Intelligent welding system and method based on binocular vision and line scanning laser

Abstract

The invention relates to the technical field of intelligent welding, in particular to an intelligent welding system and method based on binocular vision and line scanning laser. The intelligent welding system based on binocular vision and line scanning laser comprises a welding management assembly, a welding monitoring assembly, a welding execution assembly and a welding seam tracking assembly. The welding management assembly comprises a scanning unit and a processing unit. The scanning unit forms welding part contour information and welding groove shape information; and the processing unit edits a welding process characteristic curve and plans multiple layers of welding paths. And the welding monitoring assembly generates welding process real-time correction information. And the execution assembly controls the welding execution equipment to perform welding operation. The intelligent welding system has the advantages that the welding process is monitored in real time, welding parameters are adjusted in time, the positioning precision of a workpiece is improved, the multi-layer and multi-pass welding requirement is met, a multi-layer and multi-pass planning algorithm is provided, process parameters in the welding process can be corrected in real time, and the welding quality is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent welding, and particularly to an intelligent welding system and a welding method based on binocular vision and line-scanning laser. Background Art

[0002] With the development of intelligent device technology, intelligent welding systems have been widely applied in the welding field, improving the processing efficiency, enhancing the welding quality, reducing the labor cost, and decreasing the occurrence of personal injury accidents. Generally, automated intelligent welding equipment is equipped with a camera to collect the image information of the operation site and the welding workpiece, facilitating the positioning of the welding equipment. Through the analysis of the images and in combination with process parameters, the welding equipment is controlled to operate and perform filling welding operations on the groove.

[0003] Existing intelligent welding equipment usually uses a monocular image acquisition device to collect images of the operation site. During the welding process, it is difficult to monitor the welding environment and welding accuracy, and it is hard to ensure the welding quality. For example, the Chinese invention patent with the publication number CN118720571A provides an intelligent welding system and method based on three-dimensional space vision recognition and positioning, suitable for welding a workpiece to be welded with pipe holes. The intelligent welding system based on three-dimensional space vision recognition and positioning includes: a welding robot system, a 2D vision module, a welding robot system, a 2D vision module suitable for obtaining the pipe hole image information of the pipe holes and sending it; a 3D vision module suitable for obtaining the spatial pose information of the workpiece to be welded and sending it; a control system. The welding robot system, the 2D vision module, and the 3D vision module are respectively communicatively connected. The control system corrects the pose of the welding robot system according to the spatial pose 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. For the welding system of this patent, since a binocular image acquisition device is not used, the positioning accuracy of the workpiece is relatively low, it is difficult to meet the requirements of multi-layer and multi-pass welding of thick plates in the ship welding field, and it does not have the ability of multi-layer and multi-pass welding planning, does not have the pulse MIG welding process parameter implementation adjustment and correction and multi-layer and multi-pass planning algorithm, and cannot perform real-time correction of the process parameters during the welding process. Summary of the Invention

[0004] In view of this, the present invention aims to provide an intelligent welding system based on binocular vision and line-scanning laser. By using a binocular vision unit to collect image information of the working space of the station, the position of the workpiece is accurately located. The line laser scanning unit scans the weld information, plans the line laser scanning path, obtains the point cloud data of the workpiece, extracts the groove information data, and establishes a welding filling strategy according to requirements such as different materials and process parameters. During the welding process, the welding state information is monitored in real time through a molten pool monitoring unit, the interlayer temperature information is collected through a temperature acquisition unit, and the characteristic spectral lines of various elements in the pulsed welding arc are collected through a spectral acquisition unit, realizing real-time monitoring of the welding process, timely adjustment of welding parameters, and solving the problems of low positioning accuracy of the workpiece, difficulty in meeting the requirements of thick plate multi-layer and multi-pass welding in the ship welding field, lack of multi-layer and multi-pass welding planning ability, lack of pulse MIG welding process parameter implementation adjustment and correction and multi-layer and multi-pass planning algorithms, inability to perform real-time correction of process parameters during welding, and poor welding quality.

