Online repairing device and method for additive and subtractive flange face

Abstract

The invention discloses an additive and subtractive flange face online repairing device which comprises a base module, a transmission module, an additive module and a turning module, and the additive module and the turning module are arranged at the two ends of a dragging tool row in the transmission module correspondingly. The transmission module is arranged on the base module, the whole device is arranged in a to-be-repaired flange cylinder, and the base module can be finely adjusted relative to the position of the inner wall of the to-be-repaired flange cylinder. The invention further discloses an online repairing method for the additive and subtractive flange face. The method has the advantages that on-site repair can be carried out without dismounting the flange, production loss caused by equipment shutdown is reduced, the production efficiency is remarkably improved, and continuity of an industrial process and maximization of economic benefits are ensured.

Description

Technical Field

[0001] The present application relates to the technical field of flange surface repair, and in particular to a flange repair method using additive and subtractive materials. Background Art

[0002] As a key component connecting pipes and equipment interfaces, flange surfaces play an important role in the industrial field. Their intact condition is crucial to ensuring the sealing, safety and stability of the entire system.

[0003] Long-term exposure to harsh industrial environments can easily cause damage such as corrosion and wear on the flange surface, which can lead to reduced sealing performance and affect production safety and efficiency. Due to the inconvenience of flange removal and the high requirements for stability during production, it is urgent to develop an online flange surface repair device to reduce downtime and related losses.

[0004] It is known that the authorization announcement number is CN 217898380 U, and the invention name is: A flange repair machine for repairing flanges, which can only perform processing and repair, and cannot perform corrosion-resistant treatment on the surface material of the flange surface. Summary of the invention

[0005] The first object of the present invention is to provide an online repair device for flange surface by adding or subtracting materials, and an online flange surface repair device to reduce downtime and related losses.

[0006] An online repair device for an additive and subtractive flange surface is implemented by the following technical scheme, comprising a base module, a transmission module, an additive module and a turning module, wherein the additive module and the turning module are respectively arranged at two ends of a drag tool row in the transmission module, and both the additive module and the turning module make linear motion on the drag tool row; the transmission module is arranged on the base module, and the transmission module drives the drag tool row to rotate; the device as a whole is arranged in a flange cylinder to be repaired through the base module, and the radial position and axial position of the base module relative to the inner wall of the flange cylinder to be repaired can be fine-tuned; wherein the welding gun head end in the additive module is located on the flange surface to be repaired, and the blade in the turning module is located on the flange surface to be repaired.

[0007] In a further embodiment of the technical solution of the present invention, the additive module is slidably arranged on the drag tool row through a transverse slider of the additive module, and includes an additive welding gun, a six-axis collaborative robotic arm and an additive module screw slider mechanism. The additive module screw slider mechanism is fixed to the transverse slider of the additive module, and the sliding directions of the two are perpendicular. The fixed end of the six-axis collaborative robotic arm is connected to the screw slider mechanism of the additive module. The additive welding gun is arranged at the end of the six-axis collaborative robotic arm. The additive welding gun automatically feeds the welding wire in a quantitative manner and melts it at the defective position on the flange surface to be repaired.

[0008] In a further embodiment of the technical solution of the present invention, the turning module is slidably arranged on the drag tool row through a turning module transverse slider, and includes an external cylindrical turning tool and a turning module screw slider mechanism. The turning module screw slider mechanism is fixed to the turning module transverse slider, and the sliding directions of the two are perpendicular. The external cylindrical turning tool is mounted on the turning module screw slider mechanism. Under the condition that the turning module screw slider mechanism and the drag tool row rotate, the external cylindrical turning tool realizes spiral turning.

[0009] In a further embodiment of the technical solution of the present invention, the transmission module also includes a servo mechanism that drives the drag tool row to rotate and drives the additive module and the turning module to make linear motion on the drag tool row, and the transmission module is arranged on the base module through a flange; wherein the servo mechanism includes a servo motor and a transmission gear box, and a housing of the servo motor and the transmission gear box.

[0010] In a further preferred embodiment of the technical solution of the present invention, the online repair device for the additive and subtractive flange surface further includes a counterweight module, an additive module and a turning module are arranged on one side of the drag tool row, and the counterweight module is arranged on the other side of the drag tool row. The counterweight module is used to balance the weight of the entire device to ensure stability during processing and operation, and to avoid precision loss or safety hazards caused by imbalance.

[0011] The second purpose of the present invention is to propose an online repair method for an additive and subtractive flange surface, the method comprising: using a TOF industrial camera to collect three-dimensional data of the flange surface to be repaired, and inputting the data into a convolutional neural network training model to output the coordinates of the defect position on the flange surface to be repaired; driving the additive module to move according to the defect position coordinate information and performing additive cladding repair on the defects on the flange surface to be repaired; after completion, moving the turning module to perform contour processing correction on the flange surface to be repaired.

[0012] Further preferred embodiment of the technical solution of the present invention is an additive flange surface online repair method, comprising the following steps: Step 1, the TOF industrial camera is fixed above the flange cylinder to be repaired, the additive and subtractive flange surface online repair device is arranged as a whole in the flange cylinder to be repaired, the welding gun head in the additive module is located on the flange surface to be repaired, and the blade in the turning module is located on the flange surface to be repaired; Step 2: The TOF industrial camera takes a three-dimensional image of the flange cylinder to be repaired; Step 3: preprocess the three-dimensional image data, and construct a three-dimensional reconstruction model of the flange cylinder to be repaired based on the preprocessed data; Step 4: Input the 3D reconstruction model into the trained convolutional neural network training model, output the defect location coordinates, and input them into the ARM controller in real time; Step 5: The ARM controller drives the additive module to move according to the defect position coordinate information input in step 4, and the additive module itself evenly clads the flux-cored welding wire to each defect position on the flange surface to be repaired according to the parameters preset by the ARM controller; After step 6 and step 5 are completed, the ARM controller starts the turning module, and the turning module itself corrects the flange surface to be repaired according to the spiral line based on the parameters preset by the ARM controller.

