Laser cleaning robot trajectory planning method and system

Abstract

A laser cleaning robot trajectory planning method comprises the steps that large three-dimensional curved surface part image information is obtained, and point cloud data are obtained through splicing; using the point cloud data to extract track key points through a recursive plane fitting method; processing the track key points according to an optimal sequence to generate a track line; and the laser cleaning head device moves according to the track line, and robot track planning of curved surface part laser cleaning is completed. The problems that in the prior art, time and labor are wasted, the efficiency is low, the cleaning quality is not uniform, and the labor cost is high during cleaning of large three-dimensional curved surface parts are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of laser cleaning, and in particular to a trajectory planning method and device for laser cleaning. Background Art

[0002] Large three-dimensional curved surface parts are widely used parts, including aircraft fuselages, armored vehicle shells, ship hulls, etc. After being used for a period of time, large three-dimensional curved surface parts often need to be rusted and painted to facilitate repainting. The traditional method of rust and paint removal is mainly for workers to hold relevant cleaning equipment and manually clean based on visual observation. This method has the disadvantages of being time-consuming and labor-intensive, inefficient, uneven cleaning quality, and high labor costs. In contrast, laser cleaning is a cleaning method that uses lasers to remove rust and paint from the surface of parts. It has the advantages of high efficiency, good results, environmental protection, and convenience for automation. For large curved surface parts, traditional laser cleaning equipment is mounted on robots or machine tool platforms, and requires manual teaching and other methods to complete trajectory planning, which is inefficient. Therefore, there is an urgent need for a trajectory planning method and device for laser cleaning of large three-dimensional curved surface parts to solve the problems of the existing technology. Summary of the Invention

[0003] In view of the above problems, the present invention is proposed to provide a robot trajectory planning method and system for laser cleaning that overcomes the above problems or at least partially solves the above problems.

[0004] In order to solve the above technical problems, the embodiments of the present application disclose the following technical solutions:

[0005] In a first aspect, an embodiment of the present invention discloses a robot trajectory planning method for laser cleaning, comprising:

[0006] S100. Acquire image information of large-scale three-dimensional curved surface parts and obtain point cloud data by stitching;

[0007] S200. Using the point cloud data, extracting trajectory key points by recursive plane fitting method;

[0008] S300. Processing the key points of the trajectory in an optimal order to generate a trajectory line;

[0009] S400. Control the laser cleaning head device to move along the trajectory line to complete the robot trajectory planning for laser cleaning of curved surface parts.

[0010] Furthermore, in S100, image information of a large three-dimensional curved surface part is obtained, and the specific steps include:

[0011] Based on the size of the part, determine N photo points on the part and set the pose of the photo points. The guide rail and robot drive the 3D structured light camera to the i-th position above the part. Use the 3D structured light camera to take a photo of the local range of the part to obtain the i-th local point cloud, 1≤i≤N. The photo taking process is repeated until the number of photo points reaches N.

[0012] Furthermore, point cloud data is obtained by stitching, and the specific steps include: rotating and translating all local point clouds according to the coordinates of the robot and the guide rail when the camera takes the picture, and stitching them into the same coordinate system to obtain the overall point cloud.

[0013] Furthermore, in S200, the point cloud data is used to extract trajectory key points through a recursive plane fitting method, and the specific steps include:

[0014] S201. Project the entire point cloud onto a horizontal plane and divide the point cloud into multiple strip point clouds using multiple parallel vertical planes according to the length of the laser beam;

[0015] S202. For a single strip point cloud, use the RANSAC plane fitting method to fit a plane and calculate the fitting error.

[0016] S203. For the planes fitted in S202, extract the center point of each plane and use the center point as the trajectory key point.

[0017] Furthermore, in S202, the fitting error is compared with the allowable value of defocus. If the fitting error is greater than the allowable value of defocus, the current point cloud is divided equally from the middle, and then the sub-point cloud is plane fitted, and the recursion is continuously performed until the fitting error of the plane fitted by each sub-point cloud is less than the allowable value of defocus; if the size of the fitting plane is less than the threshold, the recursion is stopped and returned to S201, the length of the laser light is adjusted, and the strip point cloud is redivided.

