A high-efficiency double-wire welding path generation system and path generation method

Abstract

The application discloses a kind of high-efficiency double-wire welding path generation system and path generation method, it is related to welding technical field, including: weld recognition extraction and library module, workpiece is sequentially detected by visual sensing by visual system, obtains image information, extracts weld feature, searches database, calls the highest similarity weld model and relevant welding process data.Weld model knowledge base and weld recognition extraction need to obtain by a large number of visual recognition test, welding process database needs to be obtained by welding test for different groove forms, assembly conditions, weld recognition extraction library, weld intelligent welding trajectory path posture planning, robot welding program automatic generation and simulation issue etc. need to be guaranteed by software algorithm design, and verified and optimized by test, guarantee the welding quality and precision of double-wire welding path, avoid the problem of misplacement and false welding, maximum degree improves the efficiency of welding.

Description

Technical Field

[0001] This invention relates to the field of welding technology, specifically to an efficient dual-wire welding path generation system and method. Background Technology

[0002] Welding, also known as fusion welding, is a manufacturing process and technology that joins metals or other thermoplastic materials such as plastics by heating, high temperature, or high pressure. Welding achieves the purpose of joining through the following three methods: Fusion welding – heating the workpieces to be joined to partially melt them to form a molten pool, which then cools and solidifies to join the parts. Filler material can be added if necessary. It is suitable for welding various metals and alloys and does not require pressure. Pressure welding – the welding process requires applying pressure to the workpieces. It is suitable for various metal materials and some metal materials. Brazing – using a metal material with a lower melting point than the base metal as a filler metal, the liquid filler metal wets the base metal, fills the joint gap, and diffuses with the base metal to achieve the connection of the workpieces. It is suitable for welding various materials and also for welding different metals or dissimilar materials. Modern welding has many energy sources, including gas flame, electric arc, laser, electron beam, friction, and ultrasound.

[0003] The TCP of a traditional single-wire welding torch is a single point, while the TCP of a high-efficiency dual-wire welding torch is a line segment consisting of two points. Fitting a single point to the weld trajectory is relatively simple, but fitting a line segment to the weld trajectory for precise welding is much more difficult. Furthermore, welding requires a certain welding tilt angle, which exponentially increases the difficulty. This makes it impossible to solve the TCP calibration problem of the welding torch and cannot adapt to the multi-position adaptability of the welding torch TCP. Summary of the Invention

[0004] The purpose of this invention is to provide an efficient dual-wire welding path generation system and path generation method to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a high-efficiency dual-wire welding path generation system, comprising:

[0006] The weld seam recognition, extraction, and database access module involves the vision system sequentially performing visual sensing detection on the workpiece, acquiring image information, extracting weld seam features, searching the database, and calling the weld seam model with the highest similarity and related welding process data.

[0007] The weld seam model knowledge base and welding process database include: the weld seam knowledge base and the assembly weld seam functional process database.

[0008] The intelligent welding trajectory path and posture planning module for weld seams, after workpiece recognition and weld seam trajectory extraction, uses the extracted weld seam trajectory as the planning basis to plan the robot welding path and welding torch posture.

[0009] The robot welding program automatic generation and simulation module generates a "robot welding series (or one complete program)". After the robot welding program automatic generation and simulation module generates a "robot welding series (or one complete program)", the system converts the program into an executable "robot running program" and verifies the robot's accessibility and collision avoidance in the background.

[0010] The weld seam identification, extraction, and database adjustment module is connected to the weld seam model knowledge base and the welding process database via signal connections. The weld seam identification, extraction, and database adjustment module is also connected to the weld seam intelligent welding trajectory path posture planning module via signal connections. The welding process database is connected to the weld seam intelligent welding trajectory path posture planning module via bidirectional signal connections. The robot welding program automatic generation and simulation issuance module is also connected to the weld seam intelligent welding trajectory path posture planning module via signal connections.

[0011] Furthermore, the weld seam recognition, extraction, and database retrieval module sequentially performs visual sensing detection on the workpiece based on the vision system to acquire image information, extract weld seam features, search the database, and retrieve the weld seam model with the highest similarity and related welding process data. Simultaneously, it extracts the weld seam contour image and weld seam trajectory, compares and verifies them with the data retrieved from the database, and confirms the weld seam trajectory of the robot welding.

