k consecutive points is less than a preset threshold is defined as a smooth segment. A point is defined as an inflection point when the absolute value of the curvature difference between it and the two points before and after it is greater than or equal to a preset threshold. A continuous segment consisting of an inflection point and z points before and after it is defined as a non-smooth segment, where k and z are both integers greater than or equal to 1.

8. The chassis welding system as described in claim 1, characterized in that, The vision system includes a GRBD camera, or the vision system includes an RGB camera.

9. The frame repair welding system as described in claim 1, characterized in that, The preset focal length is based on statistical analysis of historical chassis weld inspection data, and the imaging distance with the highest weld image defect recognition accuracy among multiple batches of inspections is selected as the preset focal length.

10. A method for repairing a vehicle frame by welding, characterized in that, The welding system applied to the chassis as described in any one of claims 1 to 9, the method comprising: The vision system is driven to acquire weld images at different locations along the scanning path. The scanning path ensures that the absolute value of the difference between the lens's focusing distance and the preset focal length is less than a preset first difference threshold during the scanning process of each weld on the vehicle frame. Defect information of the corresponding weld is identified based on the images of each weld. A corresponding repair welding control program is generated based on the defect information of each weld. The welding robot is driven by a welding control program to complete the welding repair work on the chassis.

Images