[0013] The repair method includes CNN defect recognition algorithm, dynamic tracking recognition module and six-axis collaborative execution module. The three-dimensional data of the flange surface is collected by a high-precision TOF (Time-of-Flight) industrial camera. The data is input into the CNN convolutional neural network for deep learning and special training on defect recognition such as cracks and dents. The recognition results are converted into precise spatial coordinate information to determine the precise positioning of the defect position. The coordinate information is used as the target position coordinate of the additive flange repair equipment to drive the additive welding gun to the location of the defect. Under the guidance of the six-axis collaborative robot arm, the additive welding gun performs additive cladding repair on the defects on the flange surface, and melts the flux-cored welding wire containing wear-resistant and corrosion-resistant elements onto the surface of the flange surface to improve the wear resistance and corrosion resistance of the flange surface. Subsequently, the turning module is used for fine turning, and the geometric dimensions of the repaired flange surface are accurately repaired to ensure that the flange surface is restored to the original design requirements.

[0014] In a further preferred embodiment of the repair method of the present invention, the ARM controller includes a data input and processing module, a CNN model module, a feedback and monitoring module, and a linkage control module. The data input and processing module collects the 3D image data of the TOF industrial camera and performs denoising and filtering. Based on the processed 3D image data, the ICP point cloud registration algorithm is used to achieve registration through the error between point clouds, and a 3D reconstruction model of the flange cylinder to be repaired is constructed and output; The CNN model module receives the 3D reconstruction model of the flange cylinder to be repaired output by the data input and processing module, uses the convolution kernel to perform convolution operation on the 3D reconstruction model to extract feature patterns, and samples the feature patterns to output the defect position coordinates; The linkage control module reads the defect position coordinates output by the CNN model module, performs first-order position inverse solution and second-order velocity inverse solution, obtains the position, velocity and acceleration of each joint respectively, calculates the rotation angle of each joint motor through motion planning, and then drives each joint motor to rotate through motion execution, thereby controlling the movement of the six-axis collaborative robot arm, and the movement position of the six-axis collaborative robot arm is fed back to the monitoring module through the encoder input; The feedback and monitoring module inputs the real-time monitored position information of the additive welding gun and the six-axis collaborative robotic arm, and adjusts the position of the additive welding gun and the six-axis collaborative robotic arm in real time by comparing it with the preset position information, and outputs it to the additive welding gun and the six-axis collaborative robotic arm.

[0015] The beneficial effects of the present invention compared with the prior art are as follows: 1. The repair method of the present invention combines visual recognition and detection technology with additive cladding technology. Under the guidance of a six-axis collaborative robot arm, the additive welding gun can accurately spray the cladding material onto the defective part, thus achieving accurate repair of the defective part. 2. The repair method of the present invention uses a defect processing algorithm based on the Jacobian matrix inverse solution algorithm and the speed and position resolution visual algorithm to achieve efficient defect recognition and precise motion control, thereby improving the intelligence of the online repair equipment; 3. The repair method of the present invention combines arc additive technology with traditional turning processing, ensuring high standards of surface quality and geometric accuracy after repair, significantly enhancing the corrosion resistance of the flange surface, reducing maintenance costs, and extending the repair cycle; 4. The repair method of the present invention can be repaired on site without removing the flange, which reduces production losses caused by equipment downtime, significantly improves production efficiency, and ensures the continuity of the industrial process and maximizes economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a structural schematic diagram of an embodiment of the present application in which the online repair device for the flange surface of the additive and subtractive material is installed as a whole in the flange cylinder to be repaired; Figure 2 It is a schematic diagram of the overall structure of an embodiment of the present application; Figure 3 is a schematic diagram of a base module in an embodiment of the present application; Figure 4 is a schematic diagram of a transmission module in an embodiment of the present application; Figure 5 is a schematic diagram of an additive module in an embodiment of the present application; Figure 6 is a schematic diagram of a turning module in an embodiment of the present application; Figure 7 is a schematic diagram of a counterweight module in an embodiment of the present application; Figure 8 is a structural block diagram of the ARM controller in an embodiment of the present application; Fig. 9 is a flowchart of a data input and processing module in an embodiment of the present application; Fig.10 is a flow chart of the motion control system of the six-axis collaborative robot arm in the embodiment of the present application; Fig.11 It is the inverse solution logic diagram of the Jacobian matrix of the online repair method of the flange surface with additive and subtractive materials in the embodiment of the present application; Fig.12 It is a visual method diagram of distinguishing speed position of the online repair method of the flange surface of the additive and subtractive material in the embodiment of the present application; In the figure, 1-base module, 110-base, 130-screw, 140-locking nut, 150-fixed support, 160-adjustable support, 161-fixed part. 162-adjustment part, 2-transmission module, 210-servo motor, 220-transmission gearbox, 230-drag knife row, 231-drag knife row frame, 232-linear guide, 233-additive module screw, 234-turning module screw, 235-screw fixed support, 236-additive module screw nut, 237-turning module screw nut, 238-gearbox shaft sleeve, 239-gearbox cover, 240-gearbox housing, 3-additive module, 310-additive welding gun, 3 11-welding power supply connection line, 312-insulating sleeve, 313-welding gun head, 320-six-axis collaborative robot arm, 321-front end clamp, 322-first joint Arm_link, 323-second joint Arm_link, 324-third joint Arm_link, 325-fourth joint Arm_link, 326-fixed joint Arm_link, 330-additive module lead screw slider mechanism, 331-additive module dovetail groove A Plate, 332-additive module screw nut fixing screw, 333-additive module screw, 334-additive module dovetail groove B plate, 335-additive module screw fixed end, 340-additive module lateral slider, 341-additive module lateral motion slider, 342-additive module guide rail matching end, 4-turning module, 410-external turning tool, 411-tool bar, 412-blade, 420-turning module screw slider mechanism, 421-turning module dovetail groove A plate, 422-turning mold Block screw nut fixing screw, 423-external cylindrical turning tool fixing screw, 424-turning module dovetail groove B plate, 425-turning module screw fixed end, 426-turning module screw, 430-turning module transverse slider, 431-turning module transverse motion slider, 432-matching end fixing screw, 433-turning module guide rail matching end, 5-counterweight module, 511-counterweight fixing plate, 512-counterweight block, 513-counterweight block bolt through hole, 514-counterweight fixing plate bolt through hole. DETAILED DESCRIPTION

