An automated grinding method for repairing defects of fan blades

Abstract

The present invention relates to an automated grinding method for repairing defects of fan blades, belonging to the technical field of fan repair, and comprising the following steps: a drone collects a point cloud map of the fan blade and transmits it to an industrial control computer; the industrial control computer reconstructs the fan blade model according to the point cloud map, marks the areas with a surface depth greater than 1 mm, and transmits them to a repair robot; the repair robot rises to the front of the blade of the wind turbine through a self-lifting device, adsorbs on the fan through a suction cup, and executes a pre-grinding detection program on the fan blade to determine the area and nature to be ground; the repair robot executes a grinding program on the area to be ground according to its nature; the repair robot executes a post-grinding detection program on the fan blade; the repair robot descends to the ground through a self-lifting device, ending the automated grinding work. The present invention can complete the grinding process of the blade in the air, has good safety, and ensures the accuracy of the grinding position of the robotic arm.

Description

Technical Field

[0001] The present invention belongs to the technical field of fan repair, and particularly relates to an automated grinding method for repairing defects of fan blades. Background Art

[0002] At present, the grinding work in the repair of fan blade defects depends entirely on manual labor. Before the operation, the blade needs to be removed from the main tower and transported to the ground. Then, tools such as a hoisting vehicle and ropes are used to send maintenance personnel to the maintenance area for operation. Considering safety, repair efficiency, and repair cost, the current manual repair is not the best choice. Moreover, the efficiency of manual grinding is low, time-consuming, and labor-intensive. A five-person maintenance team can usually only complete the grinding work of one fan in a day, and the entire operation process cannot meet the rapidly increasing business volume in the fan repair industry. Summary of the Invention

[0003] The purpose of the present invention is to solve the problems existing in the prior art, and provides an automated grinding method for fan blades. Through the cooperation of a repair robot and a drone, the grinding process of the blade is completed in the air without disassembling the blade, solving the safety hazard problem of manual high-altitude operation, effectively reducing the labor intensity of the staff, and improving work efficiency.

[0004] The technical solution is as follows:

[0005] An automated grinding method for repairing defects of fan blades includes a drone, a repair robot, and an industrial control computer that are communicatively connected to each other through a wireless network. The repair robot includes a self-lifting device, a carrying platform, a robotic arm, a tool library, a control module, and an energy power module. One end of the robotic arm is provided with a 3D line laser scanner, a camera, and a quick-change device. The other ends of the self-lifting device and the robotic arm are respectively installed at both ends above the carrying platform. The tool library, the control module, and the energy power module are respectively arranged in the middle above the carrying platform. At least four suction cups are arranged around the bottom of the carrying platform. The energy power module is electrically connected to the self-lifting device, the carrying platform, the robotic arm, and the control module. The control module includes a single-chip microcomputer and a database. The tool library is provided with a grinding component and a spray gun component with acetone. The upper ends of each component are respectively adapted to the quick-change device of the robotic arm through a circular ring member. And the following steps are executed:

[0006] Step S1: The drone flies along a preset path, connects the self-lifting device of the repair robot to the fan, and collects the point cloud map of the fan blade through the lidar carried by the fuselage and transmits it to the industrial control computer;

[0007] Step S2: The industrial control computer reconstructs the fan blade model according to the point cloud map, uses software to segment the reconstructed model, marks the area with a surface depth greater than 1 mm, and transmits the marked model diagram to the control module of the repair robot;

[0008] Step S3: The repair robot rises in front of the blade of the wind turbine through the self-lifting device, adsorbs on the wind turbine through the suction cup, and executes a pre-grinding detection program on the wind turbine blade to determine the area and nature to be ground;

[0009] Step S4: The repair robot executes a grinding program on the area to be ground according to its nature;

[0010] Step S5: The repair robot executes a post-grinding detection program on the wind turbine blade;

[0011] Step S6: The repair robot descends to the ground through the self-lifting device to end the automated grinding work.

