High-altitude double-arm cooperative intelligent welding robot system

Abstract

The invention provides a high-altitude double-arm cooperative intelligent welding robot system. According to the system, by integrating the ladder truck platform, the telescopic arm system, the three-axis rotary clamping mechanism and the double-collaborative robot, full automation, high precision and high adaptability of high-altitude pipeline welding are achieved. The core technical effect is that the problems of high operation risk, unstable quality, strong environmental dependence and the like in the traditional high-altitude welding are solved, and the safety, the welding quality, the efficiency and the intelligent level are improved.

Description

Technical Field

[0001] This invention relates to the field of high-end equipment manufacturing, and in particular to a high-altitude dual-arm collaborative intelligent welding robot system. Background Technology

[0002] Pipeline systems play a crucial role in material transportation in energy, chemical, construction, and municipal industries. Welding, as the most important and reliable method of pipeline connection, directly determines the structural integrity, sealing performance, operational safety, and service life of the entire pipeline system. Ensuring welding quality is of paramount importance in pipeline engineering construction.

[0003] With the continuous development of modern industry, the application of pipelines is becoming increasingly widespread, especially in petrochemical, natural gas, urban heating, and large industrial facilities, where high-altitude pipeline installations have become commonplace. However, high-altitude pipeline welding operations face numerous severe challenges. Their unique locations and extremely limited construction space, coupled with significant interference from natural environmental factors such as wind, temperature, and humidity, not only greatly increase construction difficulty and safety risks but also impose more stringent requirements on key aspects such as heat input control, weld quality, and shielding gas effectiveness during the welding process. Currently, welding technologies and systems applied to high-altitude pipelines have significant limitations, hindering industry development and the improvement of safety levels.

[0004] Traditional welding methods heavily rely on welders using aerial work platforms for manual operations at height, making it difficult to guarantee consistent welding quality and posing significant safety risks to personnel. Existing automated welding equipment is mostly ground-based, lacking spatial mobility and struggling to effectively reach and adapt to complex high-altitude work sites. Even when equipment is transported to high altitudes, common single-arm welding robots often struggle to achieve high-quality welding across all pipe positions, especially at the bottom, resulting in blind spots. Furthermore, existing systems generally lack the ability to identify weld seams and cannot perform real-time path planning and adaptive adjustments based on actual deviations in pipe assembly, easily leading to welding defects. Additionally, high-altitude winds can easily compromise gas shielding effectiveness, and uneven heat input during unilateral welding can cause significant deformation.

[0005] Therefore, the development of a high-altitude dual-arm collaborative intelligent welding robot system is of great significance. Summary of the Invention

[0006] The purpose of this invention is to provide a high-altitude dual-arm collaborative intelligent welding robot system to solve the problems existing in the prior art.

[0007] The technical solution adopted to achieve the purpose of this invention is as follows: a high-altitude dual-arm collaborative intelligent welding robot system, including a welding execution unit, a telescopic arm unit, and a mobile support unit.

[0008] The telescopic boom unit includes welding auxiliary equipment and a telescopic boom structure. The welding auxiliary equipment includes a welding machine, a top electrical control cabinet, an end wire feeder, a gas cylinder for MIG / MAG welding, and an industrial control computer. The industrial control computer is equipped with an intelligent algorithm unit that controls the entire system's workflow. The intelligent algorithm unit includes a weld seam positioning module, a scanning module, and a welding path planning module. The telescopic boom structure includes a platform rotary boom, a first-stage telescopic boom, a second-stage telescopic boom, and a third-stage telescopic boom. The platform rotary boom includes a rotary base and a boom body. The rotary base is rotatably mounted on a mobile support unit using a rotary support mechanism, enabling horizontal rotation. The boom body, welding machine, MIG / MAG welding gas cylinder, and industrial control computer are fixed to the rotary base. A first-stage telescopic boom capable of pitch and rotation is installed at the tail of the boom body. An auxiliary telescopic support rod is provided between the boom body and the first-stage telescopic boom. The base of the auxiliary telescopic support rod is fixed to the boom body, and the telescopic end is fixed to the first-stage telescopic boom. A first-stage telescopic boom cable chain and a top electrical control cabinet are fixed to the first-stage telescopic boom. A secondary telescopic arm is installed at the front end of the primary telescopic arm to enable forward and backward sliding extension. A tertiary telescopic arm is installed at the front end of the secondary telescopic arm to enable forward and backward sliding extension. A tertiary telescopic arm cable chain and an end wire feeder are installed on the tertiary telescopic arm to enable cable following and end-effector welding wire feeding.

