Intelligent welding system and method based on three-dimensional space visual recognition and positioning

Abstract

The present invention belongs to the technical field of intelligent welding, and provides an intelligent welding system and method based on stereoscopic spatial visual recognition and positioning, which are suitable for welding a workpiece to be welded, wherein the workpiece to be welded has a pipe hole. The intelligent welding system based on stereoscopic spatial visual recognition and positioning comprises: a welding robot system, a 2D vision module, a welding robot system, a 2D vision module, which is suitable for obtaining and sending pipe hole image information of the pipe hole; a 3D vision module, which is suitable for obtaining and sending spatial posture information of the workpiece to be welded; a control system, a welding robot system, a 2D vision module and a 3D vision module, which are respectively connected to communication, the control system performs posture correction on the welding robot system according to the spatial posture information sent by the 3D vision module, and the control system generates a welding path trajectory according to the pipe hole image information sent by the 2D vision module, and controls the welding robot to perform welding operations according to the welding path trajectory.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent welding, and in particular relates to an intelligent welding system and method based on three-dimensional space visual recognition and positioning. Background Art

[0002] The tube sheet welding process is widely used in tube sheet energy heat exchange equipment in the petroleum, chemical, power station and other industries. The tube sheet heat exchanger is a pressure-resistant circulation heat conduction device that is welded together by a number of wound tubes and flange sockets. The thin-walled tube is only 0.8mm~1.2mm, the diameter is between 8mm~12mm, and the arrangement is diversified and the density is high. Therefore, the welding workload and welding difficulty are very large. In addition, due to the high pressure during use, the welding quality requirements are very high, and leakage welding is not allowed.

[0003] In the relevant technology, semi-automatic and specialized welding equipment such as welding operation racks and automatic welding carriages have been used in tube-to-sheet welding production, but the use and installation process of these devices are complicated and the degree of automation is not high. Welding operations still need to be completed by experienced professional technicians, and the welding quality of welders with different proficiency varies greatly, resulting in reduced work efficiency and waste of energy and raw materials. Summary of the invention

[0004] In view of this, the present invention provides an intelligent welding system based on three-dimensional space visual recognition and positioning to solve the problem of low efficiency of manual welding in the prior art.

[0005] The present invention provides an intelligent welding system based on stereoscopic spatial visual recognition and positioning, which is suitable for welding a workpiece to be welded, wherein the workpiece to be welded has a tube hole. The intelligent welding system based on stereoscopic spatial visual recognition and positioning comprises: a welding robot system; a 2D vision module, which is arranged on the welding robot system, and the 2D vision module is suitable for obtaining and sending tube hole image information of the tube hole; a 3D vision module, which is suitable for obtaining and sending spatial posture information of the workpiece to be welded; a control system, which is respectively connected to the welding robot system, the 2D vision module and the 3D vision module for communication, the control system performs posture correction on the welding robot system according to the spatial posture information sent by the 3D vision module, and the control system generates a welding path trajectory according to the tube hole image information sent by the 2D vision module, and controls the welding robot system to perform welding operations according to the welding path trajectory.

[0006] As an optional implementation, the 2D vision module includes a 2D camera, and the 3D vision module includes a 3D camera.

[0007] As an optional embodiment, the control system includes: a human-computer interaction device, suitable for inputting workpiece parameters and process parameters; an image information processor, which is communicatively connected to the human-computer interaction device, the 2D vision module and the 3D vision module respectively, and the image information processor generates the welding path trajectory based on the tube hole image information sent by the 2D vision module.

[0008] The present invention also provides an intelligent welding method based on three-dimensional space vision recognition and positioning, which adopts any one of the intelligent welding systems based on three-dimensional space vision recognition and positioning. The intelligent welding method based on three-dimensional space vision recognition and positioning includes: obtaining spatial posture information through a 3D vision module; controlling the welding robot system to drive the 2D vision module to perform posture correction according to the spatial posture information; obtaining tube hole image information through the 2D vision module; generating a welding path trajectory according to the tube hole image information; and controlling the welding robot system to perform welding operations on the workpiece to be welded according to the welding path trajectory.