[0005] To solve the above problems, the present invention provides an intelligent welding system and a welding method based on binocular vision and line-scanning laser, including:

[0006] A welding management assembly that stores welding workpiece information and welding process parameter information and generates welding motion trajectory information. The welding management assembly includes:

[0007] A scanning unit that scans the contour of the welding workpiece, generates point cloud data, and forms welding part contour information and welding groove shape information;

[0008] A processing unit that, based on the received welding part contour information and welding groove shape information, matches welding parameters based on the welding process characteristic curve, edits the welding process characteristic curve, plans the multi-layer and multi-pass welding path, and performs virtual welding simulation;

[0009] A welding monitoring assembly that transmits the received welding quality information to the welding management assembly and generates real-time welding process correction information;

[0010] A welding execution assembly that controls the welding execution device to perform welding operations according to the welding motion trajectory information and welding process parameter information sent by the welding management assembly;

[0011] A weld tracking assembly. The welding management assembly generates motion position and speed real-time correction information according to the real-time welding process correction information sent by the weld tracking assembly, and the weld tracking assembly performs real-time correction on the welding execution device.

[0012] Further, the welding management assembly further includes:

[0013] The binocular vision unit collects images of the welded workpiece, and the processing unit generates three-dimensional point cloud data of the workpiece based on the received images of the welded workpiece to identify the spatial position of the weld seam.

[0014] Furthermore, the processing unit includes:

[0015] The process parameter adaptive module: Based on the welding process characteristic curve database, it automatically matches the pulse MIG welding process characteristic curve and welding parameters according to the weld seam characteristics;

[0016] The characteristic curve editing module: For various welding materials, it corrects the pulse welding waveform and welding process characteristic curve, and conducts targeted optimization for various physical and chemical property welding materials;

[0017] The layer and pass planning module: Combining the weld depth and groove angle, it dynamically plans the multi-layer and multi-pass welding path;

[0018] The simulation module: Through digital twin technology, it simulates the welding process to verify the feasibility of the parameters.

[0019] Furthermore, the welding monitoring assembly includes:

[0020] The molten pool monitoring unit, which collects welding state information in real time. The welding state information includes the molten pool shape, arc stability, and wire position;

[0021] The temperature acquisition unit, which collects interlayer temperature information in real time;

[0022] The spectral acquisition unit, which collects the characteristic spectral lines of various elements in the pulsed welding arc in real time;

[0023] The monitoring and processing unit analyzes the received welding state information and the characteristic spectral lines to obtain real-time welding process parameters and optimize the process library.

[0024] Furthermore, the welding execution assembly includes:

[0025] The control unit receives the welding motion trajectory information and welding process parameter information sent by the welding management assembly and generates welding execution instructions;

[0026] The execution unit receives the welding execution instructions sent by the control unit and performs welding operations.

[0027] Furthermore, the execution unit uses a six-axis welding robot.

[0028] An adaptive intelligent welding method based on binocular vision and line-scanning laser, a method of welding through the adaptive intelligent welding system as described above. The welding method includes:

[0029] S100, collecting work space information;

[0030] S110, collecting the image information of the working space through the binocular vision unit to determine the placement position of the welding workpiece and the position of the welding working area;

[0031] S120, calibrating the start and end points of the scanning unit scanning, and planning the outgoing laser scanning path;

[0032] S130, scanning the welding operation area along the laser scanning path, performing surface reconstruction on the obtained point cloud data, and extracting groove information data;

[0033] S140, match the selected welding materials and parent materials with their materials and specifications, and select welding process characteristic curves in a targeted manner;

[0034] S150, adjusting the required number of layers and passes for the reconstructed welding groove, and establishing a welding filling strategy;

[0035] S200, programming the welding filling strategy information to form an execution unit to execute a numerical control program, and the execution unit performs a welding operation;

[0036] S300, welding quality monitoring, real-time correction of welding parameters;

[0037] S400, storing welding data.