[0012] Furthermore, the weld seam model knowledge base and welding process database also have the functions of importing, creating, storing and managing all weld seam 3D models or weld seam rules, the best welding processes for various weld seams that have been verified through batch production, welding processes that have been designed or verified through experiments but have not been verified through batch production, and data such as assembly weld seam welding design rules. This database module transforms the product model through algorithms and performs parameterized management, serving as the basis for the visual scanning and recognition system to effectively compare and analyze the actual workpiece and workpiece model data in the later stages, and is also the basis for all trajectory, posture and path planning software.

[0013] Furthermore, the intelligent welding trajectory path posture planning module is also used to combine the retrieved welding process, based on the "welding process" level given in the database, and according to the data and model-driven approach, to design the welding process and parameters for each weld, and form a series of robot welding programs.

[0014] Furthermore, the automatic generation and simulation delivery module for robot welding programs is also used to push the "robot running program" to each robot controller to form the robot program to be executed.

[0015] A path generation method for an efficient dual-wire welding path generation system, applied to any one of the above-mentioned efficient dual-wire welding path generation systems, includes the following steps:

[0016] Step 1: First, use the weld identification, extraction, and database adjustment module to determine the path of the weld. Then, interact the extracted data with the weld model knowledge base and the welding process database to obtain the optimal welding process and weld model.

[0017] Step 2: Send the welding process and weld model data to the intelligent welding trajectory path and posture planning module. After workpiece recognition and weld trajectory extraction, use the extracted weld trajectory as the planning basis to plan the robot welding path and welding torch posture.

[0018] Step 3: Based on the data and model-driven approach, design the welding process and parameters for each weld seam, and form a series of robot welding programs.

[0019] Step 4: Send the path data planned by the intelligent welding trajectory path attitude planning module to the robot welding program automatic generation and simulation module for data simulation. After passing the simulation, send the data.

[0020] Step 5: Push the "robot operation program" to each robot controller to form the robot program to be executed, and finally carry out the welding process.

[0021] Furthermore, the following steps are included: According to the operation steps in step one, the weld trajectory obtained by the visual recognition of the weld recognition extraction and library adjustment module is transformed into the motion trajectory of the welding torch. In addition to requiring specialized execution software, it is also necessary to integrate the vision system, the robot's external axis, and the robot's TCP (welding torch end) into a common coordinate system through a specialized calibration method to form a whole.

[0022] Furthermore, the following steps are included: According to the operation steps in step two, the weld seam identification extraction and library adjustment module is based on the hand-eye calibration stereo vision measurement and recognition module, which is the foundation of high-precision three-dimensional measurement. The module mainly includes a high-resolution board camera, a high-speed image acquisition board, an image processing control module, etc. The calibration process is optimized, and a spatial circle fitting method is proposed to perform hand-eye calibration of the standard sphere, so that the light plane can be any plane equation, thereby improving the accuracy of the standard sphere hand-eye calibration and expanding the adaptability range of structural components. For large and complex components, multiple modules are used to reconstruct multiple areas through three-dimensional imaging, and then multi-view stitching is performed to obtain the panoramic shape.

[0023] Furthermore, the following steps are included: according to the operation steps in step four, 180 or 360 attitude data points of the line segment direction formed by the two welding wires of the TCP are compiled and stored in the database. When the intelligent welding trajectory path attitude planning module of the weld is executed, the optimal welding gun attitude is automatically matched according to the algorithm to generate the welding trajectory.

[0024] Furthermore, the process includes the following steps: according to the operation steps in step five, the welding process employs an adaptive laser sensor to perform high-resolution full-frame image processing and weld seam tracking.

[0025] This invention provides a highly efficient dual-wire welding path generation system and method. It offers the following advantages:

[0026] This efficient dual-wire welding path generation system and its usage method require extensive visual recognition testing to obtain the weld model knowledge base and weld recognition extraction. The welding process database needs to be obtained through welding experiments for different bevel forms and assembly conditions. The weld recognition extraction database adjustment, intelligent welding trajectory path posture planning, automatic generation of robot welding programs, and simulation issuance all need to be guaranteed through software algorithm design and verified and optimized through experiments. This ensures the welding quality and accuracy of the dual-wire welding path, avoids misalignment and incomplete welding problems, and maximizes welding efficiency. Attached Figure Description

[0027] Figure 1 This is a general system diagram of the efficient dual-wire welding path generation system and path generation method of the present invention;

[0028] Figure 2 This is a schematic diagram of a high-efficiency dual-wire welding path generation system and path generation method according to the present invention.