[0017] The specific implementation of the present application is described in detail below with reference to the accompanying drawings.

[0018] refer to Figure 1-6An online repair device for an additive and subtractive flange surface mainly includes five key modules: a base module 1, a transmission module 2, an additive module 3, a turning module 4 and a counterweight module 5.

[0019] like Figure 1 As shown, the base module 1 is the foundation of the entire device. The entire additive and subtractive flange surface online repair device is arranged as a whole in the flange cylinder to be repaired through the base module 1. The radial position and axial position of the base module 1 relative to the inner wall of the flange cylinder to be repaired can be fine-tuned.

[0020] like Figure 2 As shown, the transmission module 2 is arranged on the base module 1. The main function of this module is to provide the necessary power and speed change for the entire device, and ensure that each module can work in a coordinated manner. The transmission module 2 drives the drag blade row 230 to rotate.

[0021] like Figure 2 As shown, the additive module 3 and the turning module 4 are respectively arranged at the two ends of the drag knife row 230 in the transmission module 2, and both the additive module 3 and the turning module 4 make linear motion on the drag knife row 230. The additive module 3 is responsible for adding corrosion-resistant materials on the flange surface to improve the corrosion resistance of the flange surface. The turning module 4 is mainly used to turn the flange surface, remove excess materials, and restore the flatness and precision of the flange surface.

[0022] like Figure 2 As shown, the counterweight module 5 is arranged on one side of the transmission module 2, specifically: the additive module 3 and the turning module 4 are arranged on one side of the drag tool row 230, and the counterweight module 5 is arranged on the other side of the drag tool row 230. The counterweight module 5 is used to balance the weight of the entire device to ensure stability during processing and operation, and avoid precision loss or safety hazards caused by imbalance.

[0023] like Figure 2 As shown, the base module 1 includes a base 110 and a plurality of screw rods 130 evenly distributed on the side of the base 110 . The screw rods 130 are arranged perpendicular to the side of the base 110 and are threadedly connected to the side of the base 110 . A locking nut 140 is arranged at the connection between the screw rods 130 and the side of the base 110 .

[0024] like Figure 2 and 3 As shown, in this embodiment, eight screw rods 130 are preferably designed, and the screw rods 130 are locked on the base 110 by means of locking nuts 140. The purpose of adjusting the locking nuts 140 is to adjust the length along the direction of the screw rods 130. The length of the screw rods retracted into the base 110 is adjusted by rotating the threads. The length adjustment of the threaded rods 130 is used to adjust the radial position of the base module 1 relative to the inner diameter of the flange tube to be repaired.

[0025] like Figure 3 As shown, fixed supports 150 and adjustable supports 160 are alternately arranged at the ends of eight screw rods 130, and the adjustable supports 160 include a fixed portion 161 fixed to the end of the screw rod 130, and an adjusting portion 162 arranged in cooperation with the fixed portion. The fixed portion 161 and the adjusting portion 162 are matched through a dovetail groove and connected by a screw 163 to adjust the axial position of the base module 1 relative to the inner diameter of the flange cylinder to be repaired; specifically: By tightening the screw 163, the adjusting portion 162 can move vertically with the fixing portion 161. By this fine adjustment, the levelness of the entire base 110 can be adjusted, that is, the axial position of the base module 1 relative to the inner diameter of the flange cylinder to be repaired can be adjusted.

[0026] like Figure 2 and 4 As shown, the transmission module 2 is located above the base module 1 , and mainly includes a servo mechanism and a drag blade row 230 .

[0027] The end of the servo mechanism is fixed to the middle bottom plate of the base 110 through a flange, and the drag tool row 230 is set on the servo mechanism. It rotates under the drive of the servo mechanism, and the additive module lead screw nut 236 and the turning module lead screw nut 237 set on the drag tool row 230 respectively drive the additive module 3 and the turning module 4 to make linear motion on the drag tool row 230.

[0028] like Figure 4 As shown, the drag tool row 230 includes a drag tool row frame 231, two linear guide rails 232, an additive module lead screw 233, a turning module lead screw 234, a lead screw fixing support 235, an additive module lead screw nut 236, a turning module lead screw nut 237 and a gear box sleeve 238.