[0038] Further, in step S130, the surface reconstruction method includes: the execution unit carries a line scanning laser sensor, scans the welding operation area along the line laser scanning path, and performs surface reconstruction on the obtained point cloud data by using a Delaunay triangulation method;

[0039] The groove information data includes groove depth, top width and bottom width.

[0040] Furthermore, in step S150, the welding filling strategy includes data such as welding current, voltage, wire feeding speed, etc. required for base welding, filling welding and cap welding, and plans the welding filling path and welding speed.

[0041] Furthermore, in step S300, the method for monitoring welding quality and correcting welding parameters in real time includes:

[0042] The welding monitoring assembly monitors the characteristic spectrum, welding status information and interlayer temperature information in real time. When characteristic signals of welding defects are found, the welding management assembly provides feedback adjustment to the execution unit to correct the welding parameters.

[0043] Compared with the prior art, the intelligent welding system based on binocular vision and line scanning laser described in the present invention has the following advantages:

[0044] The advantages of this technical solution are as follows: The binocular vision unit is used to collect the image information of the working space of the station, accurately locate the position of the workpiece. The line laser scanning unit scans the weld information, plans the line laser scanning path, obtains the point cloud data of the workpiece, and extracts the groove information data. Welding filling strategies are established according to requirements such as different materials and process parameters. During the welding process, the molten pool monitoring unit monitors the welding state information in real time, the temperature acquisition unit collects the interlayer temperature information, and the spectral acquisition unit collects the characteristic spectral lines of various elements in the pulsed welding arc, realizing the real-time monitoring of the welding process, timely adjusting the welding parameters, improving the positioning accuracy of the workpiece, meeting the requirements of thick plate multi-layer and multi-pass welding in the ship welding field, having the ability of multi-layer and multi-pass welding planning, having the pulsed MIG welding process parameter implementation adjustment and correction and multi-layer and multi-pass planning algorithms, being able to correct the process parameters in real time during the welding process, and improving the welding quality. Brief Description of the Drawings

[0045] Figure 1 It is a structural block diagram of the intelligent welding system based on binocular vision and line-scanning laser according to the embodiment of the present invention;

[0046] Figure 2 It is a flowchart of the adaptive intelligent welding method based on binocular vision and line-scanning laser according to the embodiment of the present invention. Detailed Embodiments

[0047] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention will be given in conjunction with the accompanying drawings.

[0048] In the present invention, the descriptions involving "first", "second", "upper", "lower", etc. are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first", "second", "upper", "lower" may explicitly or implicitly include at least one of such features. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the fact that those skilled in the art can implement them. When the technical solutions between the embodiments can be combined, they are all within the protection scope required by the present invention.

[0049] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0050] As Figure 1As shown in the figure, an intelligent welding system based on binocular vision and line-scanning laser includes: a welding management assembly, a welding monitoring assembly, a welding execution assembly, and a weld seam tracking assembly. The welding management assembly stores welding workpiece information and welding process parameter information, and generates welding motion trajectory information. The welding management assembly includes: a scanning unit and a processing unit. The scanning unit scans the contour of the welding workpiece to generate point cloud data, forming welding part contour information and welding groove shape information; the processing unit matches welding parameters based on the received welding part contour information and welding groove shape information, edits the welding process characteristic curve, plans the multi-layer and multi-pass welding path, and performs virtual welding simulation. The welding monitoring assembly transmits the received welding quality information to the welding management assembly to generate real-time welding process correction information. The welding execution assembly controls the welding execution device to perform welding operations according to the welding motion trajectory information and welding process parameter information sent by the welding management assembly. The weld seam tracking assembly, the welding management assembly generates motion position and speed real-time correction information according to the real-time welding process correction information sent by the weld seam tracking assembly, and the weld seam tracking assembly performs real-time correction on the welding execution device.