[0009] As an optional implementation, the obtaining of spatial posture information through a 3D vision module specifically includes: obtaining three-dimensional image information of the workpiece to be welded; generating three-dimensional point cloud data based on the three-dimensional image information to obtain position information of the center of the workpiece to be welded in the world coordinate system and deflection angle information of the welding plane of the workpiece to be welded relative to the plane of the 3D vision module; and generating the spatial posture information based on the position information and the deflection angle information.

[0010] As an optional implementation, the obtaining of pore image information through a 2D vision module specifically includes: obtaining a pore image of the pore; identifying pore contour feature information of the pore based on the pore image; generating a pore center coordinate sequence based on the contour feature information; and generating the pore image information based on the pore contour feature information and the pore center coordinate sequence.

[0011] As an optional implementation, generating a welding path trajectory according to the tube hole image information specifically includes: calculating and analyzing the number of tube holes and the contour radius according to the series of tube hole images taken; planning a welding path according to the series of tube hole center coordinates and the number of tube holes; planning a welding trajectory according to the contour radius and the tube hole center coordinates; wherein the welding path trajectory includes the welding path and the welding trajectory.

[0012] As an optional implementation manner, the method of calculating and analyzing the number of tube holes and the contour radius based on the series of tube hole images taken specifically includes: obtaining sub-pixel points of the tube hole contour feature information to obtain the tube hole contour frame; taking the center of the camera field of view (x0, y0) as the coordinate origin to obtain the center coordinate value (x1, y1) of the tube hole contour; taking (x1, y1) as the four circumferential edge coordinate values ​​on the coordinate axis of the relative coordinate origin of the circle, the circumferential coordinates on the x-axis are marked as (x+, y1), (x-, y1), and the circumferential coordinates on the y-axis are marked as (x1, y-), (x1, y+), performing difference operation on the four circumferential coordinate values ​​and the center coordinate value (x1, y1) of the tube hole contour to obtain the contour radius by taking the average; processing each tube hole image to count the number of the tube holes to be welded in the image field of view.

[0013] As an optional implementation, after obtaining the contour radius, it also includes: subtracting the calculated contour radius from the pre-input pipe diameter to obtain a difference; if the difference is greater than a preset threshold, it is eliminated; if the difference is within the threshold range, the corresponding pipe hole is judged to be a pipe hole to be welded.

[0014] As an optional implementation, the intelligent welding method based on three-dimensional space visual recognition and positioning also includes: obtaining the outer contour size, tube hole diameter and camera field of view parameters of the workpiece to be welded; and planning the shooting route trajectory of the 2D vision module according to the outer contour size, tube hole diameter and camera field of view parameters.

[0015] The beneficial effects of the present invention are:

[0016] The 3D vision module is used to obtain the spatial posture information of the parts to be welded, and guide the welding robot system to drive the 2D vision module to realize the correction of the spatial posture. The 2D vision module is used to identify the hole image information of the hole to be welded. The hole image information is the plane contour information of the hole. The 3D vision module can automatically identify the spatial posture of the tube plate, communicate with the welding robot system in real time, correct the shooting plane of the 2D vision module, and automatically generate welding paths and trajectories, which greatly improves the welding efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a block diagram of the composition of an intelligent welding system based on three-dimensional space visual recognition and positioning according to an embodiment of the present invention;

[0019] Figure 2 A block diagram of a control system of an intelligent welding system based on three-dimensional space visual recognition and positioning according to an embodiment of the present invention;

[0020] Figure 3 This is a flow chart of an intelligent welding method based on three-dimensional space visual recognition and positioning according to an embodiment of the present invention;

[0021] Figure 4 It is a flow chart of an intelligent welding method based on three-dimensional space visual recognition and positioning according to another embodiment of the present invention;

[0022] Figure 5 It is a schematic diagram of contour radius calculation of an intelligent welding method based on three-dimensional space visual recognition and positioning according to an embodiment of the present invention.

[0023] Description of reference numerals:

[0024] 110. Welding robot system; 120. 2D vision module; 130. 3D vision module; 140. Control system; 141. Human-computer interaction device; 142. Image information processor; 143. Verification submodule; 144. Welding path trajectory planning module; 145. Data sending module. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0026] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0027] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0028] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0029] Combine the following Figures 1 to 5 , describing an embodiment of the present invention.