[0029] The two linear guide rails 232 are fixed at the upper and lower edges of the outer side of the drag tool row frame 231 by bolts at upper and lower intervals. One end of the additive module lead screw 233 and the turning module lead screw 234 are rotatably set on the lead screw fixing support 235 through bearings, and the other end is set on the drag tool row frame 231 through a bearing seat; the two lead screw fixing supports 235 are bolted to the drag tool row frame 231.

[0030] In this embodiment, considering the stability of the additive module 3 and the turning module 4 in linear motion on the drag tool row 230, the additive module lead screw 233 and the turning module lead screw 234 are parallel and spaced apart with a guide slide 233-1, the two ends of the guide slide are respectively fixed on the lead screw fixing support 235 and the bearing seat, and the two guide slides pass through the additive module lead screw nut 236 and the turning module lead screw nut 237. Figure 4 shown.

[0031] A third gear is evenly keyed to the additive module lead screw 233 and the turning module lead screw 234, and the third gear is used to transmit the power transmitted by the transmission gear box 220; the third gear on the additive module lead screw 233 or the turning module lead screw 234 is engaged respectively by changing the gear position of the gear inside the transmission gear box 220 to control the start and stop of the additive module lead screw 233 and the turning module lead screw 234.

[0032] The additive module lead screw nut 236 and the turning module lead screw nut 237 are threadedly connected to the additive module lead screw 233 and the turning module lead screw 234 respectively; the additive module 3 and the turning module 4 are fixed to the additive module lead screw nut 236 and the turning module lead screw nut 237 respectively; the gear box sleeve 238 is fixed to the drag tool row frame 231 by bolts.

[0033] The transmission gearbox 220 rotates to drive the gearbox shaft sleeve 238 to rotate through the meshing of the internal gear, thereby driving the drag tool rack 231 to rotate. At the same time, the power is distributed through the meshing of the external gears of the transmission gearbox 220 and the reducer structure to drive the additive module lead screw 233 and the turning module lead screw 234 to rotate, thereby driving the radial feed movement of the additive module 3 and the turning module 4.

[0034] The additive module lead screw 233 and the turning module lead screw 234 are independent drive lead screws, and their start and stop are realized by the gear shift change inside the transmission gear box 220. The details are as follows: The servo mechanism includes a servo motor 210 and a transmission gear box 220. The housing of the servo motor 210 is fixed to the gear box cover 239 of the transmission gear box 220 by bolts. At the same time, the gear box cover 239 is fixed to the gear box housing 240 of the transmission gear box 220 by bolts. The power of the servo motor 210 is output to the first gear inside the transmission gear box 220. The first gear and the second gear transmit power through internal engagement. The second gear is fixed to the gear box sleeve 238 in the transmission gear box 220. Power transmission is carried out through the servo motor 210 and the gears inside the transmission gear box 220, driving the gear box sleeve 238 to rotate. The gear box housing 240 is a fixed part.

[0035] The start and stop of the additive module lead screw 233 and the turning module lead screw 234 are controlled by changing the gear position of the gear inside the transmission gear box 220. Specifically, when the additive module lead screw 233 needs to be started, the gear and synchronizer inside the transmission gear box 220 are manually adjusted to mesh the third gear of the additive module lead screw 233 to complete the power transmission. When the turning module lead screw 234 needs to be started, the meshing state of the current gear is released, the gear and synchronizer inside the transmission gear box 220 are readjusted, and the third gear of the turning module lead screw 234 is meshed.

[0036] like Figure 5 As shown, the additive module 3 is arranged at one end of the drag blade row 230 in the transmission module 2 , and is specifically slidably arranged on the drag blade row 230 through an additive module transverse slider 340 .

[0037] The additive module 3 includes an additive welding gun 310, a six-axis collaborative robot arm 320, and an additive module lead screw slider mechanism 330. The additive module 3 moves to the location of the defect, and the additive welding gun 310, under the guidance of the six-axis collaborative robot arm 320, performs additive cladding repair on the defect on the flange surface to be repaired, and clads the flux-cored welding wire containing wear-resistant and corrosion-resistant elements onto the flange surface to improve the wear resistance and corrosion resistance of the flange surface.

[0038] The specific implementation structure of the additive module 3 is as follows: The additive module lead screw slider mechanism 330 is fixed to the additive module transverse slider 340, and the sliding directions of the two are perpendicular. The fixed end of the six-axis collaborative robot arm 320 is connected to the additive module lead screw slider mechanism 330. The additive welding gun 310 is arranged at the end of the six-axis collaborative robot arm 320. The additive welding gun 310 automatically feeds the welding wire in a quantitative manner and melts it on the defective position on the flange surface to be repaired.

[0039] In this embodiment, the additive welding gun 310 is a purchased part, preferably the model: MB EVO PRO 36, purchased from the German Binzel Group. The six-axis collaborative robot arm 320 is a purchased part, preferably the model: Model: ordinary mechanical claw with STM32 electronic control + adapter, purchased from WHEELTEC Wheel Fun Technology.

[0040] The additive welding gun 310 is a core component, mainly including a welding power connection line 311, an insulating sleeve 312 and a welding gun head 313. The six-axis collaborative robot arm 320 includes a front end clamp 321 and five joints, which are used to support and adjust the welding gun angle, namely: the first joint Arm_link322, the second joint Arm_link323, the third joint Arm_link324, the fourth joint Arm_link325 and the fixed joint Base_link326.