[0012] Furthermore, the self-lifting device includes a guiding bracket, a driving motor, two rope winders, and three cables; the driving motor is installed in the middle of the two and is connected to the rope winder through a rotating shaft. One end of the two cables is wound in the two rope winders respectively, and the other end is provided with a hook. After passing through the ring fixed on the guiding bracket, it is fixed to the top of the wind turbine by a drone. By rotating the driving motor forward and backward, the rope winder winds and unwinds the cable to pull the repair robot up and down. The other cable is fixed to the ground at both ends after passing through the ring and pulley fixed on the guiding bracket.

[0013] Furthermore, the quick-change device is an electric control component with a convex cross-section. Two or more steel balls are arranged on the side of the protruding part. The rear part of the steel ball is a telescopic structure and can freely expand and contract under the control of the quick-change device; holes are opened at corresponding positions on the inner side of the circular ring components on each component. When the steel ball extends, it is stuck into the hole, so that the protruding part of the quick-change device is tightly connected to the circular ring.

[0014] Furthermore, the pre-grinding detection program in Step S3 includes the following steps:

[0015] Step S31: Take an image of the wind turbine blade through the camera on the robotic arm and save it to the database of the control module;

[0016] Step S32: Perform binaryzation and filtering operations on the image in the database through software to sharpen the information such as stripes and particles in the image;

[0017] Step S33: Statistically calculate the total area of the information such as stripes and particles, compare the difference with the model diagram transmitted from the industrial computer, and determine the area with a difference area greater than twenty square millimeters as the area to be ground, and mark the numbers in sequence;

[0018] Step S34: Detect the defect depth of the area on the blade surface that needs to be polished by the 3D line laser scanner installed at the end of the robotic arm to determine the nature of the area to be polished. If the depth is less than the set value of 3 mm, it is set as a normal polishing area; if the depth is greater than the set value of 3 mm, it is set as a stepped polishing area.

[0019] Furthermore, the polishing procedure in step S4 includes the following steps:

[0020] Step S41: The robotic arm moves to the tool library and connects the polishing component through the quick-change device;

[0021] Step S42: The robotic arm moves in front of the wind turbine blade and uses the polishing component to perform normal polishing or stepped polishing on the marked area in sequence according to the number and according to the polishing area settings. The polishing time is 1 minute;

[0022] Step S43: After the robotic arm moves the polishing component back to its original position in the tool library, disconnect the connection and reconnect the spray gun component in the tool library;

[0023] Step S44: The robotic arm moves in front of the wind turbine blade and uses the spray gun component to spray acetone on the polishing area in step S42 for cleaning in sequence according to the number;

[0024] Step S45: After the robotic arm moves the spray gun component back to its original position in the tool library, disconnect the connection.

[0025] Furthermore, the so-called normal polishing in step S42 means that during polishing, the same size is repeatedly polished on each layer of fiberglass cloth according to a preset trajectory.

[0026] Furthermore, the so-called stepped polishing in step S42 means that when polishing the fiberglass cloth layer, the polishing sizes in the axial and chordal directions of each layer of fiberglass cloth are different. The axial single-sided step polishing size is 50 mm - 100 mm, the chordal single-sided step size is 0 - 50 mm, and it is polished in a stepped shape that shrinks from the outside to the inside.

[0027] Furthermore, in step S5, the detection procedure after polishing includes the following steps:

[0028] Step S51: Use three-dimensional laser scanning technology to detect the defect depth of the surface of the polished area. If it is less than the set value of 1 mm, it is considered to pass the polishing depth detection and enter step S52; otherwise, return to step S4;

[0029] Step S52: Take a picture of the polished area through the camera on the robotic arm and save it to the database of the control module;

[0030] Step S53: Perform binarization and enhanced filtering operations on the images in the database through software to sharpen the information such as stripes and particles in the images;

[0031] Step S54: Statistically calculate the total area of information such as stripes and particles, compare the difference with the model diagram transmitted by the industrial control computer, determine that the grinding is completed if the difference area is less than four square millimeters, and enter Step S6; if not satisfied, return to Step S4.

[0032] Beneficial effects:

[0033] 1) Through the cooperation of the repair robot and the drone, the present invention completes the grinding process of the blade in the air without disassembling the blade, does not require manual operation, has good safety, and is cost-effective in the long run.

[0034] 2) The robot ensures the accuracy of the grinding position of the robotic arm through the comparison and detection of the pictures taken by the camera and the blade modeling.