[0018] Furthermore, after obtaining the trajectory key points, interference check and singularity check will be performed on the trajectory key points. The specific steps include: for each trajectory key point, calculating the guide rail movement distance and the corresponding robot joint angle, performing interference check according to the CAD model, performing singularity check according to the robot's kinematic model, adjusting the joint angle of abnormal points, and optimizing the robot joint angle.

[0019] Furthermore, in S300, the trajectory key points are processed in the optimal order to generate a trajectory line. The specific method includes: for each strip point cloud, the trajectory key points are connected in order to form a broken line; for each broken line, the current broken endpoint is connected to the unconnected endpoint on the nearest other broken line to form a trajectory line that is connected end to end.

[0020] Furthermore, in S400, the laser cleaning head moves along the trajectory line to complete the robot trajectory planning for laser cleaning of curved parts. The specific steps include: the midpoint of the one-dimensional light emitted by the laser cleaning head is located at the key point of the trajectory, the axis of the cleaning head is parallel to the normal vector of the fitting plane corresponding to the key point of the trajectory, and the cleaning light is perpendicular to the forward direction of the trajectory.

[0021] Furthermore, the laser cleaning head device includes: a guide rail, a 6-axis robot, a 3D structured light camera, a laser cleaning head, and a tooling fixture: the 6-axis robot is installed on the guide rail, the 3D camera and the laser cleaning head are installed on the end flange of the robot, and the parts to be cleaned are fixed on the tooling fixture; the laser cleaning head generates one-dimensional laser light for cleaning through a galvanometer, and the light length can be adjusted.

[0022] In a second aspect, an embodiment of the present invention discloses a robot trajectory planning system for laser cleaning, comprising: a point cloud data acquisition unit, a trajectory key point extraction unit, a trajectory line generation unit, and a robot trajectory planning unit; wherein:

[0023] Point cloud data acquisition unit, used to obtain image information of large three-dimensional curved surface parts and obtain point cloud data by splicing;

[0024] A trajectory key point extraction unit, configured to extract trajectory key points using the point cloud data through a recursive plane fitting method;

[0025] A trajectory line generating unit, configured to process the trajectory key points in an optimal order to generate a trajectory line;

[0026] The robot trajectory planning unit is used to control the laser cleaning head device to move along the trajectory line to complete the robot trajectory planning for laser cleaning of curved surface parts.

[0027] The beneficial effects of the above technical solutions provided by the embodiments of the present invention include at least:

[0028] An embodiment of the present invention discloses a robot trajectory planning method for laser cleaning, comprising: acquiring image information of large three-dimensional curved surface parts and stitching it together to generate point cloud data; extracting key points of the trajectory using a recursive plane fitting method from the point cloud data; processing the key points in an optimal order to generate a trajectory line; and moving a laser cleaning head device along the trajectory line to complete the robot trajectory planning for laser cleaning of curved surface parts. This invention addresses existing issues in the cleaning of large three-dimensional curved surface parts, such as time-consuming and labor-intensive processes, low efficiency, uneven cleaning quality, and high labor costs.

[0029] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0031] Figure 1 This is a flow chart of a robot trajectory planning method for laser cleaning in Example 1 of the present invention;

[0032] Figure 2 This is a flow chart for obtaining image information of a large three-dimensional curved surface part in Example 1 of the present invention;

[0033] Figure 3 Schematic diagram of overall 3D point cloud division in Example 1 of the present invention;

[0034] Figure 4 This is a schematic diagram of the strip point cloud plane fitting and trajectory point extraction, and the connection trajectory in Example 1 of the present invention;

[0035] Figure 5 This is a schematic diagram of the overall trajectory point connection and the laser cleaning head posture in Example 1 of the present invention;

[0036] Figure 6 This is a structural schematic diagram of a laser cleaning head device in Example 1 of the present invention. DETAILED DESCRIPTION

[0037] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0038] In order to solve the problems existing in the prior art, an embodiment of the present invention provides a robot trajectory planning method and system for laser cleaning.