[0009] The welding execution unit includes a pipe clamping mechanism, a welding execution mechanism, and a vision recognition mechanism. The pipe clamping mechanism includes a three-axis rotary table assembly, a pipe clamping frame, a left clamping mechanism, and a right clamping mechanism. The three-axis rotary table assembly is connected to the end of a three-stage telescopic arm, providing three degrees of freedom adjustment: roll, pitch, and yaw. The pipe clamping frame is fixed to the end of the three-axis rotary table assembly. The left and right clamping mechanisms are symmetrically installed inside the pipe clamping frame for gripping the pipe. The welding execution mechanism includes a left-arm welding collaborative robot, a right-arm welding collaborative robot, and a robot end-effector welding tool assembly. The left and right arm welding collaborative robots are symmetrically installed on the outside of the pipe clamping frame. The ends of the left and right arm welding collaborative robots are each connected to the robot end-effector welding tool assembly via flanges. The welding tool assembly integrates a vision sensor and a welding torch. The vision recognition mechanism includes a frame observation camera. The frame observation camera is installed on the side of the pipe clamping frame and communicates in real time with an industrial control computer for remote monitoring of high-altitude operations.

[0010] During operation, the mobile support unit travels to the work point, and the vehicle platform's rotary arm rotates horizontally to adjust the direction of the telescopic arm. The first-stage telescopic arm pitches and swings, combined with the extension and retraction of the second and third-stage telescopic arms, to coarsely position the end effector near the high-altitude pipeline area. The frame observation camera captures pipeline images, and after the industrial control computer analyzes the position, it drives the three-axis rotary table for fine-tuning, aligning the clamping frame with the pipeline axis. The left and right clamping mechanisms move synchronously, gripping the pipeline from both sides to form a stable working reference. The camera acquires pipeline images and transmits them to the industrial control computer. The weld seam positioning module extracts features from the pipeline images, generates an initial 3D scanning trajectory, and segments the trajectory according to the dual-arm collaboration rules. The scanning module uses the visual sensor of the welding tool component to collect weld seam point cloud data along the segmented trajectory and filters the point cloud data. The welding path planning module groups the compressed point cloud by timestamp, fits the target welding point, and generates a continuous welding path. The industrial control computer issues symmetrical welding commands to the left-arm and right-arm welding collaborative robots. The left and right arm welding collaborative robots perform welding operations synchronously according to the planned path, ensuring balanced heat input on both sides of the weld and avoiding deformation. During the welding process, the welding machine, gas cylinder, and wire feeder are powered by a primary telescopic arm cable chain and a tertiary telescopic arm cable chain.

[0011] Furthermore, the mobile support unit is an aerial work platform.

[0012] Furthermore, the left-arm welding collaborative robot performs counter-clockwise welding, while the right-arm welding collaborative robot performs clockwise welding to achieve stress balance.

[0013] Furthermore, the industrial control computer is equipped with a display.

[0014] Furthermore, the weld seam positioning module uses the pipe image captured by the frame observation camera as input, and predicts a three-dimensional scanning trajectory through a convolutional neural network. The form of the three-dimensional trajectory is as follows:

[0015]

[0016] In the formula, This represents a point in three-dimensional space. n points together form a complete trajectory Traj. Refers to three-dimensional Euclidean space.

[0017] Furthermore, the 3D scanning trajectory is generated by increasing the path by 10% based on the radius of the pipe. The scanning trajectory is divided into two 180° arcs by the vertical plane of the pipe's center, and assigned to the left-arm welding collaborative robot and the right-arm welding collaborative robot respectively.

[0018] Furthermore, the scanning module compresses the data volume through outlier removal and downsampling algorithms.

[0019] Furthermore, a statistical filtering algorithm is used to calculate the mean s and variance σ of the 3D coordinates of the point cloud, retaining those that satisfy the following conditions: The points are used to remove outliers.

[0020] Furthermore, the three-dimensional space is divided into 10mm×10mm×10mm cubes, and the arithmetic mean of the coordinates of points within the same cube is calculated as the new point to reduce the amount of data.

[0021] Furthermore, the welding path planning module groups the point cloud data processed by the scanning module by timestamp, uses weighted least squares to fit the target welding point for each group of point clouds, and then performs curve fitting on all target points to generate a continuous welding path for the left arm welding collaborative robot and the right arm welding collaborative robot to perform welding operations.