[0030] According to an embodiment of the present invention, Figure 1 and Figure 2 As shown, an intelligent welding system based on three-dimensional spatial visual recognition and positioning is provided, which is suitable for welding a workpiece to be welded, and there are pipe holes on the workpiece to be welded. The intelligent welding system based on three-dimensional spatial visual recognition and positioning includes a welding robot system 110, a 2D vision module 120, a 3D vision module 130 and a control system 140. The 2D vision module 120 is arranged on the welding robot system 110, and the 2D vision module 120 is suitable for obtaining and sending the pipe hole image information of the pipe hole, and the 3D vision module 130 is suitable for obtaining and sending the spatial posture information of the workpiece to be welded. The control system 140 is communicatively connected with the welding robot system 110, the 2D vision module 120 and the 3D vision module 130 respectively, and the control system 140 performs posture correction on the welding robot system 110 according to the spatial posture information sent by the 3D vision module 130, and the control system 140 generates a welding path trajectory according to the pipe hole image information sent by the 2D vision module 120, and controls the welding robot system 110 to perform welding operations according to the welding path trajectory.

[0031] In this embodiment, the workpiece to be welded is a tube sheet. The 2D vision module 120 is a two-dimensional camera module, and the 3D vision module 130 is a three-dimensional camera module. The welding robot system 110 includes a welding robot and a welding robot control module.

[0032] The 3D vision module 130 is used to obtain the spatial posture information of the workpiece to be welded, guide the welding robot to drive the 2D vision module 120 to realize the correction of the spatial posture, and the 2D vision module 120 is used to identify the tube hole image information of the tube hole to be welded. The tube hole image information is the plane contour information of the tube hole. The 3D vision module 130 can automatically identify the spatial posture of the tube plate, communicate with the welding robot control module of the welding robot system in real time, correct the shooting plane of the 2D vision module 120, and automatically generate welding paths and trajectories, which greatly improves the welding production efficiency.

[0033] The 2D vision module 120 can automatically plan the shooting path according to the field of view of the 2D vision module 120, realize the accurate acquisition of the image of the tube hole to be welded, avoid the inaccurate size measurement caused by imaging distortion, and greatly improve the reliability of the welding image contour information. It can solve the problem of low welding position accuracy of high-density holes and large-diameter tube sheets, and the difficulty of fully automatic welding of thin-walled tube sheets, and improve the welding quality and welding stability.

[0034] Optionally, the 2D vision module 120 includes a 2D camera, and the 3D vision module 130 includes a 3D camera.

[0035] In this embodiment, the 2D vision module 120 includes a 2D camera, an image recognition unit, a coordinate information generating unit, and a data sending unit. The 2D camera is arranged on the end effector of the welding robot, and automatically corrects the position and posture as the welding robot moves. It is used to capture and obtain the plane image information of the pipe holes entering the same photographing area, and send the image information to the image recognition unit. The image recognition unit uses a preset template image to match and identify the image information, and uses the outer contour of the pipe hole to be welded as a matching feature, in order to achieve precise welding operations.

[0036] In this application, the center point coordinates of the welding tube hole are used as the welding coordinates. There are multiple tube holes to be welded in the photographing area. After contour feature matching and identification, a series of coordinate values ​​relative to the center of the tube plate can be obtained through the coordinate information generation unit, which is the center point coordinate value sequence of the tube hole contour to be welded; the data sending unit is used to send the tube hole contour information and the tube hole contour center point coordinate sequence to the control system 140.

[0037] Since there may be cases of mis-sending or mis-identification, resulting in the re-welding of the welded tube hole, before planning the welding path trajectory, the upper control system also needs to determine whether the difference between the contour radius of the target welding tube hole sent by the 2D vision module 120 and the standard tube hole radius input by the user is within the threshold satisfaction range. If it is not satisfied, it is regarded as a welded tube hole and is removed, or it is recorded in the human-computer interaction interface as a tube hole that does not meet the welding standard for manual processing. Therefore, the method proposed in the present invention can ensure the reliability of welding quality and avoid the occurrence of miswelding and leakage caused by dense holes or excessive tube plate size.