[0051] The weld seam information is scanned by the line laser scanning unit, the line laser scanning path is planned, the point cloud data of the workpiece is obtained, the groove information data is extracted, and the welding filling strategy is established according to different material and process parameter requirements. During the welding process, the welding state information is monitored in real time by the molten pool monitoring unit, the interlayer temperature information is collected by the temperature acquisition unit, and the characteristic spectral lines of various elements in the pulsed welding arc are collected by the spectral acquisition unit, realizing the real-time monitoring of the welding process, timely adjusting the welding parameters, improving the positioning accuracy of the workpiece, meeting the requirements of multi-layer and multi-pass welding of thick plates in the ship welding field, not having the multi-layer and multi-pass welding planning ability, having the pulsed MIG welding process parameter implementation adjustment and correction and multi-layer and multi-pass planning algorithms, being able to perform real-time correction of the process parameters during the welding process, and improving the welding quality.

[0052] It should be noted that the welding workpiece information and welding process parameter information can be input manually, for example, by inputting the corresponding project content on the management interface, or by inputting through various media. The scanning unit uses an integrated high-precision laser sensor (accuracy ±0.5mm) to scan the weld seam morphology and reconstruct the three-dimensional contour, and measures the features such as gap, misalignment, and groove angle in real time. In the processing unit of the welding management assembly, the binocular vision and line laser scanning data are fused through Kalman filtering to eliminate the reflection interference and improve the positioning accuracy to ±0.3mm.

[0053] Furthermore, the welding management assembly further includes: a binocular vision unit that collects the image of the welding workpiece, and the processing unit generates the workpiece three-dimensional point cloud data according to the received image of the welding workpiece to identify the spatial position of the weld seam.

[0054] The binocular vision unit uses a high-resolution 3D camera (1280×1024 pixels), generates three-dimensional point cloud data of the workpiece based on the feature point matching algorithm, and identifies the spatial position of the weld seam.

[0055] Furthermore, the processing unit includes: a process parameter adaptive module, a feature curve editing module, a layer and pass planning module, and a simulation module. The process parameter adaptive module, based on the welding process feature curve database (supporting materials such as carbon steel and stainless steel), automatically matches the pulse MIG welding process feature curve and welding parameters according to the weld seam features. The welding parameters include parameters such as pulse current, voltage, wire feeding speed, peak current, and peak time. The feature curve editing module corrects the pulse welding waveform and welding process feature curve for various welding materials, and optimizes various physical and chemical property welding materials specifically. The layer and pass planning module: combines the weld seam depth and groove angle to dynamically plan the multi-layer and multi-pass welding path, and supports manual correction. The simulation module simulates the welding process through digital twin technology to verify the feasibility of the parameters.

[0056] Furthermore, the welding monitoring assembly includes: a molten pool monitoring unit, a temperature acquisition unit, a spectrum acquisition unit, and a monitoring and processing unit. The molten pool monitoring unit uses a high-dynamic CMOS camera (frame rate 58fps, HDR≥140dB) to collect welding state information in real time. The welding state information includes molten pool morphology, arc stability, and wire position. The temperature acquisition unit uses an infrared camera (accuracy ±1°C) to collect interlayer temperature information in real time and triggers the threshold to control the welding rhythm. The spectrum acquisition unit collects the characteristic spectral lines of various elements in the pulsed welding arc through an eight-channel fiber optic spectrometer in real time. The monitoring and processing unit records parameters such as current, voltage, and welding speed, analyzes the received welding state information and the characteristic spectral lines, obtains real-time welding process parameters, and optimizes the process library through machine learning.

[0057] Furthermore, the welding execution assembly includes: a control unit and an execution unit. The control unit receives the welding motion trajectory information and welding process parameter information sent by the welding management assembly and generates a welding execution instruction. The execution unit receives the welding execution instruction sent by the control unit and performs the welding operation.

[0058] Furthermore, the execution unit uses a six-axis welding robot.

[0059] The execution unit uses a welding robot and a welding power source. The welding robot is, for example, a Kuka six-axis robot (repeat positioning accuracy ±0.05mm). The welding power source uses a Panasonic digital welding machine, which supports the pulse MAG / MIG process.