[0022] The technical effects of this invention are beyond doubt:

[0023] A. The multi-level positioning capability of the aerial work platform's telescopic boom assembly enables rapid arrival and stable operation at any location at high altitudes, significantly reducing the risks associated with manual high-altitude work. The frame's observation camera allows ground personnel to remotely monitor the operation via an industrial control computer, avoiding direct exposure of personnel to the high-altitude environment. This reduces the probability of safety accidents, adapts to complex spaces, and covers a 360° horizontal and multi-angle vertical operating range.

[0024] B. Dual-robot collaborative welding, combined with intelligent path planning, ensures uniform weld formation and balanced heat input, eliminating the deformation or defects of traditional single-sided welding. Weld strength and airtightness meet high industrial standards, making it particularly suitable for high-risk fields such as petrochemicals and natural gas.

[0025] C. The device utilizes visual recognition and automatic control algorithms to achieve intelligent weld seam identification, path planning, and real-time adjustment, reducing reliance on manual experience. It adapts to interference from high-altitude winds and temperature fluctuations, improving process consistency and reducing the need for skilled welders. Attached Figure Description

[0026] Figure 1 A schematic diagram of a high-altitude dual-arm collaborative intelligent welding robot system;

[0027] Figure 2 This is a schematic diagram of the telescopic arm unit structure;

[0028] Figure 3 This is a schematic diagram of the telescopic boom unit in operation;

[0029] Figure 4 This is a schematic diagram of the welding execution unit structure;

[0030] Figure 5 This is a schematic diagram of the welding execution unit. Detailed Implementation

[0031] The present invention will be further described below with reference to embodiments, but it should not be construed that the scope of the present invention is limited to the following embodiments. Various substitutions and modifications made based on ordinary technical knowledge and common practices in the art without departing from the above-described technical concept of the present invention should be included within the scope of protection of the present invention.

[0032] Example 1:

[0033] See Figures 1-5 This embodiment provides a high-altitude dual-arm collaborative intelligent welding robot system, including a welding execution unit A, a telescopic arm unit B, and a mobile support unit C.

[0034] The telescopic boom unit B includes welding auxiliary equipment and a telescopic boom structure. The welding auxiliary equipment includes a welding machine B2, a top electrical control cabinet B4, an end wire feeder B6, a gas cylinder for MIG / MAG welding B7, and an industrial control computer B8. The industrial control computer B8 is equipped with an intelligent algorithm unit. This intelligent algorithm unit controls the workflow of the entire system. The intelligent algorithm unit includes a weld seam positioning module, a scanning module, and a welding path planning module. The telescopic boom structure includes a platform rotary boom B9, a first-stage telescopic boom B10, a second-stage telescopic boom B11, and a third-stage telescopic boom B12. The platform rotary boom B9 includes a rotary base and a boom body. The rotary base is rotatably mounted on the mobile support unit C using a rotary support mechanism, enabling horizontal rotation. The boom body, welding machine B2, MIG / MAG welding gas cylinder B7, and industrial control computer B8 are fixed to the rotary base. A first-stage telescopic boom B10, capable of pitch and rotation, is installed at the tail of the boom body. An auxiliary telescopic support rod is provided between the boom body and the first-stage telescopic boom B10. The base of the auxiliary telescopic support rod is fixed to the arm body, and the telescopic end is fixed to the primary telescopic arm B10. A primary telescopic arm cable chain B3 and a top electrical control cabinet B4 are fixed to the primary telescopic arm B10. A secondary telescopic arm B11 is installed at the front end of the primary telescopic arm B10 to achieve forward and backward sliding extension. A tertiary telescopic arm B12 is installed at the front end of the secondary telescopic arm B11 to achieve forward and backward sliding extension. A tertiary telescopic arm cable chain B5 and an end wire feeder B6 are installed on the tertiary telescopic arm B12 to achieve cable following and end-of-line robot welding wire feeding.