[0041] The additive module lead screw slider mechanism 330 includes an additive module dovetail groove A plate 331, an additive module lead screw nut fixing screw 332, an additive module lead screw 333, an additive module dovetail groove B plate 334 and an additive module lead screw fixing end 335. The additive welding gun 310 is fixed to the front end of the first joint Arm_link322 of the six-axis collaborative robot 320 through the front end clamp 321, and the fixed joint Base_link326 of the six-axis collaborative robot 320 is fixed to the additive module dovetail groove A plate 331 through the additive module lead screw nut fixing screw 332. The additive module dovetail groove A plate 331 and the additive module dovetail groove B plate 334 are installed in coordination, and the additive module lead screw 333 is threadedly connected between the two. By rotating the additive module lead screw 333, the longitudinal movement between the additive module dovetail groove A plate 331 and the additive module dovetail groove B plate 334 is realized, thereby realizing the longitudinal movement of the additive module 3.

[0042] like Figure 4 and 5 As shown, the additive module transverse slider 340 includes an additive module transverse motion slider 341 and an additive module guide rail mating end 342. The additive module transverse motion slider 341 is fixed to the additive module lead screw nut 236 on the drag knife row 230 by screws, and the additive module guide rail mating end 342 is installed in cooperation with the two linear guide rails 232 on the drag knife row 230 to jointly realize transverse linear motion.

[0043] like Figure 6 As shown, the turning module 4 is arranged at the other end of the dragging tool row 230 in the transmission module 2, and is specifically slidably arranged on the dragging tool row 230 through a turning module transverse slider 430.

[0044] The turning module 4 includes an external turning tool 410 and a turning module lead screw slider mechanism 420. The turning module 4 performs precision turning on the flange surface to be repaired after material addition, and accurately repairs the geometric dimensions of the repaired flange surface to ensure that the flange surface is restored to the original design requirements.

[0045] The specific implementation structure of the turning module 4 is as follows: The turning module screw slider mechanism 420 is fixed to the turning module transverse slider 430, and the sliding directions of the two are perpendicular. The external cylindrical turning tool 410 is installed on the turning module screw slider mechanism 420. The external cylindrical turning tool 410 realizes spiral turning under the condition that the turning module screw slider mechanism 420 and the drag tool row 230 rotate.

[0046] In this embodiment, the external cylindrical turning tool 410 is a core component of the turning module, including a tool rod 411 and a blade 412. The turning module lead screw slider mechanism 420 includes a turning module dovetail groove A plate 421, a turning module lead screw nut fixing screw 422, an external cylindrical turning tool fixing screw 423, a turning module dovetail groove B plate 424, a turning module lead screw fixing end 425 and a turning module lead screw 426. The external cylindrical turning tool 410 is fixed to the turning module dovetail groove A plate 421 by a plurality of external cylindrical turning tool fixing screws 423, ensuring the stability and reliability of the tool. The turning module dovetail groove A plate 421 and the turning module dovetail groove B plate 424 are tightly matched and installed, and the turning module lead screw 426 is threadedly connected between the two. By rotating the turning module lead screw 426, the longitudinal movement between the turning module dovetail groove A plate 421 and the turning module dovetail groove B plate 424 is realized, thereby realizing the longitudinal movement of the turning module 4.

[0047] The turning module transverse slider 430 includes a turning module transverse motion slider 431, a mating end fixing screw 432 and a turning module guide rail mating end 433. The turning module transverse motion slider 431 and the turning module guide rail mating end 433 are fixed by the mating end fixing screw 432, the turning module transverse motion slider 431 is fixed to the turning module lead screw nut 237 on the drag tool row 230 by screws, and the turning module guide rail mating end 433 is installed in cooperation with the two linear guide rails 232 on the drag tool row 230 to jointly realize transverse linear motion.

[0048] like Figure 7 As shown, the counterweight module 5 includes a counterweight fixing plate 511, a counterweight block 512, a counterweight block bolt through hole 513 and a counterweight fixing plate bolt through hole 514. The counterweight fixing plate 511 is used to fix the counterweight block 512 to the transmission gear box 220, and the counterweight block 512 is fixed to the counterweight fixing plate 511 through the counterweight block bolt through hole 513.

[0049] Based on the online repair device for flange surface using additive and subtractive materials, this embodiment proposes an online repair method for flange surface using additive and subtractive materials: a TOF industrial camera is used to collect three-dimensional data of the flange surface to be repaired, and the data is input into a convolutional neural network training model to output the coordinates of the defect position on the flange surface to be repaired; the additive module 2 is driven to move according to the defect position coordinate information and additive cladding repair is performed on the defects on the flange surface to be repaired; after completion, the turning module 4 is moved to perform contour processing correction on the flange surface to be repaired.

[0050] The online repair method for the flange surface of the additive and subtractive material specifically comprises the following steps: Step 1, the TOF industrial camera is fixed above the flange cylinder to be repaired, the additive and subtractive flange surface online repair device is arranged as a whole in the flange cylinder to be repaired, and the welding gun head end 313 in the additive module 3 is located on the flange surface to be repaired, and the blade 412 in the turning module 4 is located on the flange surface to be repaired; Step 2: The TOF industrial camera takes a three-dimensional image of the flange cylinder to be repaired; Step 3: preprocess the three-dimensional image data, and construct a three-dimensional reconstruction model of the flange cylinder to be repaired based on the preprocessed data; Step 4: Input the 3D reconstruction model into the trained convolutional neural network training model, output the defect location coordinates, and input them into the ARM controller in real time; Step 5: The ARM controller drives the additive module 3 to move according to the defect position coordinate information input in step 4, and the additive module 3 itself evenly clads the flux-cored welding wire to each defect position on the flange surface to be repaired according to the parameters preset by the ARM controller; After step 6 and step 5 are completed, the ARM controller starts the turning module 4, and the turning module 4 corrects the flange surface to be repaired according to the spiral line based on the parameters preset by the ARM controller.