[0060] Such as Figure 2As shown, an adaptive intelligent welding method based on binocular vision and line scanning laser is used to perform welding by the adaptive intelligent welding system as described above. The welding method includes:

[0061] S100, collecting work space information;

[0062] S110, collecting the image information of the working space through the binocular vision unit to determine the placement position of the welding workpiece and the position of the welding working area;

[0063] Taking advantage of the 3D camera's large field of view and fast shooting speed, the grayscale and physical depth information recorded by the 3D camera is extracted and visualized to quickly understand the overall situation of the welding operation space, determine the placement of the welding workpiece, the location of the welding operation area, etc.

[0064] S120, calibrating the start and end points of the scanning unit scanning, and planning the outgoing laser scanning path;

[0065] As an implementation method of this step, in the welding management assembly, manually determine the welding operation area and click to mark the start and end points of the laser detailed scan.

[0066] S130, the execution unit carries a line scanning laser sensor, scans the welding operation area along the line laser scanning path, reconstructs the surface of the obtained point cloud data by using the Delaunay triangulation method, and extracts the groove information data;

[0067] Groove information data includes key groove shape data such as groove depth, top width, bottom width, etc., which provide a basis for welding layer planning.

[0068] S140, match the selected welding materials and parent materials with their materials and specifications, and select welding process characteristic curves in a targeted manner;

[0069] S150. Through the weld bead filling-process parameter data model, the required number of layers and passes are adjusted for the reconstructed welding groove, and a welding filling strategy is established, including the welding current, voltage, wire feeding speed and other data required for the base weld, filling weld and cover weld, and the welding filling path and welding speed are planned.

[0070] S200, programming the welding filling strategy information to form a numerical control program for execution of the execution unit, and sending it to the execution unit, and the execution unit performs the welding operation;

[0071] S300, welding quality monitoring, real-time correction of welding parameters, including:

[0072] The welding monitoring assembly monitors the characteristic spectral lines, welding status information, and interlayer temperature information in real time for real-time monitoring. When a welding defect characteristic signal is detected, the welding management assembly provides feedback adjustment to the execution unit to correct the welding parameters.;

[0073] Through the welding monitoring assembly, the spectral data, the data collected by the molten pool camera, and the interlayer temperature data collected by the welding infrared camera are used to monitor the quality information of the welding process in real time. If a welding defect characteristic signal is found, the welding intelligent management system will provide feedback adjustment to the welding robot and the digital welding machine to correct the welding parameters to avoid continuous welding quality problems.

[0074] S400, store welding data.

[0075] Record and store the process parameter information and quality monitoring information of the welding process in the welding traceability database, which is convenient for iteratively optimizing the welding process through big data analysis or algorithms in the future.

[0076] Furthermore, in step S300, the method for welding quality monitoring and real-time correction of welding parameters

[0077] The following uses specific embodiments to illustrate the welding method of the present invention.

[0078] Take the butt welding of a stainless steel test plate (750×750mm) as an example.

[0079] Through the above welding method, the system automatically identifies a weld gap of 1.2 mm, plans 3 layers and 4 passes of welding, matches a current of 280 A and a voltage of 28 V, and monitors the molten pool width fluctuation ≤ 0.5 mm in real time. Finally, the weld formation is uniform and defect-free.

[0080] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention should be subject to the scope defined by the claims.

Claims

1. An intelligent welding system based on binocular vision and line scanning laser, characterized in that: include: The welding management assembly stores welding workpiece information and welding process parameter information and generates welding motion trajectory information. The welding management assembly includes: The scanning unit scans the contour of the welding workpiece, generates point cloud data, and forms the contour information of the welding part and the shape information of the welding groove; The processing unit matches welding parameters based on the welding process characteristic curve through the received welding part contour information and welding groove shape information, edits the welding process characteristic curve, plans multi-layer and multi-pass welding paths, and implements virtual welding simulation; The welding monitoring assembly transmits the received welding quality information to the welding management assembly to generate real-time correction information of the welding process; A welding execution assembly controls the welding execution equipment to perform welding operations according to the welding motion trajectory information and welding process parameter information sent by the welding management assembly; The weld tracking assembly, the welding management assembly generates motion position and speed correction information according to the real-time correction information of the welding process sent by the weld tracking assembly, and the weld tracking assembly performs real-time correction on the welding execution equipment.