[0035] The welding execution unit A includes a pipe clamping mechanism, a welding execution mechanism, and a vision recognition mechanism. The pipe clamping mechanism includes a three-axis rotary table assembly A1, a pipe clamping frame A2, a left clamping mechanism A5, and a right clamping mechanism A6. The three-axis rotary table assembly A1 is connected to the end of a three-stage telescopic arm B12, providing three degrees of freedom adjustment: roll, pitch, and yaw. The pipe clamping frame A2 is fixed to the end of the three-axis rotary table assembly A1. The left clamping mechanism A5 and the right clamping mechanism A6 are symmetrically installed inside the pipe clamping frame A2 for clamping the pipe. The welding execution mechanism includes a left-arm welding collaborative robot A3, a right-arm welding collaborative robot A4, and a robot end-effector welding tool assembly A8. The left-arm welding collaborative robot A3 and the right-arm welding collaborative robot A4 are symmetrically installed on the outside of the pipe clamping frame A2. The ends of the left-arm welding collaborative robot A3 and the right-arm welding collaborative robot A4 are each connected to the robot end-effector welding tool assembly A8 via flanges. The welding tool assembly A8 integrates a vision sensor and a welding torch. The vision recognition mechanism includes a frame observation camera A7. The frame observation camera A7 is installed on the side of the pipe clamping frame A2 and communicates with the industrial control computer B8 in real time for remote monitoring of high-altitude operations.

[0036] During operation, the mobile support unit C travels to the work point, and the vehicle platform's rotary arm B9 rotates horizontally to adjust the direction of the telescopic arm. The first-stage telescopic arm B10 pitches and swings, combined with the extension and retraction of the second-stage B11 and third-stage telescopic arms B12, to roughly position the end component A near the high-altitude pipeline area. The frame observation camera A7 captures pipeline images, and after the industrial control computer B8 resolves the position, it drives the three-axis rotary table A1 for fine-tuning, aligning the clamping frame A2 with the pipeline axis. The left clamping mechanism A5 and the right clamping mechanism A6 move synchronously, gripping the pipeline from both sides to form a stable working reference. The camera A7 acquires pipeline images and transmits them to the industrial control computer B8. The weld positioning module extracts features from the pipeline images, generates an initial three-dimensional scanning trajectory, and segments the trajectory according to the dual-arm cooperation rule. The scanning module uses the vision sensor of the welding tool component A8 to acquire weld point cloud data along the segmented trajectory and performs filtering operations on the point cloud data. The welding path planning module groups the compressed point cloud by timestamp, fits the target welding points using weighted least squares, and generates a continuous welding path, ensuring that the welding angle, speed, and heat input parameters match the high-altitude environment. The industrial control computer B8 issues symmetrical welding commands to the left-arm welding collaborative robot A3 and the right-arm welding collaborative robot A4. The left-arm welding collaborative robot A3 and the right-arm welding collaborative robot A4 perform welding operations synchronously according to the planned path, ensuring balanced heat input on both sides of the weld and avoiding deformation. During welding, the welding machine B2, gas cylinder B7, and wire feeder B6 are powered by the primary telescopic arm cable chain B3 and the tertiary telescopic arm cable chain B5, supporting continuous welding.

[0037] Example 2:

[0038] The main content of this embodiment is the same as that of Embodiment 1, except that the mobile support unit C is an aerial work platform. The industrial control computer B8 is equipped with a display.

[0039] Example 3:

[0040] The main content of this embodiment is the same as that of embodiment 1 or 2, wherein the left arm welding collaborative robot A3 performs counterclockwise welding and the right arm welding collaborative robot A4 performs clockwise welding to achieve stress balance.

[0041] Example 4:

[0042] The main content of this embodiment is the same as any one of embodiments 1 to 3. The weld seam positioning module uses the pipe image captured by the frame observation camera A7 as input and predicts a three-dimensional scanning trajectory through a convolutional neural network. The form of the three-dimensional trajectory is as follows:

[0043]

[0044] In the formula, This represents a point in three-dimensional space. n points together form a complete trajectory Traj. This refers to three-dimensional Euclidean space. In this embodiment, n is set to 10, which allows for a detailed description of a three-dimensional trajectory without causing information redundancy due to excessively dense points.

[0045] The 3D scanning trajectory is generated by increasing the path by 10% based on the radius of the pipe. In order to enable both robotic arms to perform scanning and welding operations normally, the scanning trajectory is divided into two 180° arcs by a vertical plane through the center of the pipe, and assigned to the left-arm welding collaborative robot A3 and the right-arm welding collaborative robot A4 respectively.