[0051] like Figure 8 As shown, the ARM controller in the modification method mainly includes four key modules, a data input and processing module, a CNN model module, a feedback and monitoring module, and a linkage control module. The data input and processing module is responsible for the collection, filtering and three-dimensional reconstruction of point cloud data; the three-dimensional reconstruction model is input into the three-dimensional graphics input layer in the CNN model module to perform convolution kernel morphological detection, surface feature extraction, and defect recognition, and output the location information of the defect; after the linkage control module receives the defect location information, it performs first-order position inverse solution, second-order velocity inverse solution and motion execution; at the same time, the position state will be sent to the feedback and monitoring module for position adjustment. Among them, the feedback and monitoring module and the linkage control module together constitute the motion control system of the six-axis collaborative robot arm 320. As shown Fig.10 shown.

[0052] The data input and processing module collects the three-dimensional image data of the TOF industrial camera and performs denoising and filtering. Based on the processed three-dimensional image data, the ICP point cloud registration algorithm is used to achieve registration through the error between point clouds, and a three-dimensional reconstruction model of the flange cylinder to be repaired is constructed and output.

[0053] like Fig. 9 As shown, the data input and processing module is specifically composed as follows: The data input and processing module includes a data acquisition circuit, a filtering processing circuit and a point cloud three-dimensional reconstruction module. First, in the data acquisition circuit, the sensor of the TOF industrial camera acquires information, and then stores the acquired data. Then, the analog signal is converted into a digital signal through the analog-to-digital conversion circuit, and then the point cloud data that meets the requirements is output through the amplification circuit and the filtering circuit. Secondly, in the filtering processing circuit, the point cloud data output by the data acquisition circuit is obtained, and the error is calculated by matching the nearest points of the source point cloud and the target point cloud. The distance error of the corresponding point pairs between the source point cloud and the target point cloud is calculated, and the transformation calculation is performed based on minimizing the error, and the required rotation and translation transformation is calculated. Then, the calculated transformation is applied to the source point cloud by applying the transformation. Then, through iteration, the nearest point matching, error calculation, transformation calculation and application transformation are repeated until the error converges or the predetermined number of iterations is reached. Finally, the successfully registered point cloud is surface reconstructed, and the smoothness and detail enhancement are performed through post-processing, and then the three-dimensional reconstructed model is output.

[0054] The CNN model module receives the 3D reconstruction model of the flange cylinder to be repaired output by the data input and processing module, uses the convolution kernel to perform convolution operation on the 3D reconstruction model to extract feature patterns, downsamples the feature patterns, and outputs the defect position coordinates.

[0055] Specifically, the three-dimensional reconstruction model output by the data input and processing module is input into the three-dimensional graphics input layer in the CNN model module, and convolution kernel morphology detection, surface feature extraction, and defect recognition are performed. At the same time, the location information of the defect is output. This information is used as target information and sent to the motion control system for motion control of the six-axis collaborative robot arm.

[0056] The CNN model module in the repair method is a known model and is known to those skilled in the art.

[0057] like Fig.10 As shown, the linkage control module reads the defect position coordinates output by the CNN model module, performs first-order position inverse solution and second-order velocity inverse solution, obtains each joint position, velocity and acceleration respectively, calculates the rotation angle of each joint motor through motion planning, and then drives each joint motor to rotate through motion execution, thereby controlling the movement of the six-axis collaborative robot arm, and feeds the movement position of the six-axis collaborative robot arm to the monitoring module through the encoder input feedback.

[0058] The feedback and monitoring module inputs the real-time monitored position information of the additive welding gun 310 and the six-axis collaborative robot arm 320, and adjusts the position of the additive welding gun 310 and the six-axis collaborative robot arm 320 in real time by comparing with the preset position information, and outputs it to the additive welding gun 310 and the six-axis collaborative robot arm 320.

[0059] like Fig.10 As shown in the figure, the linkage control module and the feedback monitoring module together constitute the motion control system of the six-axis collaborative robot arm. The specific process is as follows: The motion control system first reads the target position output by the CNN model module, then performs the first-order position inverse solution and the second-order velocity inverse solution, calculates the angle, velocity and acceleration required for each joint, and then calculates the rotation angle of the motor in each joint through motion planning, and then drives the motor to rotate through motion execution, thereby controlling the motion of the six-axis collaborative robot 320. The motion position of the six-axis collaborative robot 320 is input to the feedback and monitoring module through the encoder, and the actual position is compared with the target position, and the difference between the target position and the actual position is calculated to obtain the position error signal. Then, the PID controller adjusts the control signal according to the signal generated by the error through the comprehensive action of the three parts of proportion (P), integration (I) and differentiation (D) to minimize the error, and then outputs the control signal and transmits it to the actuator. The actuator receives the control signal and performs the corresponding motion, and the drive system realizes position adjustment. After the actuator moves, the actual position is updated, and the actual position signal is transmitted back to the feedback and monitoring module, thereby realizing closed-loop control.

[0060] like Fig.10 As shown, the six-axis collaborative robot 320 communicates and interacts with the ARM controller through the CAN OPEN communication protocol. The height information processed by the intelligent algorithm is converted into the matrix position information in the absolute space coordinate system. The kinematic inverse solution algorithm written at the bottom of the six-axis collaborative robot 320 calculates the specific angle or position that the six-axis collaborative robot 320 needs to reach according to the target position matrix, and then controls the movement of each joint motor to make the six-axis collaborative execution module move to the preset position in the absolute coordinate system, forming a closed-loop control system, and the additive welding gun 310, the six-axis collaborative robot 320 and the additive module lead screw slider mechanism 330 realize linkage control. The IK position inverse solution of the six-axis collaborative robot 320 is to establish the position parameters of each joint through the DH parameters, establish the inverse kinematics equation based on the forward kinematics model and the displacement rotation conversion matrix, and use the numerical analysis method to solve and obtain the joint angle.

