A method and system for milling the weld seam of a rotor support based on an industrial robot.

Abstract

This invention relates to the field of generator bracket manufacturing technology, and more particularly to a method for milling the weld seam of a rotor bracket based on an industrial robot. The method includes obtaining a first plane model of the rotor bracket, a second plane model of the rotor bracket, and an actual weld seam model; obtaining a target weld seam model based on the first plane model, the second plane model, and a preset weld seam radius; obtaining the weld seam milling amount, the weld seam milling direction, and the weld seam milling path based on the first plane model, the second plane model, the target weld seam model, and the actual weld seam model; and milling and grinding the rotor bracket weld seam according to the weld seam milling amount and the weld seam milling path. This invention can significantly improve product quality and shorten the manufacturing cycle by 50%. The method of this invention can achieve standardized and automated operation of rotor bracket measurement and milling, thereby reducing the manufacturing cost of turbine rotor brackets and enhancing enterprise competitiveness.

Description

Technical Field

[0001] This invention relates to the field of generator bracket manufacturing technology, and in particular to a method and system for milling the weld seam of a rotor bracket based on an industrial robot. Background Technology

[0002] Currently, the key process for rotor supports in hydroelectric generator sets is the assembly and connection between the rotor and the support, which is currently done manually, with the weld seams ground into rounded corners. The weld seams retain a 10mm weld scar, which is currently mainly manually ground and smoothed into a transition arc. This process is labor-intensive, inefficient, results in poor product consistency, and generates significant dust, which is detrimental to occupational health and does not meet the requirements of green production. Therefore, it is necessary to develop a robotic milling method and system for rotor support weld seams to achieve automated milling, ensure milling efficiency and accuracy, improve the milling environment, and promote the development of green production. Summary of the Invention

[0003] This invention provides a method and system for milling the weld seam of a rotor support based on an industrial robot, in order to solve the technical problems mentioned in the background art.

[0004] One technical solution of the present invention is as follows: A method for milling the weld seam of a rotor support based on an industrial robot, comprising:

[0005] S10: Obtain the first plane model of the rotor support, the second plane model of the rotor support, and the actual weld model of the rotor support;

[0006] S20: Obtain the target weld model based on the first plane model of the rotor support, the second plane model of the rotor support, and the preset weld radius;

[0007] S30: Based on the first plane model of the rotor support, the second plane model of the rotor support, the target weld model, and the actual weld model, obtain the weld milling amount, the weld milling direction, and the weld milling path;

[0008] S40: Mill and grind the rotor support weld according to the weld milling amount and weld milling path.

[0009] Further, step S10 includes:

[0010] The rotor support weld was measured to obtain point cloud data of the rotor support weld;

[0011] The point cloud data is filtered and denoised to obtain simplified point cloud data;

[0012] Based on the simplified point cloud data, the first plane model of the rotor support, the second plane model of the rotor support, and the actual weld model are obtained.

[0013] Furthermore, based on the simplified point cloud data, the first plane model of the rotor support, the second plane model of the rotor support, and the actual weld model are obtained, including:

[0014] The random sampling consensus algorithm and recursive method are used to obtain all fitting planes of the simplified point cloud data. Based on the point cloud data in the fitting plane and the plane fitting deviation, all fitting planes are filtered to obtain the first plane model and the second plane model of the rotor support.

[0015] Further, the point cloud data is filtered and denoised to obtain simplified point cloud data, including:

[0016] Point cloud data is downsampled by a voxel filtering algorithm, and point cloud data with distributions exceeding a preset variance threshold is removed by a statistical outlier filtering algorithm to obtain simplified point cloud data.

[0017] Further, step S20 includes:

[0018] Based on the first plane model of the rotor support, the second plane model of the rotor support, and the preset weld radius, the tangent circle between the first plane model of the rotor support and the second plane model of the rotor support is determined, and the arc of the tangent circle between the first plane model of the rotor support and the second plane model of the rotor support is the target weld model.