[0039] Example 1: The present invention discloses a robot trajectory planning method for laser cleaning, such as Figure 1 ,include:

[0040] S100. Acquire image information of a large three-dimensional curved surface part and obtain point cloud data by stitching. In S100 of the present embodiment, the image information of the large three-dimensional curved surface part is acquired. The specific steps include: determining N photographic points of the part according to the size of the part, and setting the positions of the photographic points; using a guide rail and a robot to drive a 3D structured light camera to move to the i-th position above the part, and using the 3D structured light camera to photograph a local range of the part to obtain the i-th local point cloud, 1≤i≤N; and looping the photographing until the number of photographic points reaches N.

[0041] Specifically, such as Figure 2 As shown in the figure, the process for acquiring image information of large 3D curved surface parts is as follows: First, the number of photographic points N is determined based on the part size, and the poses of the photographic points are set. Then, the 3D camera is moved by the guide rail and the robot to the i-th position above the part. The 3D structured light camera is used to photograph a local area of ​​the part, generating the i-th local point cloud. The photographic process repeats until the number of photographic points reaches N.

[0042] In some preferred embodiments, after obtaining the part image information, point cloud data is obtained by stitching. The specific steps include: rotating and translating all local point clouds according to the coordinates of the robot and the guide rail when the camera takes the picture, and stitching them into the same coordinate system to obtain the overall point cloud.

[0043] After acquiring part image information, obtaining point cloud data through stitching is a key step in 3D reconstruction and reverse engineering. The specific steps include:

[0044] Feature extraction and matching uses SIFT, SURF, or ORB algorithms to extract key points and their descriptors, generating local features. Preliminary matching pairs are screened using the nearest neighbor ratio test, and false matches are eliminated using the RANSAC algorithm. The geometric transformation matrix between viewpoints is estimated for precise matching. Based on the feature matching results, a dense depth map is generated using semi-global matching (SGM) or a depth map fusion algorithm, which then generates a single-view point cloud. This point cloud is then filtered and denoised. Through multi-view image acquisition, feature matching, point cloud generation, and stitching, 3D point cloud data for parts can be efficiently acquired.

[0045] S200. Utilize the point cloud data to extract trajectory keypoints using a recursive plane fitting method. The recursive plane fitting method is an effective method for extracting trajectory keypoints from point cloud data. By properly selecting algorithms and parameters, combined with appropriate post-processing steps, high-quality trajectory keypoints can be obtained.

[0046] In S200 of this embodiment, the point cloud data is used to extract trajectory key points through a recursive plane fitting method. The specific steps include:

[0047] S201. Project the entire point cloud onto a horizontal plane, and divide the point cloud into multiple strip point clouds using multiple parallel vertical planes according to the length of the laser beam; specifically, Figure 3 As shown in the figure, the entire point cloud is projected onto the horizontal plane, and according to the length of the laser light, several parallel vertical planes are used to divide the point cloud into several strip point clouds.

[0048] S202. For a single strip point cloud, use the RANSAC plane fitting method to fit the plane and obtain the fitting error; in S202 of this embodiment, the fitting error is compared with the allowable defocus value. If the fitting error is greater than the allowable defocus value, the current point cloud is divided equally from the middle, and then the sub-point cloud is plane fitted, and the recursion is continuously performed until the fitting error of the plane fitted by each sub-point cloud is less than the allowable defocus value; if the size of the fitted plane is less than the threshold, the recursion is stopped and the process returns to S201, the laser beam length is adjusted, and the strip point cloud is re-divided.

[0049] S203. For the planes fitted in S202, extract the center point of each plane and use the center point as the key point of the trajectory. Figure 4 As shown, for each fitted plane in the strip point cloud, the center point of each plane is extracted as the trajectory key point;

[0050] In some preferred embodiments, after the trajectory key points are obtained, interference check and singularity check will be performed on the trajectory key points. The specific steps include: for each trajectory key point, calculating the guide rail movement distance and the corresponding robot joint angle, performing interference check according to the CAD model, performing singularity check according to the robot's kinematic model, adjusting the joint angle of abnormal points, and optimizing the robot joint angle.