[0061] like Fig.11 As shown, the inverse solution of the Jacobian matrix is ​​then used to solve the joint velocity required to achieve the desired velocity and acceleration. Describes the mathematical relationship between joint space acceleration and operation space acceleration. Describes the mathematical relationship between joint space velocity and operation space velocity, The mathematical relationship between the joint space position and the operation space position is described. The differential variable produces positive regulation on KD (differential regulation coefficient) and KP (proportional regulation coefficient). Jacobian matrix solution produces real-time negative feedback regulation on KD (differential regulation coefficient) and KP (proportional regulation coefficient). The second-order derivative JA matrix is ​​introduced and the true solution is approximated through iterative calculation to regulate the joint angle of the six-axis collaborative execution module ( ).

[0062] Based on the real-time measurement of TOF industrial cameras, the target position information can be obtained in real time, and the speed of the trajectory can be derived by calculating the motion trajectory in space. , realize the real-time movement of each joint of the six-axis collaborative robot arm 320 in the spatial coordinate system, where Satisfies the following equation:

[0063] At the same time, the motion position q and motion speed of each joint of the six-axis collaborative execution module The TOF camera then captures it and uses the feedback as the control input. The adjustment of K affects the transient behavior of the TOF camera, thus achieving closed-loop control.

[0064] like Fig.12 As shown in the figure, it is the framework of the motion feedback vision algorithm. The target position of the system is represented by x. d Indicates that an error signal is generated by comparing with the current state. The error signal is calculated by an adder, and the input is the current state, and the output is the error signal. The control gain is used to adjust the size of the error signal and enhance or weaken the system response. This component converts the control input into a command in the joint space, which depends on the current joint position and the initial state of the system, and the joint velocity command obtained from the Jacobian matrix. The time delay between the object and the sensor is calculated by the TOF module, and then the distance or position information is obtained. The system obtains the current state xc through feedback and compares it with the target state to achieve closed-loop control. The final output signal represents the actual state of the system, which is adjusted by feedback control to achieve the set goal.

[0065] This repair method, based on the defect processing algorithm of the Jacobian matrix inverse solution algorithm and the speed and position resolution vision algorithm, provides a high-precision and high-efficiency flange surface defect processing method; the Jacobian matrix inverse solution algorithm provides motion control of the six-axis collaborative robot arm, and the speed and position resolution vision algorithm ensures real-time capture and processing of dynamic defects; the collaborative work of these two algorithms enables the flange surface repair equipment to complete defect repair with high precision and efficiency.

[0066] like Figure 1As shown, in step 1 of the modified method, the online repair device for the flange surface with additive and subtractive materials is integrally arranged in the flange cylinder to be repaired, specifically: First, fix the base module 1 in the flange tube to be repaired. Adjust the X, Y, Pitch and Roll positions of the base module 1 to ensure that the base module is located at the center of the flange surface. Adjust the base module 1 by adjusting the locking nut 140 to adjust the length along the direction of the screw rod 130, so that the fixed supports 150 and the adjustable supports 160 at the ends of the eight screw rods 130 are against the inner wall of the flange tube to be repaired; then adjust the screws 163 on the four adjustable supports 160 to adjust the entire base 110 to a horizontal level.

[0067] Then, after the base module 1 is fixed, the transmission gear box 220, the drag tool row 230, the additive module 3, the turning module 4 and the counterweight module 5 are installed. At this time, the additive and subtractive flange surface online repair device is installed.

[0068] Finally, the TOF industrial camera uses a camera fixing bracket to fix the TOF industrial camera above the flange cylinder to be repaired, so that the shooting surface of the TOF industrial camera covers the entire flange surface of the flange cylinder to be repaired; the TOF industrial camera is used to accurately obtain the flange surface of the flange cylinder to be repaired, so as to accurately measure the target position information of the flange surface defects in the subsequent measurement.

[0069] Among them, the TOF industrial camera is an outsourced part and is purchased directly.

[0070] In step 5 of the repair method, the additive module 3 will uniformly coat the surface of the flange with a flux-cored welding wire containing specific nickel and copper alloy elements according to the parameters preset by the ARM controller. These alloy elements can form a protective oxide film alloy layer during the welding process, thereby significantly improving the corrosion resistance of the flange end face. The flux-cored welding wire mentioned here is a known product, known to those skilled in the art.

[0071] In step 6 of the repair method, after the additive process is completed, the turning module 4 is started to accurately process the added material on the flange surface by a turning tool to restore the original shape and size of the flange surface. The entire repair process is monitored and adjusted in real time by the ARM controller to ensure that the accuracy and effect of the repair are at the best level.

[0072] The above is only a preferred implementation of the present application. It should be pointed out that a person skilled in the art can make several modifications and improvements without departing from the inventive concept of the present application, and these all fall within the scope of protection of the present application.

Claims

1. An online repair device for flange surface with additive and subtractive materials, characterized in that: The invention comprises a base module (1), a transmission module (2), an additive module (3) and a turning module (4), wherein the additive module (3) and the turning module (4) are respectively arranged at two ends of a drag knife row (230) in the transmission module (2), and both the additive module (3) and the turning module (4) perform linear motion on the drag knife row (230); the transmission module (2) is arranged on the base module (1), and the transmission module (2) drives the drag knife row (230) to rotate; the device as a whole is arranged in a flange cylinder to be repaired through the base module (1), and the radial position and axial position of the base module (1) relative to the inner wall of the flange cylinder to be repaired can be finely adjusted; wherein the welding gun head end (313) in the additive module (3) is located on the flange surface to be repaired, and the blade (412) in the turning module (4) is located on the flange surface to be repaired.