[0019] Further, step S30 includes:

[0020] Obtain the plane intersection line between the first plane model and the second plane model of the rotor support;

[0021] The processing start point is obtained based on the bounding box of the point cloud data of the plane intersection line and the target weld model;

[0022] Obtain the angle bisector of the first plane model and the second plane model of the rotor support, wherein the perpendicular feature line of the angle bisector is the weld milling direction;

[0023] Based on the target weld model and the actual weld model, the weld milling amount and the actual weld scar distribution are obtained through the point cloud ICP matching algorithm.

[0024] The weld milling path is determined based on the actual weld scar distribution and the processing start point.

[0025] Further, step S40 includes:

[0026] When the milling amount of the weld is less than the preset value, the weld scar of the weld is ground.

[0027] When the amount of milling required for the weld exceeds the preset value, the weld scar is first milled and then ground.

[0028] Another technical solution of the present invention is as follows: a rotor support weld milling system, used to implement any of the above-described rotor support weld milling methods, comprising:

[0029] The system includes a control module, an industrial robot, and a positioning mechanism. The control module is communicatively connected to the industrial robot, and a rotor support is mounted on the positioning mechanism. The industrial robot faces the rotor support.

[0030] The control module is able to acquire the first plane model of the rotor support, the second plane model of the rotor support, and the actual weld model of the rotor support.

[0031] It can obtain the target weld model based on the first plane model of the rotor support, the second plane model of the rotor support, and the preset weld radius;

[0032] It can obtain the weld milling amount, weld milling direction, and weld milling path based on the first plane model of the rotor support, the second plane model of the rotor support, the target weld model, and the actual weld model.

[0033] It can control an industrial robot to mill and grind the weld seam of the rotor support according to the amount and path of weld seam milling.

[0034] Furthermore, the displacement mechanism includes a motor and a hub, the hub is connected to the motor drive shaft, the rotor support is placed on the hub, and the motor is used to drive the hub to rotate, thereby causing the rotor support to rotate and shift.

[0035] Furthermore, during the installation of the industrial robot, hand-eye calibration is performed on the industrial robot.

[0036] The beneficial effects of this invention are as follows: By establishing a target weld model and obtaining an actual weld model, and by comparing the target weld model and the actual weld model, this invention controls an industrial robot to mill the weld. This invention can steadily improve product quality and shorten the manufacturing cycle by 50%. The method of this invention can achieve standardized and automated operation of rotor support measurement and milling, thereby reducing the manufacturing cost of turbine rotor supports and enhancing enterprise competitiveness. Attached Figure Description

[0037] Figure 1 This is a flowchart of the rotor support weld milling method based on industrial robots according to the present invention.

[0038] Figure 2 This is a schematic diagram of the rotor support weld milling system based on an industrial robot according to the present invention.

[0039] Figure 3 This is a schematic diagram of the rotor support model in this invention. Detailed Implementation

[0040] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. The described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0041] In an embodiment of the present invention, Figure 1 This is a flowchart provided by the present invention regarding the specific process of the rotor support weld milling method based on an industrial robot, as shown below. Figure 1 As shown, the present invention specifically includes:

[0042] S10: Obtain the first plane model of the rotor support, the second plane model of the rotor support, and the actual weld model of the rotor support.

[0043] The industrial robot of this invention measures the rotor support weld seam using a camera and photographs the weld scar and the surface features on both sides of the weld seam. This step is repeated multiple times at different positions to ensure that the rotor support morphology can be measured at all robot positions, thereby obtaining point cloud data of the rotor support weld seam.

[0044] The point cloud data is filtered and denoised to obtain simplified point cloud data.

[0045] Specifically, point cloud data is downsampled using a voxel filtering algorithm, where each voxel is a cube with a side length of 1 mm. A statistical outlier filtering algorithm is then used to remove point cloud data with distributions exceeding a preset variance threshold, resulting in simplified point cloud data. The statistical neighborhood radius is set to 5 mm, and the variance threshold is set to 1 standard deviation.