[0035] 3) The 3D line laser scanner is used to detect the defect depth so that the robot can intelligently select ordinary grinding or staggered grinding. Description of the drawings

[0036] Figure 1 is the automated grinding flow chart of the present invention;

[0037] Figure 2 is the structural schematic diagram of the repair robot;

[0038] Figure 3 is the structural schematic diagram of the quick-change device and the ring member;

[0039] Figure 4 is the schematic diagram of the robotic arm connected to the grinding assembly through the quick-change device.

[0040] Among them: 1 is the robotic arm, 11 is the camera, 12 is the quick-change device, 121 is the steel ball, 13 is the ring member, 131 is the hole, 2 is the bearing platform, 3 is the lifting device, 31 is the guiding bracket, 32 is the driving motor, 33 is the rope reel, 34 is the cable, 4 is the tool library, 5 is the control module, and 6 is the energy power module. Detailed implementation manners

[0041] In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the drawings and embodiments:

[0042] As Figure 1 and Figure 2An automated grinding method for repairing defects of fan blades is shown, which includes a drone, a repair robot, and an industrial control computer that communicate with each other through a wireless network. The repair robot includes a self-lifting device 3, a carrying platform 2, a robotic arm 1, a tool library 4, a control module 5, and an energy and power module 6. One end of the robotic arm is provided with a 3D line laser scanner, a camera 11, and a quick-change device 12. The other ends of the self-lifting device and the robotic arm are respectively installed at both ends above the carrying platform. The tool library, the control module, and the energy and power module are respectively arranged in the middle above the carrying platform. At least four suction cups are arranged around the bottom of the carrying platform. The energy and power module is electrically connected to the self-lifting device, the carrying platform, the robotic arm, and the control module. The control module includes a single-chip microcomputer and a database. The tool library is provided with a grinding component and a spray gun component with acetone. The upper ends of each component are adapted to the quick-change device of the robotic arm through a circular ring member 13. And the following steps are executed:

[0043] Step S1: The drone flies along a preset path, connects the self-lifting device of the repair robot to the fan, and collects the point cloud map of the fan blade through the lidar carried by the fuselage and transmits it to the industrial control computer;

[0044] Step S2: The industrial control computer reconstructs the fan blade model according to the point cloud map, uses software to segment the reconstructed model, marks the areas with a surface depth greater than 1 mm, and transmits the marked model diagram to the control module of the repair robot;

[0045] Step S3: The repair robot rises to the front of the blade of the wind turbine through the self-lifting device, adsorbs on the fan through the suction cup, and executes a pre-grinding detection program on the fan blade to determine the area and nature to be ground; Step S4: The repair robot executes a grinding program on the area to be ground according to its nature;

[0046] Step S5: The repair robot executes a post-grinding detection program on the fan blade;

[0047] Step S6: The repair robot descends to the ground through the self-lifting device to end the automated grinding work.

[0048] The self-lifting device includes a guiding bracket 31, a driving motor 32, two rope winders 33, and three cables 34. The driving motor is installed in the middle of the two and is connected to the rope winder through a rotating shaft. One ends of two cables are respectively wound in the two rope winders, and the other ends are provided with hooks. After passing through the rings fixed on the guiding bracket, they are fixed to the top of the fan through the drone. By rotating the driving motor forward and backward, the rope winder winds and unwinds the cables to pull the repair robot up and down. The other cable is fixed at both ends on the ground after passing through the ring and pulley fixed on the guiding bracket.

[0049] Such as Figure 3As shown in the figure, the quick-change device is an electronically controlled component with a convex-shaped cross-section. Two or more steel balls 121 are provided on the side of the protruding part. The rear part of the steel ball is a telescopic structure, which can freely expand and contract under the control of the quick-change device. Corresponding holes 131 are opened at the inner side of the circular ring components on each component. When the steel balls protrude, they are stuck into the holes, so that the protruding part of the quick-change device is tightly connected to the circular ring.