[0051] S300. Process the trajectory key points in the optimal order to generate a trajectory line; in S300 of this embodiment, the trajectory key points are processed in the optimal order to generate a trajectory line. The specific method includes: for each strip point cloud, connecting the trajectory key points in order to form a broken line; for each broken line, connecting the current broken endpoint with the unconnected endpoint on the nearest other broken line to form a trajectory line that is connected end to end.

[0052] Specifically, such as Figure 5 As shown in the figure, for each strip point cloud, the obtained trajectory key points are connected into a broken line in the optimal order; for each generated broken line, its endpoint is connected to the unconnected endpoint of the nearest other broken line to form a trajectory line that is connected end to end.

[0053] S400. Control the laser cleaning head device to move along the trajectory line to complete the robot trajectory planning for laser cleaning of curved parts. Figure 5 As shown, in S400 of this embodiment, the laser cleaning head moves along the trajectory line to complete the robot trajectory planning for laser cleaning of curved parts. The specific steps include: the midpoint of the one-dimensional light emitted by the laser cleaning head is located at the key point of the trajectory, the axis of the cleaning head is parallel to the normal vector of the fitting plane corresponding to the key point of the trajectory, and the cleaning light is perpendicular to the direction of advance of the trajectory.

[0054] In this embodiment, if Figure 6 The laser cleaning head device based on this trajectory planning method includes: 1 - a guide rail, 2 - a 6-axis robot, 3 - a 3D structured light camera, 4 - a laser cleaning head, and 5 - a fixture. The 6-axis robot is mounted on the guide rail, 3 - the 3D camera and 4 - the laser cleaning head are mounted on the end flange of the robot, and 6 - the part being cleaned is fixed to the fixture. 1 - the guide rail can move horizontally, driving the robot 2 in motion. 3 - the 3D structured light camera can capture a point cloud within a surface area at once, remaining stationary during capture.

[0055] 4- The laser cleaning head generates a one-dimensional laser beam through a galvanometer, and the length of the beam can be adjusted. 5- The fixture can fix the curved part in a specific posture, so that it maintains the same posture during the cleaning process.

[0056] This embodiment discloses a robot trajectory planning method for laser cleaning, comprising: acquiring image information of large three-dimensional curved surface parts and stitching it together to generate point cloud data; extracting key points of the trajectory using a recursive plane fitting method from the point cloud data; processing the key points in an optimal order to generate a trajectory line; and moving a laser cleaning head device along the trajectory line to complete the robot trajectory planning for laser cleaning of curved surface parts. This invention addresses the existing problems of large three-dimensional curved surface part cleaning, which are time-consuming and labor-intensive, inefficient, uneven cleaning quality, and high labor costs.

[0057] Example 2: Based on the same inventive concept, the present disclosure also provides a robot trajectory planning system for laser cleaning, comprising: a point cloud data acquisition unit, a trajectory key point extraction unit, a trajectory line generation unit, and a robot trajectory planning unit; wherein:

[0058] Point cloud data acquisition unit, used to obtain image information of large three-dimensional curved surface parts and obtain point cloud data by splicing;

[0059] A trajectory key point extraction unit, configured to extract trajectory key points using the point cloud data through a recursive plane fitting method;

[0060] A trajectory line generating unit, configured to process the trajectory key points in an optimal order to generate a trajectory line;

[0061] The robot trajectory planning unit is used to control the laser cleaning head device to move along the trajectory line to complete the robot trajectory planning for laser cleaning of curved surface parts.

[0062] The trajectory key point extraction unit uses the point cloud data to extract trajectory key points through a recursive plane fitting method, and the specific steps include:

[0063] S201. Project the entire point cloud onto a horizontal plane, and divide the point cloud into multiple strip point clouds using multiple parallel vertical planes according to the length of the laser beam; specifically, Figure 3 As shown in the figure, the entire point cloud is projected onto the horizontal plane, and according to the length of the laser light, several parallel vertical planes are used to divide the point cloud into several strip point clouds.

[0064] S202. For a single strip point cloud, use the RANSAC plane fitting method to fit the plane and obtain the fitting error; in S202 of this embodiment, the fitting error is compared with the allowable defocus value. If the fitting error is greater than the allowable defocus value, the current point cloud is divided equally from the middle, and then the sub-point cloud is plane fitted, and the recursion is continuously performed until the fitting error of the plane fitted by each sub-point cloud is less than the allowable defocus value; if the size of the fitted plane is less than the threshold, the recursion is stopped and the process returns to S201, the laser beam length is adjusted, and the strip point cloud is re-divided.