[0046] Example 5:

[0047] This embodiment is similar in content to any one of embodiments 1-4. However, due to the large amount of point cloud data collected, a single scan typically contains 1 million 3D points. Processing all points simultaneously requires a significant amount of time, severely impacting the overall system efficiency. Therefore, after acquiring the scanned point cloud data, a filtering operation is necessary. The scanning module compresses the data volume through outlier removal and downsampling algorithms. This primarily achieves two objectives: filtering outliers to avoid affecting subsequent path planning, and point cloud downsampling to effectively reduce the amount of point cloud data. Specifically, a statistical filtering algorithm is used to filter the initial point cloud. First, the mean of the 3D coordinates of all points in the point cloud is calculated. With variance Retain points in the point cloud that meet the following requirements:

[0048]

[0049] Since the points in a point cloud are relatively concentrated, using the above range as the screening criterion can effectively eliminate outliers while ensuring the geometric characteristics of the weld. Point cloud downsampling divides the three-dimensional space into several... For a cube of a certain size, for points within the same cube, calculate the arithmetic mean of the three-dimensional coordinates and use it as the coordinates of a new point. Use the new point to represent the remaining points in the cube and remove the remaining points. Through the above two steps, outliers can be effectively removed and the amount of point cloud data can be reduced, while the geometric features of the weld can be better preserved.

[0050] Example 6:

[0051] This embodiment is essentially the same as any one of embodiments 1-5. Specifically, the welding path planning module uses the point cloud data collected and processed by the scanning module to solve for a welding path. The scanned point cloud has timestamp labels. First, the point cloud data is grouped according to the timestamps. Then, each group is processed. For each group, a target welding point is obtained using the least squares method and weighted summation. The target welding point for each group is recorded. Finally, curve fitting is performed on all target welding points to obtain a welding path. Finally, the left-arm welding collaborative robot A3 and the right-arm welding collaborative robot A4 can each perform welding operations according to their respective welding paths.

Claims

1. A high-altitude dual-arm collaborative intelligent welding robot system, characterized in that: The welding execution unit (A), the telescopic arm unit (B) and the mobile support unit (C) are included. The telescopic arm unit (B) includes welding auxiliary equipment and a telescopic arm structure; the welding auxiliary equipment includes a welding machine (B2), a top electric control cabinet (B4), a tail end wire feeder (B6), a two-protection welding gas cylinder (B7) and an industrial computer (B8); the industrial computer (B8) is disposed and runs an intelligent algorithm unit; the intelligent algorithm unit is used for controlling the work flow of the whole system; the intelligent algorithm unit includes a weld positioning module, a scanning module and a welding path planning module; the telescopic arm structure includes a platform rotary arm (B9), a first telescopic arm (B10), a second telescopic arm (B11) and a third telescopic arm (B12); the platform rotary arm (B9) includes a rotary base and an arm body; the rotary base is rotatably installed on the mobile support unit (C) by using a rotary supporting mechanism to realize horizontal rotation; the arm body, the welding machine (B2), the two-protection welding gas cylinder (B7) and the industrial computer (B8) are fixed on the rotary base; a first telescopic arm (B10) capable of pitching rotation is installed at the tail of the arm body; an auxiliary telescopic support rod is arranged between the arm body and the first telescopic arm (B10); the base of the auxiliary telescopic support rod is fixed on the arm body, and the telescopic end is fixed on the first telescopic arm (B10); the first telescopic arm (B10) is fixed with a first telescopic arm drag chain (B3) and a top electric control cabinet (B4); the first telescopic arm (B10) is provided with a second telescopic arm (B11) at the front end to realize forward and backward sliding telescopic movement; the second telescopic arm (B11) is provided with a third telescopic arm (B12) at the front end to realize forward and backward sliding telescopic movement; the third telescopic arm (B12) is provided with a third telescopic arm drag chain (B5) and a tail end wire feeder (B6) to realize cable follow-up and tail end robot welding wire feeding. The welding execution unit (A) includes a pipeline clamping mechanism, a welding execution mechanism and a visual identification mechanism; the pipeline clamping mechanism includes a three-axis rotary disc assembly (A1), a pipeline clamping frame (A2), a left clamping mechanism (A5) and a right clamping mechanism (A6); the three-axis rotary disc assembly (A1) is connected at the end of a three-stage telescopic arm (B12) to provide three degrees of freedom adjustment of rolling, pitching and yawing; the three-axis rotary disc assembly (A1) is fixed at the end of the pipeline clamping frame (A2); the pipeline clamping frame (A2) is symmetrically provided with the left clamping mechanism (A5) and the right clamping mechanism (A6) inside for clamping the pipeline; the welding execution mechanism includes a left-arm welding collaborative robot (A3), a right-arm welding collaborative robot (A4) and a robot end welding tool assembly (A8); the left-arm welding collaborative robot (A3) and the right-arm welding collaborative robot (A4) are symmetrically installed outside the pipeline clamping frame (A2); the left-arm welding collaborative robot (A3) and the right-arm welding collaborative robot (A4) are respectively connected with the robot end welding tool assembly (A8) through flanges; the welding tool assembly (A8) is integrated with a visual sensor and a welding torch; the visual identification mechanism includes a frame observation camera (A7); the frame observation camera (A7) is installed on the side of the pipeline clamping frame (A2) and communicates with a work computer (B8) in real time for remote monitoring of high-altitude operation. During operation, the mobile support unit (C) travels to a work point, the vehicle platform swivel arm (B9) is horizontally rotated to adjust the direction of the telescopic arm; the first-stage telescopic arm (B10) is pitched and swung in combination with the second-stage (B11) and third-stage telescopic arms (B12) to coarsely position the end assembly (A) near the high-altitude pipeline area; the frame observation camera (A7) captures a pipeline image, the work computer (B8) analyzes the position and drives the three-axis rotary disc (A1) to finely adjust the clamping frame (A2) to align with the pipeline axis; the left clamping mechanism (A5) and the right clamping mechanism (A6) simultaneously act to tightly clamp the pipeline from both sides to form a stable work reference; the camera (A7) collects the pipeline image and transmits it to the work computer (B8); the welding seam positioning module extracts features from the pipeline image, generates an initial three-dimensional scanning track and divides the track according to the double-arm collaborative rule; the scanning module collects welding point cloud data along the divided track by using the visual sensor of the welding tool assembly (A8) and filters the point cloud data; the welding path planning module groups the compressed point cloud according to timestamps, fits target welding points and generates a continuous welding path; the work computer (B8) issues a symmetric welding instruction to the left-arm welding collaborative robot (A3) and the right-arm welding collaborative robot (A4); the left-arm welding collaborative robot (A3) and the right-arm welding collaborative robot (A4) simultaneously perform welding operation according to the planned path to ensure balanced heat input on both sides of the welding seam and avoid deformation; during welding, the welding machine (B2), the gas cylinder (B7) and the wire feeder (B6) are driven to supply energy through the first-stage telescopic arm drag chain (B3) and the third-stage telescopic arm drag chain (B5).