[0050] The detection program before grinding in step S3 includes the following steps:

[0051] Step S31: Take an image of the fan blade through the camera on the robotic arm and save it to the database of the control module;

[0052] Step S32: Perform binarization and filtering operations on the image in the database through software to sharpen the information such as stripes and particles in the image;

[0053] Step S33: Calculate the total area of the information such as stripes and particles, compare the difference with the model diagram transmitted from the industrial computer, determine the area that needs to be ground if the difference area is greater than 20 square millimeters, and mark the numbers in sequence;

[0054] Step S34: Detect the defect depth of the area that needs to be ground on the blade surface through the 3D line laser scanner installed at the end of the robotic arm to determine the nature of the area that needs to be ground. If the depth is less than the set value of 3 millimeters, it is set as a normal grinding area. If the depth is greater than the set value of 3 millimeters, it is set as a staggered grinding area.

[0055] The grinding program in step S4 includes the following steps:

[0056] Step S41: The robotic arm moves to the tool library and connects to the grinding component as shown in the figure through the quick-change device; Figure 4 Step S42: The robotic arm moves to the front of the fan blade, and uses the grinding component to perform normal grinding or staggered grinding on the marked area according to the number sequence and the grinding area setting, and the grinding time is 1 minute;

[0057] Step S43: After the robotic arm moves the grinding component to the original position in the tool library, disconnect the connection, and reconnect the spray gun component in the tool library;

[0058] Step S44: The robotic arm moves to the front of the fan blade, and uses the spray gun component to spray acetone on the grinding area in step S42 according to the number sequence for cleaning;

[0059] Step S45: After the robotic arm moves the spray gun component to the original position in the tool library, disconnect the connection.

[0060] The normal grinding mentioned in step S42 means that during grinding, the fiberglass cloth of each layer is repeatedly ground according to the same size along the preset trajectory.

[0061] The stepped grinding described in step S42 means that when grinding the fiberglass cloth layer, the grinding dimensions of the axial and chordal directions of each layer of fiberglass cloth are different. The single-sided stepped grinding dimension in the axial direction is 50 mm - 100 mm, and the single-sided stepped dimension in the chordal direction is 0 - 50 mm, and the grinding is in a stepped shape that shrinks from the outside to the inside.

[0062] Further, in step S5, the inspection procedure after grinding includes the following steps:

[0063] Step S51: Use three-dimensional laser scanning technology to detect the defect depth on the surface of the ground area. If it is less than the set value of one millimeter, it is considered to pass the grinding depth inspection and enter step S52; otherwise, return to step S4.

[0064] Step S52: Take an image of the ground area through the camera on the robotic arm and save it to the database of the control module.

[0065] Step S53: Perform binarization and enhanced filtering operations on the images in the database through software to sharpen information such as stripes and particles in the images.

[0066] Step S54: Statistically calculate the total area of information such as stripes and particles, compare the difference with the model diagram transmitted from the industrial control computer, and determine that the grinding is completed if the difference area is less than four square millimeters and enter step S6; if not satisfied, return to step S4.