[0046] Based on the simplified point cloud data, the first plane model of the rotor support, the second plane model of the rotor support, and the actual weld model are obtained.

[0047] Specifically, the Random Sample Consensus Algorithm (RANSAC) and a recursive method are used to obtain all fitting planes of the simplified point cloud data. Based on the point cloud data in the fitting planes and the plane fitting deviation, all fitting planes are filtered to obtain the first plane model and the second plane model of the rotor support.

[0048] S20: Obtain the target weld model based on the first plane model of the rotor support, the second plane model of the rotor support, and the preset weld radius.

[0049] like Figure 3As shown, based on the first plane model of the rotor support, the second plane model of the rotor support, and the preset weld radius, the tangent circle tangent to the first plane model of the rotor support and the second plane model of the rotor support is determined. The weld radius can be determined by technicians according to the actual situation.

[0050] The arc of the tangent circle between the first plane model and the second plane model of the rotor support is the target weld model.

[0051] S30: Based on the first plane model of the rotor support, the second plane model of the rotor support, the target weld model, and the actual weld model, obtain the weld milling amount, the weld milling direction, and the weld milling path.

[0052] The intersection line of the first plane model and the second plane model of the rotor support is obtained by solving the equations of the first plane model and the second plane model of the rotor support.

[0053]

[0054] Where a1, b1, c1, and d1 are the parameters of the first plane model of the rotor support, and a2, b2, c2, and d2 are the parameters of the second plane model of the rotor support.

[0055] Based on the bounding box of the point cloud data of the target weld model and the plane intersection line, the processing start point is obtained, specifically the intersection point of the plane intersection line and the bounding box.

[0056] Obtain the angle bisector of the first plane model and the second plane model of the rotor support. The vertical feature line of the angle bisector is the weld milling direction, which refers to the Z-axis tool axis vector data in the grinding tool TCP.

[0057] Specifically, the direction vector u is established perpendicular to the plane on the angle bisector plane, and the direction vector u is the milling direction of the weld. The formula for the direction vector u is as follows:

[0058] The formula for the direction vector u is as follows:

[0059] u=n1×n2=(b1*c2-c1*b2,c1*a2-a1*c2,a1*b2-b1*a2)

[0060] Where n1 and n2 are the normals of the first plane model and the second plane model of the rotor support, respectively.

[0061] Based on the target weld model and the actual weld model, the weld milling amount and the actual weld scar distribution are obtained through the point cloud ICP matching algorithm. Based on the actual weld scar distribution and the processing start point, the weld milling path is determined.

[0062] By using the point cloud ICP matching algorithm, the actual weld scar distribution and the deviation values ​​between each actual weld scar and the target model can be obtained from the actual weld scar model and the target model, which is the weld milling amount.

[0063] S40: Mill and grind the rotor support weld according to the weld milling amount and weld milling path.

[0064] The determination is made by checking whether the actual weld model can encompass the target weld model. When the weld milling amount is less than the preset value, the weld scar is ground; when the weld milling amount is greater than the preset value, the weld scar is milled first, and then ground.

[0065] Since the milling accuracy of general industrial robots is ±0.5mm, the preset value can be set to 2mm. Therefore, if the milling amount of the weld is less than 2mm, it can be directly ground. If it exceeds 2mm, the weld scar of the weld should be milled first, and then ground.

[0066] Based on the planned weld milling path, select the feature surface of the target weld model, use the machining along the line, select the plane intersection point as the tool tip point, and set the tool axis vector to the weld milling direction to ensure that the arc and the grinding distance on both sides are consistent and the transition is uniform.