2. The intelligent welding system based on binocular vision and line scanning laser according to claim 1 is characterized in that: The welding management assembly also includes: The binocular vision unit collects the image of the welding workpiece, and the processing unit generates three-dimensional point cloud data of the workpiece according to the received image of the welding workpiece to identify the spatial position of the weld.

3. The intelligent welding system based on binocular vision and line scanning laser according to claim 1 is characterized in that: The processing unit comprises: Process parameter adaptive module: Based on the welding process characteristic curve database, it automatically matches the pulse MIG welding process characteristic curve and welding parameters according to the weld characteristics; Characteristic curve editing module: for a variety of welding materials, it can modify the pulse welding waveform and welding process characteristic curve, and optimize various physical and chemical properties of welding materials; Layer planning module: Combines weld depth and groove angle to dynamically plan multi-layer and multi-pass welding paths; Simulation module: simulate the welding process through digital twin technology to verify the feasibility of parameters.

4. The intelligent welding system based on binocular vision and line scanning laser according to claim 1 is characterized in that: The welding monitoring assembly comprises: A molten pool monitoring unit collects welding status information in real time, including molten pool morphology, arc stability and welding wire position; Temperature collection unit, real-time collection of interlayer temperature information; Spectrum acquisition unit, real-time acquisition of characteristic spectra of multiple elements in pulse welding arc; The monitoring processing unit analyzes the received welding state information and the characteristic spectrum to obtain real-time welding process parameters and optimize the process library.

5. The intelligent welding system based on binocular vision and line scanning laser according to claim 1 is characterized in that: The welding execution assembly comprises: A control unit receives the welding motion trajectory information and welding process parameter information sent by the welding management assembly and generates a welding execution instruction; The execution unit receives the welding execution instruction sent by the control unit and performs the welding operation.

6. The intelligent welding system based on binocular vision and line scanning laser according to claim 5 is characterized in that: The execution unit adopts a six-axis welding robot.

7. An adaptive intelligent welding method based on binocular vision and line scanning laser, characterized in that: A method for welding by the adaptive intelligent welding system according to any one of claims 1 to 6, the welding method comprising: S100, collecting work space information; S110, collecting the image information of the working space through the binocular vision unit to determine the placement position of the welding workpiece and the position of the welding working area; S120, calibrating the start and end points of the scanning unit scanning, and planning the outgoing laser scanning path; S130, scanning the welding operation area along the laser scanning path, performing surface reconstruction on the obtained point cloud data, and extracting groove information data; S140, match the selected welding materials and parent materials with their materials and specifications, and select welding process characteristic curves in a targeted manner; S150, adjusting the required number of layers and passes for the reconstructed welding groove, and establishing a welding filling strategy; S200, programming the welding filling strategy information to form an execution unit to execute a numerical control program, and the execution unit performs a welding operation; S300, welding quality monitoring, real-time correction of welding parameters; S400, storing welding data.

8. The adaptive intelligent welding method based on binocular vision and line scanning laser according to claim 7 is characterized in that: Step S130, the surface reconstruction method includes: the execution unit carries a line scanning laser sensor, scans the welding operation area along the line laser scanning path, and performs surface reconstruction on the obtained point cloud data by using a Delaunay triangulation method; The groove information data includes groove depth, top width and bottom width.

9. The adaptive intelligent welding method based on binocular vision and line scanning laser according to claim 7 is characterized in that: In step S150, the welding filling strategy includes: welding current, voltage, wire feeding speed and other data required for base welding, filling welding and cap welding, and plans the welding filling path and welding speed.

10. The adaptive intelligent welding method based on binocular vision and line scanning laser according to claim 7, characterized in that: In step S300, the method for monitoring welding quality and correcting welding parameters in real time includes: The welding monitoring assembly monitors the characteristic spectrum, welding status information and interlayer temperature information in real time. When characteristic signals of welding defects are found, the welding management assembly provides feedback adjustment to the execution unit to correct the welding parameters.

Images