[0029] Figure 1 The module includes: 1. Weld seam identification, extraction, and database adjustment module; 2. Weld seam model knowledge base and welding process database; 3. Intelligent welding trajectory path and posture planning module; 4. Automatic generation and simulation issuance module for robot welding programs. Detailed Implementation

[0030] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0031] Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the invention, and should not be construed as limiting the invention.

[0032] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0033] Please see Figure 1-2 This invention provides a technical solution: a high-efficiency dual-wire welding path generation system, comprising:

[0034] Module 1 for weld seam recognition, extraction and database access: The vision system sequentially performs visual sensing detection on the workpiece, acquires image information, extracts weld seam features, searches the database, and calls the weld seam model with the highest similarity and related welding process data.

[0035] Weld Model Knowledge Base and Welding Process Database 2, which includes: a weld knowledge base and an assembly weld functional process database.

[0036] The intelligent welding trajectory path and posture planning module 3 for weld seams, after workpiece recognition and weld seam trajectory extraction, uses the extracted weld seam trajectory as the planning basis to plan the robot welding path and welding torch posture.

[0037] The robot welding program automatic generation and simulation release module 4, after forming a "robot welding series or a complete program", the system performs program conversion to form a "robot running program" that the robot can execute, and verifies the robot's accessibility and collision avoidance in the background.

[0038] The weld identification, extraction, and database adjustment module 1 is connected to the weld model knowledge base and welding process database 2 via signal connection. The weld identification, extraction, and database adjustment module 1 is connected to the intelligent welding trajectory path and attitude planning module 3 via signal connection. The weld model knowledge base and welding process database 2 are connected to the intelligent welding trajectory path and attitude planning module 3 via bidirectional signal connection. The robot welding program automatic generation and simulation command module 4 is connected to the intelligent welding trajectory path and attitude planning module 3 via signal connection. New wear-resistant composite steel plates (whole or partial) are laid inside the compartment. Two types of welds will be formed between the steel plates. One type is the butt weld with a square bevel (two steel plates are arranged adjacently with a small gap reserved as a weld). This type of weld needs to be fully welded. The other type is the fillet weld, which is the fillet weld formed by a single steel plate and the bottom plate. This type of weld needs to be welded. Both types of welds require multi-layer, multi-pass welding, and the bottom of the container has an irregular curvature with varying heights. This addresses the challenges of weld bevel conditions, such as: large and complex dimensions; complex measurement environment with significant background interference, making it difficult to distinguish between background objects and the target in the captured image; and uncontrollable lighting factors, resulting in no obvious grayscale gradient changes at the target boundary in the image. These issues lead to difficulties in visual recognition and laser tracking.

[0039] Specifically, the weld seam recognition, extraction, and database retrieval module 1 sequentially performs visual sensing detection on the workpiece based on the vision system, acquires image information, extracts weld seam features, searches the database, and retrieves the weld seam model with the highest similarity and related welding process data. Simultaneously, it extracts the weld seam contour image and weld seam trajectory, compares and verifies them with the data retrieved from the database, and confirms the weld seam trajectory of the robot welding.

[0040] Specifically, the weld model knowledge base and welding process database 2 also have the functions of importing, creating, storing and managing all 3D weld models or weld rules, the best welding processes for various welds that have been verified through batch production, welding processes that have been designed or verified through experiments but have not been verified through batch production, and data such as assembly weld design rules; this database module transforms product models through algorithms and performs parameterized management, which serves as the basis for the visual scanning and recognition system to effectively compare and analyze the actual workpiece and workpiece model data in the later stage, and is also the basis for all trajectory, attitude and path planning software.

[0041] Specifically, the intelligent welding trajectory path posture planning module 3 is also used to combine the retrieved welding process, based on the "welding process" level given in the database, and according to the data and model-driven approach, to design the welding process and parameters for each weld, and form a series of robot welding programs.