[0065] S203. For the planes fitted in S202, extract the center point of each plane and use the center point as the key point of the trajectory. Figure 4 As shown, for each fitted plane in the strip point cloud, the center point of each plane is extracted as the trajectory key point;

[0066] In some preferred embodiments, after the trajectory key points are obtained, interference check and singularity check will be performed on the trajectory key points. The specific steps include: for each trajectory key point, calculating the guide rail movement distance and the corresponding robot joint angle, performing interference check according to the CAD model, performing singularity check according to the robot's kinematic model, adjusting the joint angle of abnormal points, and optimizing the robot joint angle.

[0067] Among them, the specific working methods of the point cloud data acquisition unit, the trajectory key point extraction unit and the robot trajectory planning unit have been described in detail in Example 1, and will not be repeated here in this embodiment.

[0068] It should be understood that the specific order or hierarchy of steps in the disclosed processes is an example of an exemplary method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process can be rearranged without departing from the scope of the present disclosure. The accompanying method claims present elements of the various steps in an exemplary order and are not intended to be limited to the specific order or hierarchy described.

[0069] In the foregoing detailed description, various features are grouped together in a single embodiment to simplify the disclosure. This method of disclosure should not be interpreted as reflecting an intention that embodiments of the claimed subject matter require more features than are expressly recited in each claim. On the contrary, as reflected in the appended claims, the invention comprises less than all the features of any individual disclosed embodiment. The appended claims are therefore hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate preferred embodiment of the invention.

[0070] Those skilled in the art will also appreciate that the various illustrative logic blocks, modules, circuits, and algorithmic steps described in conjunction with the embodiments herein may be implemented as electronic hardware, computer software, or a combination thereof. In order to clearly illustrate the interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps described above are generally described around their functions. Whether such functions are implemented as hardware or software depends on the specific application and the design constraints imposed on the entire system. A skilled person may implement the described functions in an adaptable manner for each specific application, but such implementation decisions should not be interpreted as departing from the scope of protection of this disclosure.

[0071] The steps of the methods or algorithms described in conjunction with the embodiments herein may be directly embodied as hardware, software modules executed by a processor, or a combination thereof. The software module may be located in a RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, register, hard disk, removable disk, CD-ROM, or any other form of storage medium well known in the art. An exemplary storage medium is connected to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium may also be an integral part of the processor. The processor and storage medium may be located in an ASIC. The ASIC may be located in a user terminal. Of course, the processor and storage medium may also be present in a user terminal as discrete components.

[0072] For software implementation, the techniques described in this application can be implemented using modules (e.g., procedures, functions, etc.) that perform the functions described in this application. These software codes can be stored in a memory unit and executed by a processor. The memory unit can be implemented within the processor or external to the processor. In the latter case, it is communicatively coupled to the processor via various means, which are well known in the art.

[0073] The foregoing description includes examples of one or more embodiments. Of course, it is not possible to describe all possible combinations of components or methods for the purposes of describing the above embodiments, but one of ordinary skill in the art will recognize that the various embodiments may be further combined and arranged. Therefore, the embodiments described herein are intended to encompass all such changes, modifications and variations that fall within the scope of the appended claims. Furthermore, to the extent the term "comprising" is used in the specification or claims, the term is intended to be encompassed in a manner similar to the term "including," as explained in terms of "including," used as a transitional word in the claims. Furthermore, any use of the term "or" in the specification of the claims is intended to mean a "non-exclusive or."

Claims

1. A robot trajectory planning method for laser cleaning, characterized in that: include: S100. Acquire image information of large-scale three-dimensional curved surface parts and obtain point cloud data by stitching; S200. Using the point cloud data, extracting trajectory key points by recursive plane fitting method; S300. Processing the key points of the trajectory in an optimal order to generate a trajectory line; S400. Control the laser cleaning head device to move along the trajectory line to complete the robot trajectory planning for laser cleaning of curved surface parts.