2. The online repair device for flange surface with added or subtracted material according to claim 1 is characterized in that: The additive module (3) is slidably arranged on the drag knife row (230) via an additive module transverse slider (340), and comprises an additive welding gun (310), a six-axis collaborative robot arm (320) and an additive module screw slider mechanism (330); the additive module screw slider mechanism (330) is fixed to the additive module transverse slider (340), and the sliding directions of the two are perpendicular; the fixed end of the six-axis collaborative robot arm (320) is connected to the additive module screw slider mechanism (330); the additive welding gun (310) is arranged at the end of the six-axis collaborative robot arm (320); the additive welding gun (310) automatically feeds welding wire in a quantitative manner and melts it on the defective position on the flange surface to be repaired.

3. The online repair device for flange surface with additive and subtractive materials according to claim 1 is characterized in that: The turning module (4) is slidably arranged on the drag tool row (230) via a turning module transverse slider (430), and comprises an external cylindrical turning tool (410) and a turning module screw slider mechanism (420); the turning module screw slider mechanism (420) is fixed to the turning module transverse slider (430), and the sliding directions of the two are perpendicular; the external cylindrical turning tool (410) is mounted on the turning module screw slider mechanism (420); and the external cylindrical turning tool (410) realizes spiral turning under the condition that the turning module screw slider mechanism (420) and the drag tool row (230) rotate.

4. The online repair device for flange surface with added or subtracted material according to claim 1 is characterized in that: The transmission module (2) further comprises a servo mechanism for driving the drag knife row (230) to rotate and driving the additive module (3) and the turning module (4) to perform linear motion on the drag knife row (230); the transmission module (2) is arranged on the base module (1) via a flange; wherein the servo mechanism comprises a servo motor (210) and a transmission gear box (220); and a housing of the servo motor (210) and the transmission gear box (220).

5. The online repair device for flange surface with additive and subtractive materials according to claim 1 is characterized in that: The additive and subtractive material flange surface online repair device further comprises a counterweight module (5), an additive module (3) and a turning module (4) are arranged on one side of the drag tool row (230), and the counterweight module (5) is arranged on the other side of the drag tool row (230).

6. A repair method using the online repair device for flange surface with additive and subtractive materials according to any one of claims 1 to 5, characterized in that: The method comprises: using a TOF industrial camera to collect three-dimensional data of the flange surface to be repaired, inputting the data into a convolutional neural network training model, and outputting the coordinates of the defect position on the flange surface to be repaired; driving an additive module (2) to move according to the defect position coordinate information and performing additive cladding repair on the defect on the flange surface to be repaired; and after completion, moving a turning module (4) to perform contour processing correction on the flange surface to be repaired.

7. The repair method according to claim 6, characterized in that: The steps include: Step 1, a TOF industrial camera is fixed above the flange cylinder to be repaired, the additive and subtractive flange surface online repair device is arranged as a whole in the flange cylinder to be repaired, the welding gun head end (313) in the additive module (3) is located on the flange surface to be repaired, and the blade (412) in the turning module (4) is located on the flange surface to be repaired; Step 2: The TOF industrial camera takes a three-dimensional image of the flange tube to be repaired; Step 3: preprocess the three-dimensional image data, and construct a three-dimensional reconstruction model of the flange cylinder to be repaired based on the preprocessed data; Step 4: Input the 3D reconstruction model into the trained convolutional neural network training model, output the defect location coordinates, and input them into the ARM controller in real time; Step 5: The ARM controller drives the additive module (3) to move according to the defect position coordinate information input in step 4, and the additive module (3) itself evenly melts the flux-cored welding wire to each defect position on the flange surface to be repaired according to the parameters preset by the ARM controller; After step 6 and step 5 are completed, the ARM controller starts the turning module (4), and the turning module (4) corrects the flange surface to be repaired according to the spiral line based on the parameters preset by the ARM controller.

8. The repair method according to claim 7, characterized in that: The ARM controller includes a data input and processing module, a CNN model module, a feedback and monitoring module, and a linkage control module. The data input and processing module collects the 3D image data of the TOF industrial camera and performs denoising and filtering. Based on the processed 3D image data, the ICP point cloud registration algorithm is used to achieve registration through the error between point clouds, and a 3D reconstruction model of the flange cylinder to be repaired is constructed and output; The CNN model module receives the 3D reconstruction model of the flange cylinder to be repaired output by the data input and processing module, uses the convolution kernel to perform convolution operation on the 3D reconstruction model to extract feature patterns, and samples the feature patterns to output the defect position coordinates; The linkage control module reads the defect position coordinates output by the CNN model module, performs first-order position inverse solution and second-order velocity inverse solution, obtains the position, velocity and acceleration of each joint respectively, calculates the rotation angle of each joint motor through motion planning, and then drives each joint motor to rotate through motion execution, thereby controlling the movement of the six-axis collaborative robot arm, and the movement position of the six-axis collaborative robot arm is fed back to the monitoring module through the encoder input; The feedback and monitoring module inputs the position information of the additive welding gun (310) and the six-axis collaborative robotic arm (320) monitored in real time, compares it with the preset position information, adjusts the position of the additive welding gun (310) and the six-axis collaborative robotic arm (320) in real time, and outputs it to the additive welding gun (310) and the six-axis collaborative robotic arm (320).

Images