[0067] Embodiment: When it is decided to use the grinding method to repair the defects of the fan blade, first fix one cable of the self-lifting device of the repair robot to the ground, then set the flight path of the UAV. When passing over the top of the fan, fix the hooks at one end of the other two cables of the self-lifting device to the top of the fan. Before flying to the fan blade, collect the point cloud map of the fan blade through the lidar carried by the fuselage and transmit it to the industrial control computer; reconstruct the fan blade model by the industrial control computer according to the point cloud map, use the LiDAR360 software to segment the reconstructed model, mark the areas with a surface depth greater than 1 mm, and transmit the marked model diagram to the control module of the repair robot; the repair robot drives the rope take-up device to tighten the cable and rise to the front of the blade of the wind turbine by the positive rotation of the drive motor of the self-lifting device, and adsorbs on the fan body or blade through the suction cup, and executes the pre-grinding detection program for the fan blade, that is, takes pictures of the fan blade through the camera on the robotic arm and saves them to the database of the control module. Then, perform binary and enhanced filtering operations on the images in the database through the halcon software to sharpen the information such as stripes and particles in the images, calculate the total area of the information such as stripes and particles, compare the difference with the model diagram transmitted by the industrial control computer, determine that the area with a difference area greater than twenty square millimeters needs to be ground, and mark and number it in order. Finally, detect the defect depth of the area to be ground on the blade surface by the 3D line laser scanner installed at the end of the robotic arm to determine the nature of the area to be ground. If the depth is less than the set value of 3 mm, it is set as a normal grinding area. If the depth is greater than the set value of 3 mm, it is set as a stepped grinding area; the repair robot executes the grinding program for the determined area to be ground, specifically including moving the robotic arm to the tool library, connecting the grinding component through the quick-change device, using the grinding component to perform normal grinding or stepped grinding on the marked area according to the number order and the grinding area setting, with a grinding time of 1 minute, then replacing the grinding component with the spray gun component in the tool library, and using the spray gun component to spray acetone on the ground area for cleaning according to the number order. Finally, the robotic arm moves the spray gun component to the original position in the tool library and disconnects; where normal grinding means repeating the grinding with the same size according to the preset trajectory on each layer of fiberglass cloth during grinding, and stepped grinding means that when grinding the fiberglass cloth layer, the grinding sizes in the axial and chordal directions of each layer of fiberglass cloth are different, the axial single-sided step grinding size is 50 mm - 100 mm, the chordal single-sided step size is 0 - 50 mm, and it is ground in a stepped shape that shrinks from the outside to the inside;The repair robot executes the inspection program after grinding the fan blade. Specifically, it uses three-dimensional laser scanning technology to detect the defect depth on the surface of the ground area. If it is greater than the set value of one millimeter, it is regarded as failing the grinding depth inspection and needs to be reground. Otherwise, it takes a picture of the ground area through the camera on the robotic arm, saves it to the database of the control module, and then performs binarization and enhanced filtering operations on the image in the database through Halcon software to sharpen the information such as stripes and particles in the image, calculates the total area of the information such as stripes and particles, compares the difference with the model diagram transmitted from the industrial computer, determines that regrinding is required if the difference area is greater than four square millimeters, and determines that the grinding is completed otherwise. The repair robot drives the rope take-up device to relax the cable and descend to the ground by the reverse rotation of the drive motor of the self-lifting device, ending the entire automated grinding work.;

[0068] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the principle and spirit of the present invention shall be included in the protection scope of the present invention.

Claims

1. An automated grinding method for repairing defects of a fan blade, characterized in that: It includes a drone, a repair robot, and an industrial control computer that communicate with each other through a wireless network. The repair robot includes a self-lifting device (3), a carrying platform (2), a robotic arm (1), a tool library (4), a control module (5), and an energy and power module (6); one end of the robotic arm is provided with a 3D line laser scanner, a camera (11), and a quick-change device (12). The other ends of the self-lifting device and the robotic arm are respectively installed at both ends above the carrying platform. The tool library, the control module, and the energy and power module are respectively arranged in the middle above the carrying platform. At least four suction cups are arranged around the bottom of the carrying platform. The energy and power module is electrically connected to the self-lifting device, the carrying platform, the robotic arm, and the control module. The control module includes a single-chip microcomputer and a database; the tool library is provided with a grinding component and a spray gun component with acetone. The upper ends of the grinding component and the spray gun component are respectively adapted to the quick-change device of the robotic arm through a circular member (13); and the following steps are executed: Step S1: The drone flies along a preset path, connects the self-lifting device of the repair robot to the fan, and collects the point cloud map of the fan blade through the lidar carried by the fuselage and transmits it to the industrial control computer; Step S2: The industrial control computer reconstructs the fan blade model according to the point cloud map, uses software to segment the reconstructed model, marks the areas with a surface depth greater than 1 mm, and transmits the marked model diagram to the control module of the repair robot; Step S3: The repair robot rises to the front of the wind turbine blade through the self-lifting device, adsorbs on the fan through the suction cup, and executes a pre-grinding detection program on the fan blade to determine the area and nature to be ground; Step S4: The repair robot executes a grinding program on the area to be ground according to its nature; Step S5: The repair robot executes a post-grinding detection program on the fan blade; Step S6: The repair robot descends to the ground through the self-lifting device to end the automated grinding work; The pre-grinding detection program includes the following steps: Step S31: Take an image of the fan blade through the camera on the robotic arm and save it to the database of the control module; Step S32: Perform binarization and filtering operations on the image in the database through software to sharpen the stripe and particle information in the image; Step S33: Statistically calculate the total area of the stripe and particle information, compare the difference with the model diagram transmitted by the industrial control computer, determine the area to be ground when the difference area is greater than twenty square millimeters, and mark and number it in sequence; Step S34: Detect the defect depth of the area to be ground on the blade surface through the 3D line laser scanner installed at the end of the robotic arm to determine the nature of the area to be ground. If the depth is less than the set value of 3 mm, it is set as a normal grinding area. If the depth is greater than the set value of 3 mm, it is set as a staggered grinding area.