2. The trajectory planning method according to claim 1, wherein: In S100, image information of a large three-dimensional curved surface part is obtained. The specific steps include: Based on the size of the part, determine N photo points on the part and set the pose of the photo points. The guide rail and robot drive the 3D structured light camera to the i-th position above the part. Use the 3D structured light camera to take a photo of the local range of the part to obtain the i-th local point cloud, 1≤i≤N. The photo taking process is repeated until the number of photo points reaches N.

3. The trajectory planning method according to claim 2, wherein: Point cloud data is obtained by stitching. The specific steps include: rotating and translating all local point clouds according to the coordinates of the robot and guide rail when the camera takes the picture, and stitching them into the same coordinate system to obtain the overall point cloud.

4. The trajectory planning method according to claim 1, wherein: In S200, the point cloud data is used to extract trajectory key points through a recursive plane fitting method. The specific steps include: S201. Project the entire point cloud onto a horizontal plane and divide the point cloud into multiple strip point clouds using multiple parallel vertical planes according to the length of the laser beam; S202. For a single strip point cloud, use the RANSAC plane fitting method to fit a plane and calculate the fitting error. S203. For the planes fitted in S202, extract the center point of each plane and use the center point as the trajectory key point.

5. The trajectory planning method according to claim 4, wherein: In S202, the fitting error is compared with the allowable value of defocus. If the fitting error is greater than the allowable value of defocus, the current point cloud is divided equally from the middle, and then the sub-point cloud is plane fitted, and the recursion is continuously performed until the fitting error of the plane fitted by each sub-point cloud is less than the allowable value of defocus; if the size of the fitting plane is less than the threshold, the recursion is stopped and returned to S201, the length of the laser light is adjusted, and the strip point cloud is re-divided.

6. The trajectory planning method according to claim 4, wherein: After obtaining the trajectory key points, interference check and singularity check will be performed on the trajectory key points. The specific steps include: for each trajectory key point, calculating the guide rail movement distance and the corresponding robot joint angle, performing interference check according to the CAD model, performing singularity check according to the robot's kinematic model, adjusting the joint angle of abnormal points, and optimizing the robot joint angle.

7. The trajectory planning method according to claim 1, wherein: In S300, the trajectory key points are processed in the optimal order to generate a trajectory line. The specific method includes: for each strip point cloud, the trajectory key points are connected in order to form a broken line; for each broken line, the current broken endpoint is connected to the unconnected endpoint on the nearest other broken line to form a trajectory line that is connected end to end.

8. The trajectory planning method according to claim 1, wherein: In S400, the laser cleaning head moves along the trajectory line to complete the robot trajectory planning for laser cleaning of curved parts. The specific steps include: the midpoint of the one-dimensional light emitted by the laser cleaning head is located at the key point of the trajectory, the axis of the cleaning head is parallel to the normal vector of the fitting plane corresponding to the key point of the trajectory, and the cleaning light is perpendicular to the direction of the trajectory.

9. The trajectory planning method according to claim 8, wherein: The laser cleaning head device includes: a guide rail, a 6-axis robot, a 3D structured light camera, a laser cleaning head, and a fixture: the 6-axis robot is installed on the guide rail, the 3D camera and the laser cleaning head are installed on the end flange of the robot, and the parts to be cleaned are fixed on the fixture; the laser cleaning head generates a one-dimensional laser light for cleaning through a galvanometer, and the light length can be adjusted.

10. A robot trajectory planning system for laser cleaning, using the trajectory planning method according to any one of claims 1 to 9, characterized in that: include: Point cloud data acquisition unit, trajectory key point extraction unit, trajectory line generation unit, robot trajectory planning unit; Among them: Point cloud data acquisition unit, used to obtain image information of large three-dimensional curved surface parts and obtain point cloud data by splicing; A trajectory key point extraction unit, configured to extract trajectory key points using the point cloud data through a recursive plane fitting method; A trajectory line generating unit, configured to process the trajectory key points in an optimal order to generate a trajectory line; The robot trajectory planning unit is used to move the laser cleaning head device according to the trajectory line to complete the robot trajectory planning for laser cleaning of curved surface parts.

Images