[0067] This invention establishes a target weld model and obtains an actual weld model. By comparing the target weld model and the actual weld model, an industrial robot is controlled to mill the weld. This invention can reliably improve product quality and shorten the manufacturing cycle by 50%. The method of this invention can achieve standardized and automated operation of rotor support measurement and milling, thereby reducing the manufacturing cost of turbine rotor supports and enhancing enterprise competitiveness.

[0068] In another technical solution of the present invention: a rotor support weld seam milling system for implementing any of the above-described rotor support weld seam milling methods, comprising: a control module, an industrial robot 1 and a positioning mechanism 2, wherein the control module is communicatively connected to the industrial robot, the rotor support is mounted on the positioning mechanism 2, the industrial robot is facing the rotor support, and the positioning mechanism 2 is used to rotate the rotor support.

[0069] Industrial robot 1 can specifically be an ABB6700_200kg_2.6m. This industrial robot integrates a camera, a grinding electric spindle, and force control, and has measurement, milling, and constant force grinding functions. The control module is the controller of the industrial robot and is used in conjunction with the industrial robot.

[0070] The control module can acquire the first plane model of the rotor support, the second plane model of the rotor support, and the actual weld model; it can acquire the target weld model based on the first plane model of the rotor support, the second plane model of the rotor support, and the preset weld radius; it can acquire the weld milling amount, the weld milling direction, and the weld milling path based on the first plane model of the rotor support, the second plane model of the rotor support, the target weld model, and the actual weld model; and it can control the industrial robot to mill and grind the weld of the rotor support based on the weld milling amount and the weld milling path.

[0071] The displacement mechanism 2 includes a motor and a hub. The hub is connected to the motor drive shaft, and the rotor support is placed on the hub. The motor drives the hub to rotate, causing the rotor support to rotate and shift. Specifically, four displacement mechanisms can be provided, located at the bottom of the rotor support, to support the rotor support and allow it to rotate simultaneously.

[0072] In one embodiment of the present invention, during the installation of the industrial robot, hand-eye calibration is performed on the industrial robot. Specifically, this involves first performing binocular structured coordinate calibration and TCP calibration. With the chessboard grid fixed, the robot's pose is changed in TCP relocation mode, and multiple images are captured for hand-eye calibration to determine the relative pose between the robot's base coordinate system and the camera coordinate system.

[0073] The milling cutter for machining the rotor support weld seam is a round nose cutter, φ25mm, with insertable inserts, made of coated carbide. The cutter settings are: 8mm row spacing, 0.3-0.5mm depth of cut, 20-35mm / s feed rate, and 4500-6500r / min spindle speed. For constant force grinding, 3M carbide grinding discs are used, 80 grit, φ80mm, with a feed rate of 10-15mm / s, a spindle speed of 5500-7000r / min, and a grinding pressure of 20-35N.

[0074] The beneficial effects of this technical solution are the same as those of the rotor support weld milling method, so they will not be repeated here.

[0075] Finally, it should be noted that the above specific embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to examples, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method of rotor support weld milling based on an industrial robot, characterized in that, The method comprises the following steps: S10: obtaining a rotor support first plane model, a rotor support second plane model and an actual weld model of the rotor support; S20: obtaining a target weld model according to the rotor support first plane model, the rotor support second plane model and a preset weld radius; S30: obtaining a weld milling and grinding amount, a weld milling and grinding direction and a weld milling and grinding path according to the rotor support first plane model, the rotor support second plane model, the target weld model and the actual weld model; The plane intersection line of the rotor support first plane model and the rotor support second plane model is obtained by solving the equations of the rotor support first plane model and the rotor support second plane model: ， Wherein, a1, b1, c1 and d1 are parameters of the rotor support first plane model, and a2, b2, c2 and d2 are parameters of the rotor support second plane model; A processing starting point is obtained according to the plane intersection line and the bounding box of the point cloud data of the target weld model, and the intersection point of the plane intersection line and the bounding box is the processing starting point; An angular bisector surface of the rotor support first plane model and the rotor support second plane model is obtained, and a vertical feature line of the angular bisector surface is the weld milling and grinding direction, which refers to the Z-axis vector data in the TCP of the grinding tool; A direction vector u is established on the angular bisector surface, and the direction vector u is perpendicular to the plane intersection line, and the direction vector u is the weld milling and grinding direction, wherein the formula of the direction vector u is as follows: The formula of the direction vector u is as follows: ， wherein , are the normal to the first and second rotor support plane model, respectively. According to the target weld model and the actual weld model, the weld milling and grinding amount and the actual weld scar distribution are obtained by using the point cloud ICP matching algorithm, and the weld milling and grinding path is determined according to the actual weld scar distribution and the processing starting point; S40: milling and grinding the rotor support weld according to the weld milling and grinding amount and the weld milling and grinding path.