2. The dual-arm collaborative intelligent welding robot system at high altitude according to claim 1, characterized in that: The mobile support unit (C) adopts a climbing truck.

3. The dual-arm collaborative intelligent welding robot system at high altitude according to claim 1, characterized in that: The left-arm welding collaborative robot (A3) performs counterclockwise welding, and the right-arm welding collaborative robot (A4) performs clockwise welding, so as to balance stress.

4. The dual-arm collaborative intelligent welding robot system at high altitude according to claim 1, characterized in that: The industrial computer (B8) is provided with a display.

5. The dual-arm collaborative intelligent welding robot system at high altitude according to claim 1, characterized in that: The weld positioning module takes the pipe image captured by the frame observation camera (A7) as input, and predicts a three-dimensional scanning track through a convolutional neural network; the three-dimensional track is in the form of: In the formula, represents a point in three-dimensional space; n points collectively constitute a complete trajectory Traj; refers to three-dimensional Euclidean space.

6. The high-bay dual-arm collaborative intelligent welding robotic system of claim 5, wherein: The three-dimensional scanning track is generated based on a 10% path increase on the basis of a pipe radius arc; the scanning track is equally divided into two 180° arcs through a vertical plane of the pipe center, and is respectively allocated to the left-arm welding collaborative robot (A3) and the right-arm welding collaborative robot (A4).

7. The high-bay two-arm collaborative intelligent welding robotic system of claim 1, wherein: The scanning module compresses the data amount through an outlier removal and downsampling algorithm.

8. The high-bay dual-arm collaborative intelligent welding robot system according to claim 7, characterized in that: The statistical filtering algorithm is used to calculate the mean value s and the variance σ of the three-dimensional coordinates of the point cloud, and the points satisfying are retained to remove outliers.

9. The high-bay dual-arm collaborative intelligent welding robotic system of claim 7, wherein: The three-dimensional space is divided into 10mm×10mm×10mm cubes, and the arithmetic mean value of the coordinates of the points in the same cube is taken as a new point, so as to reduce the data amount.

10. The dual-arm collaborative intelligent welding robot system at high altitude according to claim 1, characterized in that: The welding path planning module groups the point cloud data processed by the scanning module according to timestamps, fits the target welding points for each group of point cloud data by using a weighted least square method, and then performs curve fitting on all the target points to generate a continuous welding path, so as to execute the welding operation by the left-arm welding collaborative robot (A3) and the right-arm welding collaborative robot (A4).

Images