[0042] Specifically, the automatic generation and simulation delivery module 4 for robot welding programs is also used to push the "robot running program" to each robot controller to form the robot program to be executed.

[0043] A path generation method for an efficient dual-wire welding path generation system, applicable to any of the above-mentioned efficient dual-wire welding path generation systems, includes the following steps:

[0044] Step 1: First, use the weld identification, extraction and database adjustment module 1 to determine the path of the weld. Then, interact the extracted data with the weld model knowledge base and the welding process database 2 to obtain the best welding process and weld model.

[0045] Step 2: Send the welding process and weld model data to the intelligent welding trajectory path and posture planning module 3. After workpiece recognition and weld trajectory extraction, use the extracted weld trajectory as the planning basis to plan the robot welding path and welding torch posture.

[0046] Step 3: Based on the data and model-driven approach, design the welding process and parameters for each weld seam, and form a series of robot welding programs;

[0047] Step 4: Send the path data planned by the intelligent welding trajectory path attitude planning module 3 to the robot welding program automatic generation and simulation module 4 for data simulation. After passing the simulation, send the data.

[0048] Step 5: Push the "robot operation program" to each robot controller to form the robot program to be executed, and finally carry out the welding process.

[0049] Specifically, according to the operation steps in step one, the weld trajectory obtained by the visual recognition of the weld recognition extraction and library adjustment module 1 is transformed into the motion trajectory of the welding gun. In addition to requiring specialized execution software, it is also necessary to integrate the vision system, the robot's external axis, and the robot's TCP welding gun end into a common coordinate system through a specialized calibration method to form a whole.

[0050] Specifically, according to the operation steps in step two, the weld seam identification, extraction and library adjustment module 1, based on hand-eye calibration, is the foundation of high-precision 3D measurement. The module mainly includes a high-resolution board camera, a high-speed image acquisition board, an image processing control module, etc. The calibration process is optimized, and a spatial circle fitting method is proposed for hand-eye calibration of the standard sphere, so that the light plane can be any plane equation, thereby improving the accuracy of hand-eye calibration of the standard sphere and expanding the adaptability of structural components. For large and complex components, multiple modules are used to reconstruct multiple areas through 3D imaging, and then multi-view stitching can be performed to obtain the panoramic shape.

[0051] Specifically, according to the operation steps in step four, the direction of the line segment formed by the two welding wires of TCP is compiled into 180 or 360 attitude data and stored in the database. When the intelligent welding trajectory path attitude planning module 3 of the weld is executed, the optimal welding gun attitude is automatically matched according to the algorithm to generate the welding trajectory.

[0052] Specifically, according to the operation steps in step five, the welding process uses an adaptive laser sensor to process high-resolution full-frame images and track weld seams. The SLPr tracking system uses the latest smart sensor, which integrates a megapixel high-resolution camera, unique automatic laser control, and a special optical system to ensure the high quality and stability of the generated weld seam laser stripe images. The special digital image processing software and hardware inside the sensor ensure high-speed image processing and weld seam tracking.

[0053] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept of the present invention, and these all fall within the protection scope of the present invention.

Claims

1. A high-efficiency dual-wire welding path generation system, characterized in that, include: Weld seam recognition, extraction and database access module (1): The vision system sequentially performs visual sensing detection on the workpiece, acquires image information, extracts weld seam features, searches the database, and calls the weld seam model with the highest similarity and related welding process data. Welding model knowledge base and welding process database (2), which includes: welding knowledge base and assembly welding function process database; The intelligent welding trajectory path and posture planning module (3) plans the robot welding path and welding gun posture based on the extracted weld trajectory after workpiece recognition and weld trajectory extraction. The robot welding program automatic generation and simulation release module (4) forms a "robot welding series program". After the robot welding program automatic generation and simulation release module (4) forms a "robot welding series program", the system performs program conversion to form a "robot running program" that the robot can execute, and performs robot operation reachability and anti-collision verification in the background. The weld seam identification, extraction and database adjustment module (1) is connected to the weld seam model knowledge base and the welding process database (2) by signal connection. The weld seam identification, extraction and database adjustment module (1) is connected to the weld seam intelligent welding trajectory path posture planning module (3) by signal connection. The welding process database (2) is connected to the weld seam intelligent welding trajectory path posture planning module (3) by bidirectional signal connection. The robot welding program automatic generation and simulation issuance module (4) is connected to the weld seam intelligent welding trajectory path posture planning module (3) by signal connection. The weld seam recognition, extraction and database access module (1) sequentially performs visual sensing detection on the workpiece based on the vision system, obtains image information, extracts weld seam features, searches the database, calls the weld seam model with the highest similarity and related welding process data, and at the same time extracts the weld seam contour image and weld seam trajectory, compares and verifies with the data called in the database, and confirms the weld seam trajectory of the robot welding. The intelligent welding trajectory path attitude planning module (3) is also used to combine the retrieved welding process, according to the "welding process" level given by the database, and according to the data and model driving method, to design the welding process and parameters of each weld and form a "robot welding series program". The automatic generation and simulation delivery module (4) of the robot welding program is also used to push the "robot running program" to each robot controller to form the robot program to be executed. The direction of the line segment formed by the two welding wires at the end of the welding gun is compiled into 180 or 360 posture data and stored in the database. When the intelligent welding trajectory path posture planning module (3) is executed, the optimal welding gun posture is automatically matched according to the algorithm to generate the welding trajectory.