[0038] The control system 140 includes a human-computer interaction device 141 and an image information processor 142. The human-computer interaction device 141 is suitable for inputting workpiece parameters and process parameters. The image information processor 142 is communicatively connected with the human-computer interaction device 141, the 2D vision module 120 and the 3D vision module 130 respectively. The image information processor 142 generates a welding path trajectory based on the tube hole image information sent by the 2D vision module 120.

[0039] In this embodiment, the control system 140 includes a human-computer interaction device 141 , an image information processor 142 , a verification submodule 143 , a welding path trajectory planning module 144 , and a data sending module 145 .

[0040] The human-machine interaction device 141 has a human-machine interaction interface for inputting workpiece parameters, process parameters and known variables, visualizing the tube sheet welding area image, and obtaining contour recognition information results.

[0041] The image information processor 142 is used to calculate and analyze the number and contour radius of the pores based on the series of pore images.

[0042] The verification submodule 143 further eliminates the pipe holes that have been welded or do not meet the welding standards according to the pipe hole contour radius information input by the user to ensure the reliability of the welding accuracy.

[0043] The welding path trajectory planning module 144 plans the welding path according to the pipe hole series coordinates and the number of pipe holes to be welded, plans the welding trajectory according to the contour radius and center coordinates of the pipe holes to be welded, and sends the welding path trajectory to the welding robot through the data sending module 145.

[0044] The intelligent welding robot system 110 includes a regional planning unit, a data processing unit, a control instruction receiving and sending module, and an intelligent welding module.

[0045] The intelligent welding robot is connected to a control cabinet, a 2D vision module 120, a 3D vision module 130 and a host control system, which is referred to as the control system 140. The control system 140 is connected to the control cabinet of the welding robot through a communication cable, and the 2D camera is installed on the end effector of the welding robot. The welding robot can receive the spatial posture information of the tube sheet sent by the 3D vision module 130, and correct the plane posture of the 2D camera in real time through the action of the end effector; after the welding robot drives the 2D camera to the shooting point, it sends a control instruction to the 2D vision module 120 to generate a tube hole contour image; the welding robot receives the welding path trajectory data transmitted by the host control system.

[0046] The area planning unit is used to plan the image shooting path of the tube sheet area, that is, to divide the entire tube sheet into several shooting areas. To ensure that there is no leakage, the intersection area between each area is set according to the size of the tube hole. The robot performs actions in sequence according to the established shooting path. After all the tube holes in the current area are welded, it is determined whether the shooting of all areas on the tube sheet is completed. If not, the robot drives the 2D camera to the next shooting area point. If so, the robot returns to the initial position and waits.

[0047] The data processing unit generates instructions for the pipe hole welding requirements based on the welding trajectory path sent by the control system 140 and the information input by the operator, such as welding speed, arc starting height, arc ending voltage, etc.

[0048] The control instruction transceiver module can send information to other modules or units. For example, due to disturbances in the on-site operating environment, or displacement errors caused by the robot deviating from the center of the tube sheet due to the large diameter of the tube sheet, the change in the angle of the tube sheet workpiece plane relative to the 2D camera plane will affect the recognition effect of the 2D vision module 120, and it is possible that the 2D vision module 120 cannot recognize the tube hole contour image in a certain photographing area. At this time, the 2D vision module 120 can send a command to the intelligent welding robot, requiring the 3D vision module 130 to re-acquire the spatial posture information of the tube sheet and perform secondary precise positioning; when arriving at the shooting point, a photo request is sent to the 2D vision module 120; after the welding instruction is generated, it is transmitted to the intelligent welding module through the control instruction sending module to perform the welding operation.

[0049] According to an embodiment of the present invention, Figure 3 As shown, an intelligent welding method based on three-dimensional space visual recognition and positioning is also provided, using any one of the intelligent welding systems based on three-dimensional space visual recognition and positioning, and the intelligent welding method based on three-dimensional space visual recognition and positioning includes the following steps:

[0050] Step S101: Acquire spatial posture information through the 3D vision module 130;

[0051] Step S103: Controlling the welding robot system 110 to drive the 2D vision module 120 to perform posture correction according to the spatial posture information;

[0052] Step S105: acquiring pore image information through the 2D vision module 120;

[0053] Step S107: generating a welding path trajectory according to the tube hole image information;

[0054] Step S109: Control the welding robot system 110 to perform welding operations on the workpiece according to the welding path trajectory.