2. The rotor support weld beveling method of claim 1, wherein, The step S10 comprises: The point cloud data of the rotor support weld is obtained by measuring the rotor support weld; The simplified point cloud data is obtained by filtering and denoising the point cloud data; The rotor support first plane model, the rotor support second plane model and the actual weld model are obtained according to the simplified point cloud data.

3. The industrial robot-based rotor-shaft weld-bead milling method of claim 2, wherein, The rotor support first plane model, the rotor support second plane model and the actual weld model are obtained according to the simplified point cloud data, which comprises: All fitting planes of the simplified point cloud data are obtained by using the random sampling consistency algorithm and the recursive method, all fitting planes are filtered according to the point cloud data in the fitting plane and the plane fitting deviation, and the rotor support first plane model and the rotor support second plane model are obtained.

4. The industrial robot-based rotor support weld milling method of claim 2, wherein, The simplified point cloud data is obtained by filtering and denoising the point cloud data, which comprises: The point cloud data is down-sampled by using the voxel filtering algorithm, and the point cloud data distributed beyond the preset variance threshold is stripped by using the statistical outlier filtering algorithm, and the simplified point cloud data is obtained.

5. The industrial robot-based rotor support weld beveling method of claim 1, wherein, The step S20 comprises: The tangent circle tangent to the rotor support first plane model and the rotor support second plane model is determined according to the rotor support first plane model, the rotor support second plane model and the preset weld radius, and the circular arc of the tangent circle between the rotor support first plane model and the rotor support second plane model is the target weld model.

6. The industrial robot-based rotor support weld beveling method of claim 1, wherein, The step S40 comprises: When the weld milling and grinding amount is less than the preset value, the welding scar of the weld is polished; When the weld milling and grinding amount is greater than the preset value, the welding scar of the weld is first milled and then polished.

7. A rotor support weld bead milling system for carrying out the rotor support weld bead milling method according to any one of claims 1 to 6, characterized in that Comprise: The control module, the industrial robot and the displacement mechanism, the control module is connected with the industrial robot, the rotor support is installed on the displacement mechanism, and the industrial robot is opposite to the rotor support; The control module can obtain the rotor support first plane model, the rotor support second plane model and the actual weld model of the rotor support; The target weld model can be obtained according to the rotor support first plane model, the rotor support second plane model and the preset weld radius; The weld milling and grinding amount, the weld milling and grinding direction and the weld milling and grinding path can be obtained according to the rotor support first plane model, the rotor support second plane model, the target weld model and the actual weld model; The industrial robot can be controlled to mill and polish the rotor support weld according to the weld milling and grinding amount and the weld milling and grinding path.

8. The rotor support weld beveling system of claim 7, wherein, The displacement mechanism comprises a motor and a hub, the hub is connected with the driving shaft of the motor, the rotor support is placed on the hub, and the motor is used to drive the hub to rotate, so that the rotor support rotates and displaces.

9. The rotor support weld beveling system of claim 7, wherein, When the industrial robot is installed, the hand-eye calibration of the industrial robot is carried out.

Images