2. The high-efficiency dual-wire welding path generation system according to claim 1, characterized in that: The weld model knowledge base and welding process database (2) also have the functions of importing, creating, storing and managing all weld three-dimensional models or weld rules, the best welding process of various welds that have been verified through batch production, welding processes that have been designed or verified through experiments but have not been verified through batch production, and assembly weld welding design rule data; this database module transforms the product model through algorithms and performs parameterized management, which serves as the basis for the visual scanning recognition system to effectively compare and analyze the actual workpiece and workpiece model data in the later stage, and is also the basis of all trajectory, attitude and path planning software.

3. A path generation method for an efficient dual-wire welding path generation system, characterized in that, The system for generating an efficient dual-wire welding path according to claim 1 or 2 includes the following steps: S1. First, the weld identification, extraction and database adjustment module (1) is used to determine the path of the weld. Then, the extracted data is interacted with the weld model knowledge base and the welding process database (2) to obtain the best welding process and weld model. S2. Send the welding process and weld model data to the weld intelligent welding trajectory path posture planning module (3). After workpiece recognition and weld trajectory extraction, use the extracted weld trajectory as the planning basis to plan the robot welding path and welding gun posture. S3. Based on data and model-driven methods, design the welding process and parameters for each weld seam, and form a "robot welding series program"; S4. Send the path data planned by the intelligent welding trajectory path attitude planning module (3) to the robot welding program automatic generation and simulation release module (4) for data simulation. After passing the simulation, send the data. S5. Push the "robot operation program" to each robot controller to form the robot program to be executed, and finally carry out the welding process.

4. The path generation method of the high-efficiency dual-wire welding path generation system according to claim 3, characterized in that, Includes the following steps: According to the operation steps in S1, the weld trajectory obtained by the visual recognition of the weld recognition extraction and library adjustment module (1) becomes the motion trajectory of the welding gun. In addition to special execution software, it is also necessary to integrate the vision system, the robot's external axis, and the welding gun end into a common coordinate system through a special calibration method to form a whole.

5. The path generation method of the high-efficiency twin-wire welding path generation system according to claim 3, characterized in that, The following steps are included: According to the operation steps in S2, the weld seam identification extraction and library adjustment module (1) is based on the stereo vision measurement and recognition module of hand-eye calibration, which is the foundation of high-precision three-dimensional measurement. The module mainly includes a high-resolution board camera, a high-speed image acquisition board, and an image processing control module. The calibration process is optimized, and a spatial circle fitting method is proposed to perform hand-eye calibration of the standard sphere, so that the light plane can be any plane equation, improve the accuracy of hand-eye calibration of the standard sphere, expand the adaptability range of structural components, and for large and complex components, multi-module three-dimensional imaging is coordinated to perform multi-area reconstruction, and then multi-view stitching can be performed to obtain the panoramic shape.

6. The path generation method of the high-efficiency twin-wire welding path generation system according to claim 3, characterized in that, The process includes the following steps: According to the operation steps in S5, the welding process uses an adaptive laser sensor for high-resolution full-frame image processing and weld seam tracking.

Images