[0055] The control system 140 calculates the radius of the contour trajectory points of the tube holes and the number of the tube holes, generates a welding path trajectory and sends it to the welding robot.

[0056] Optionally, obtaining the spatial pose information through the 3D vision module 130 specifically includes the following steps:

[0057] Step S201: obtaining three-dimensional image information of the workpiece to be welded;

[0058] Step S203: Generate three-dimensional point cloud data according to the three-dimensional image information to obtain the position information of the center of the workpiece to be welded in the world coordinate system and the deflection angle information of the welding plane of the workpiece to be welded relative to the plane of the 3D vision module;

[0059] Step S205: Generate spatial posture information according to the position information and the deflection angle information.

[0060] The 3D vision module 130 is mainly composed of a 3D camera and a point element analysis unit. The 3D camera captures three-dimensional image information of the tube sheet workpiece, generates three-dimensional point cloud data using the point element analysis unit, obtains the tube sheet contour diameter, and calculates the position information of the tube sheet center in the world coordinate system and the deflection angle information and pitch angle information of the tube sheet plane relative to the 3D camera plane. After coordinate transformation, the spatial posture information of the tube sheet center in the robot coordinate system is obtained, so that the posture of the 2D vision module 120 set on the robot can be adjusted in real time, thereby improving efficiency.

[0061] Among them, the 3D vision module 130 automatically takes pictures to identify the spatial posture and outer contour size of the tube sheet, obtains the pitch angle of the tube sheet plane around the X-axis, the deflection angle information around the Y-axis and the position coordinates of the center of the tube sheet in the world coordinate system, and the above data is transmitted to the welding robot system in real time. The welding robot drives the 2D camera to correct the posture, so that the 2D camera plane is parallel to the tube sheet plane, and the 2D camera center coincides with the tube sheet center.

[0062] Optionally, the 2D vision module 120 is used to obtain the pore image information, which specifically includes the following steps:

[0063] Step S301: acquiring a pore image of a pore;

[0064] Step S303: identifying pore contour feature information of the pores according to the pore image;

[0065] Step S305: generating a pipe hole center coordinate sequence according to the contour feature information;

[0066] Step S307: Generate pore image information according to the pore contour feature information and the pore center coordinate sequence.

[0067] In this embodiment, the welding robot is connected to a control cabinet, and the 2D vision module 120, the 3D vision module 130 and the control system 140 are connected to the control cabinet of the welding robot through communication cables, and the 2D camera is installed on the end effector of the welding robot. The welding robot can receive the spatial posture information of the tube sheet sent by the 3D vision module 130, and correct the plane posture of the 2D camera in real time through the action of the end effector; after the welding robot drives the 2D camera to the shooting point, it sends a control instruction to the 2D vision module 120.

[0068] The 2D vision module 120 is used to obtain the tube sheet plane image and identify the contour features of the tube hole to be welded on the tube sheet plane based on image template matching; the coordinate information generation unit is used to generate a tube hole center coordinate sequence and send the tube hole image information and the tube hole center coordinate sequence to the control system 140 through the data sending module 145.

[0069] The image recognition unit uses a preset template image to match and identify the image information, and uses the outer contour of the tube hole to be welded as the matching feature. In order to achieve accurate welding operations, the present application uses the center point coordinates of the welded tube hole as the welding coordinates. Since there are multiple tube holes to be welded in the photographing area, after the contour features are matched and identified, a series of coordinate values ​​relative to the center of the tube plate can be obtained through the coordinate information generation unit, which is a sequence of center point coordinate values ​​of the contour of the tube hole to be welded; the data sending module 145 is used to send the tube hole contour information and the tube hole contour center point coordinate sequence to the control system 140.