2. The automated grinding method for repairing defects of a fan blade according to claim 1, characterized in that: The described self-lifting device includes a guiding bracket (31), a driving motor (32), two rope winders (33) and three cables (34); the driving motor is installed between the two rope winders and connected to the rope winders through a rotating shaft. One end of two cables is respectively wound in the two rope winders, and the other end is provided with a hook. After passing through a ring fixed on the guiding bracket, it is fixed to the top of the wind turbine by a drone. By the forward and reverse rotation of the driving motor, the rope winders wind and unwind the cables to pull the repair robot up and down. The other cable is fixed to the ground at both ends after passing through a ring and a pulley fixed on the guiding bracket.

3. An automated grinding method for repairing defects of a fan blade according to claim 1, characterized in that: The quick-change device is an electrically controlled component with a convex cross-section. Two or more steel balls (121) are arranged on the side of the protruding part. The rear part of the steel ball is a telescopic structure and can freely expand and contract under the control of the quick-change device. Corresponding holes (131) are opened at the inner side of the circular ring components on the grinding assembly and the spray gun assembly. When the steel balls protrude, they are clamped into the holes, so that the protruding part of the quick-change device is tightly connected to the circular ring.

4. The automated grinding method for repairing defects of a fan blade according to claim 1, wherein The grinding procedure in step S4 includes the following steps: Step S41: The robotic arm moves to the tool library and connects to the grinding assembly through the quick-change device. Step S42: The robotic arm moves to in front of the wind turbine blade. The grinding assembly is used to perform ordinary grinding or staggered grinding on the marked area according to the number sequence and the grinding area setting. The grinding time is 1 minute. Step S43: After the robotic arm moves the grinding assembly back to the original position in the tool library, the connection is released, and the spray gun assembly in the tool library is reconnected. Step S44: The robotic arm moves to in front of the wind turbine blade. The spray gun assembly is used to spray acetone on the grinding area in step S42 according to the number sequence for cleaning. Step S45: After the robotic arm moves the spray gun assembly back to the original position in the tool library, the connection is released.

5. The automated grinding method for repairing defects of a fan blade according to claim 4, characterized in that: The ordinary grinding described in step S42 means that during grinding, the same size is repeatedly ground on each layer of fiberglass cloth according to a preset trajectory.

6. An automated grinding method for repairing defects of a fan blade according to claim 4, characterized in that: The staggered grinding described in step S42 means that when grinding the fiberglass cloth layer, the grinding sizes in the axial and chordal directions of each layer of fiberglass cloth are different. The unilateral step grinding size in the axial direction is 50 mm - 100 mm, and the unilateral step size in the chordal direction is 0 - 50 mm, and it is ground in a stepped shape that shrinks from the outside to the inside.

7. An automated grinding method for repairing defects of a fan blade according to claim 1, characterized in that The inspection procedure after grinding in step S5 includes the following steps: Step S51: Use three-dimensional laser scanning technology to detect the defect depth on the surface of the ground area. If it is less than the set value of 1 mm, it is regarded as passing the grinding depth inspection, and step S52 is entered; otherwise, return to step S4. Step S52: Take a picture of the ground area through the camera on the robotic arm and save it to the database of the control module. Step S53: Perform binarization and add filtering operations on the images in the database through software to sharpen the stripe and particle information in the images. Step S54: Count the total area of the stripe and particle information, compare the difference with the model diagram transmitted by the industrial control computer, and if the difference area is less than 4 square millimeters, it is determined that the grinding is completed, and step S6 is entered; if not satisfied, return to step S4.

Images