[0070] Generate welding path trajectory according to the tube hole image information, specifically including the following steps:

[0071] Step S401: Calculate and analyze the number of pores and the contour radius based on the series of pore images taken;

[0072] Step S403: planning a welding path according to the center coordinates of the series of tube holes and the number of tube holes;

[0073] Step S405: planning a welding trajectory according to the contour radius and the center coordinates of the tube hole;

[0074] Among them, the welding path trajectory includes the welding path and the welding trajectory.

[0075] like Figure 4 and Figure 5 As shown, the number of pores and the contour radius are calculated and analyzed based on the series of pore images taken, which specifically includes the following steps:

[0076] Step S501: obtaining sub-pixel points of pore contour feature information to obtain the pore contour frame;

[0077] Step S503: Taking the center of the camera field of view (x0, y0) as the coordinate origin, obtain the center coordinate value of the pore outline (x1, y1);

[0078] Step S505: With (x1, y1) as the coordinate values ​​of the four circumferential edges on the coordinate axis relative to the circumferential coordinate origin, the circumferential coordinates on the x-axis are marked as (x+, y1), (x-, y1), and the circumferential coordinates on the y-axis are marked as (x1, y-), (x1, y+), and the four circumferential coordinates and the center coordinates of the pore contour (x1, y1) are subtracted and averaged to obtain the contour radius;

[0079] Step S507: Process each tube hole image and count the number of tube holes to be welded within the image field of view.

[0080] After obtaining the contour radius, the following steps are also included:

[0081] Step S601: subtract the calculated contour radius from the pre-input pipe diameter to obtain a difference;

[0082] Step S603: If the difference is greater than a preset threshold, it is eliminated;

[0083] Step S605: If the difference is within the threshold range, it is determined that the corresponding pipe hole is a pipe hole in a state of being welded.

[0084] Subtract the calculated hole radius r from the pre-input diameter to determine whether it is greater than the threshold; if it is greater than the threshold, the hole is in a welded state and needs to be removed; if it is within the threshold range, the hole is in a state to be welded.

[0085] Repeat the processing for the remaining tube hole contours to count the number of tube holes to be welded within the image field of view and the center coordinates of each tube hole contour.

[0086] The welding robot generates welding instructions and starts the welding module to start welding according to the number of pipe holes to be welded, the pipe hole contour radius and coordinates. After the welding robot completes welding all the pipe holes in the current field of view, it moves to the next photo point according to the 2D camera running track until the 2D camera runs all the photo tracks or completes the welding of all the pipe holes.

[0087] Intelligent welding methods based on three-dimensional space visual recognition and positioning also include:

[0088] Obtain the outer contour dimensions, tube hole diameter and camera field of view parameters of the welded parts;

[0089] The shooting route trajectory of the 2D vision module 120 is planned according to the outer contour size, the tube hole diameter and the camera shooting field of view parameters.

[0090] The welding robot generates welding instructions and starts the welding module to start welding. After the welding robot completes welding all the tube holes in the current field of view, it moves to the next photo point according to the 2D camera running track until the 2D camera runs all the photo tracks or completes the welding of all the tube holes. The 2D camera photo path is automatically planned according to the camera field of view, which realizes the accurate acquisition of the image of the tube hole to be welded, avoids the problem of inaccurate size measurement caused by imaging distortion, and greatly improves the reliability of the welding image contour information.

[0091] Obviously, the above embodiments are merely examples for clear description and are not limitations of the implementation methods.

[0092] For ordinary technicians in the relevant field, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. However, the obvious changes or modifications derived from this are still within the protection scope of the invention.

Claims

1. An intelligent welding method based on three-dimensional space visual recognition and positioning, using an intelligent welding system based on three-dimensional space visual recognition and positioning, suitable for welding a workpiece to be welded, wherein the workpiece to be welded has a pipe hole, characterized in that: The intelligent welding system based on three-dimensional space visual recognition and positioning includes: Welding robot systems (110); A 2D vision module (120) is provided on the welding robot system (110), the 2D vision module (120) is suitable for acquiring and sending the image information of the tube hole, and comprises an image recognition unit, the image recognition unit uses a preset template image to match and recognize the image information, and uses the outer contour of the tube hole to be welded as a matching feature; A 3D vision module (130), adapted to obtain and send spatial posture information of the workpiece to be welded; a control system (140) which is communicatively connected to the welding robot system (110), the 2D vision module (120) and the 3D vision module (130), respectively; the control system (140) performs posture correction on the welding robot system (110) according to the spatial posture information sent by the 3D vision module (130); and the control system (140) generates a welding path trajectory according to the tube hole image information sent by the 2D vision module (120), and controls the welding robot system (110) to perform welding operations according to the welding path trajectory; The intelligent welding method based on three-dimensional space visual recognition and positioning includes: Acquiring spatial posture information through a 3D vision module (130); Controlling the welding robot system (110) to drive the 2D vision module (120) to perform posture correction according to the spatial posture information; Acquiring pore image information through the 2D vision module (120); Generate a welding path trajectory according to the tube hole image information; Controlling the welding robot system (110) to perform welding operations on the workpiece to be welded according to the welding path trajectory; The obtaining of spatial posture information through the 3D vision module (130) specifically includes: Acquiring three-dimensional image information of the workpiece to be welded; Generate three-dimensional point cloud data according to the three-dimensional image information to obtain the position information of the center of the workpiece to be welded in the world coordinate system and the deflection angle information of the welding plane of the workpiece to be welded relative to the plane of the 3D vision module; Generate the spatial posture information according to the position information and the deflection angle information; Generating a welding path trajectory according to the tube hole image information specifically includes: Calculate and analyze the number of pores and the contour radius based on the series of pore images taken; Planning a welding path according to the center coordinates of the series of tube holes and the number of tube holes; Planning a welding trajectory according to the contour radius and the center coordinates of the tube hole; Wherein, the welding path trajectory includes the welding path and the welding trajectory; The number of pores and the contour radius are calculated and analyzed based on the series of pore images taken, including: Obtain the sub-pixel points of the pore outline feature information to obtain the pore outline frame; Taking the center of the camera field of view (x0, y0) as the coordinate origin, the coordinate value (x1, y1) of the center of the pore contour is obtained; taking (x1, y1) as the coordinate values ​​of the four outer circumferential edges on the coordinate axis relative to the circumferential coordinate origin, the outer circumferential coordinates on the x-axis are marked as (x+, y1), (x-, y1), and the outer circumferential coordinates on the y-axis are marked as (x1, y-), (x1, y+), and the four outer circumferential coordinates and the coordinate value (x1, y1) of the center of the pore contour are subtracted and averaged to obtain the contour radius; Processing each tube hole image, and counting the number of the tube holes to be welded within the image field of view; After obtaining the contour radius, the method further includes: Subtracting the calculated contour radius from the pre-input pipe diameter to obtain a difference; If the difference is greater than the preset threshold, the pipe hole is considered to be welded and is removed; If the difference is within the threshold range, the corresponding tube hole is judged to be a tube hole in a state of being welded; The intelligent welding method based on three-dimensional space visual recognition and positioning also includes: Obtain the outer contour dimensions, tube hole diameter, and camera field of view parameters of the welded parts; The photographing route trajectory of the 2D vision module (120) is planned according to the outer contour size, the tube hole diameter, and the camera photographing field of view parameters.

2. The intelligent welding method based on three-dimensional space visual recognition and positioning according to claim 1 is characterized in that: The 2D vision module (120) includes a 2D camera, and the 3D vision module (130) includes a 3D camera.

3. The intelligent welding method based on three-dimensional space visual recognition and positioning according to claim 1 or 2, characterized in that: The control system (140) comprises: A human-machine interaction device (141) adapted to input workpiece parameters and process parameters; An image information processor (142) is communicatively connected to the human-machine interaction device (141), the 2D vision module (120), and the 3D vision module (130), respectively, and the image information processor (142) generates the welding path trajectory according to the tube hole image information sent by the 2D vision module (120).

4. The intelligent welding method based on three-dimensional space visual recognition and positioning according to claim 1 is characterized in that: Acquiring pore image information through the 2D vision module (120) specifically includes: Acquiring a pore image of the pore; Identifying pore contour feature information of the pore according to the pore image; Generate a sequence of pipe hole center coordinates according to the contour feature information; The pore image information is generated according to the pore contour feature information and the pore center coordinate sequence.

Images