Mechanical arm curved surface scanning path planning method based on least square conformal algorithm

Abstract

The invention provides a mechanical arm curved surface scanning path planning method based on a least square conformal algorithm, and the method comprises the steps: carrying out the digital-to-analog conversion of a to-be-scanned curved surface into a triangular mesh, and calculating the curvature and normal vector of the vertex of the triangular mesh; performing point clustering according to the curvature and the normal vector of the vertex to obtain a point set of a plurality of areas capable of being continuously scanned; projecting the three-dimensional vertex to a two-dimensional plane through a least square conformal algorithm; solving a two-dimensional scanning path point corresponding to a point in the point set of each continuous scanning area; the adjacent scanning areas are connected end to end according to the nearest neighbor relation, and a complete two-dimensional scanning path point array is obtained; performing inverse mapping back to three dimensions, and offsetting the path points according to the normal directions of the path points and the parameters of the scanner to obtain a complete three-dimensional scanning path point array; and finally converting the data into executable data of the mechanical arm to complete path planning. According to the method provided by the invention, the curved surface scanning path which is efficient, comprehensive in coverage and high in adaptability and can be executed by the mechanical arm is generated.

Description

Technical Field

[0001] The invention relates to the technical field of automated digital measurement, and in particular to a robot arm curved surface scanning path planning method based on a least squares angle-preserving algorithm. Background Art

[0002] With the development of digital measurement technology and its application, 3D scanning technology, which is represented by the rapid acquisition of 3D information on the surface of objects, has been widely used in aerospace, automobile manufacturing, shipbuilding and other fields. The 3D scanning technology can obtain the 3D position information of the product surface, so that the product detection is transformed from the qualitative detection of traditional analog quantity to the quantitative detection of digital quantity. It has been widely used in the detection links such as product appearance detection, dimension measurement and reverse modeling. Handheld 3D scanners are commonly used to obtain product detection data because of their portability and easy operation. In order to solve the problems of low scanning efficiency and high labor intensity of manual handheld devices, more and more automated equipment is used in product 3D scanning and detection tasks, especially industrial robots (mechanical arms) are widely used because of their high flexibility, controllability and strong operability.

[0003] The main scanning path planning method at present is to scan the product by manually planning path points or manually teaching the robot arm to drive it. However, for large surfaces, manual path planning is inefficient and the teaching process is complicated. The generated path often cannot optimize the motion trajectory of the robot arm, which limits the scanning quality and efficiency. Although the existing automated scanning path planning methods can reduce manual participation to a certain extent, most of them rely on regularized or a priori defined scanning methods, such as regular path generation methods based on simple plane projection, cylindrical projection, etc. These methods have poor scanning coverage and continuity for complex surfaces. In addition, existing methods often ignore the parameter constraints of the scanning equipment (such as the limitations of the scanning angle and scanning distance) and the local geometric characteristics of the surface (such as the impact of curvature changes on the scanning path), making it difficult to obtain high-quality scanning data. Therefore, there is an urgent need for a method that can solve the above-mentioned deficiencies in the prior art. Summary of the invention

[0004] In order to solve the shortcomings of existing methods in large-scale curved surface scanning path planning, this paper proposes a robotic arm curved surface scanning path planning method based on the least squares angle-preserving algorithm. By combining the geometric characteristics of the surface, the three-dimensional surface is projected onto a two-dimensional plane, the optimized scanning path is planned in two-dimensional space, and then the path points are inversely mapped back to three-dimensional space, while considering the angle and distance constraints of the scanning device, so as to generate an efficient, comprehensive and adaptable scanning path that can be executed by the robotic arm.

[0005] In order to achieve the above technical objectives, the present invention provides the following technical solutions: A method for planning a curved scanning path for a robotic arm based on a least squares conformal algorithm specifically comprises the following steps: S1. Convert the design digital model of the surface to be scanned into a triangular mesh model to obtain the vertex set of the triangular mesh model of the surface to be scanned. V and the vertex triangle set T , and calculate the normal vector and curvature of the vertex; S2, clustering the continuously scannable triangular mesh vertices according to the curvature and normal vector calculated in step S1 to obtain a plurality of different continuously scannable area point sets; S3, projecting the vertices of the triangular network model from three dimensions to a two-dimensional plane using a least squares conformal algorithm; S4, determining corresponding points of the points in each continuously scannable area point set on the two-dimensional plane, and then obtaining a two-dimensional scanning path point array through the corresponding points on the two-dimensional plane; S5, connecting adjacent scanning areas end to end according to the nearest neighbor relationship to obtain a complete two-dimensional scanning path point array for the entire surface to be scanned; S6, inversely mapping the complete two-dimensional scanning path points obtained in step S5 back to three-dimensional space, and offsetting the path points according to the path point normal and the scanner parameters to obtain a complete three-dimensional scanning path point array; S7, converting the complete three-dimensional scanning path point array into an executable scanning path point array of the robotic arm, and completing the robotic arm scanning path planning of the surface to be scanned.

[0006] Furthermore, step S1 specifically includes: S11. Import the design digital model of the surface to be scanned into the modeling software to extract the surface features and discretize it into a triangular mesh to obtain the vertex set of the triangular mesh model of the surface to be scanned. and the vertex triangle set ; Each triangle in ;in, Vertex Set The index of the vertex in ; S12. For each vertex , select its neighborhood radius The neighboring points within the local plane are used to calculate their normal vectors. ; S13. For each vertex , select its neighborhood radius The curvature of the neighboring points within is calculated using a method based on local surface fitting. .

[0007] Furthermore, step S2 specifically includes: S21, the vertex set The points in are sorted in ascending order according to the curvature, and the vertex set is selected. The point with the smallest curvature As the starting point of clustering, let its curvature be , and its normal vector is ; S22, traversal Neighborhood radius , calculate its neighboring points k Neighboring points The normal vector With point The normal vector The angle between ,like Smaller than the set viewing cone angle , then and grouped into one category and and Mark as clustered; S23, when Judge as After clustering into one category, calculate Curvature With point Curvature The difference ,like Less than the set curvature change threshold , then Put in the cluster starting sequence middle; S24, traversal completed After the neighboring points of Reselect the point with the smallest curvature as the starting point of the new cluster, and repeat steps S22-S23 for the points that are not marked as clustered until the sequence is empty; points that are judged to be clustered together with the new cluster starting point are also considered as group into one category; And all its cluster points are put into the same point set, that is, a point set of a continuously scannable area is obtained. ; S25, set the vertex The points that are not marked as clustered in , for the set Repeat steps S21-S24 until the vertex set All points in are marked as clustered, and multiple different point sets of continuously scannable areas are obtained.

[0008] Furthermore, step S3 specifically includes: S31, traverse the vertex triangle set , find all the edges that belong to only one triangle, take out the vertices of these edges, and get the surface boundary line point set ; S32, projecting the boundary line points onto a unit circle in a two-dimensional plane; for a boundary line point , and its projected coordinates are ,in, , , Boundary point exist The index in is the total number of boundary line points; S33, define a triangle gradient matrix for describing the relationship between vertex coordinate changes and two-dimensional gradients , for any vertex triangle , its triangular gradient matrix is: ; in, is the area of ​​the triangle, , , , The triangle The three vertices of and The coordinates of the axes; S34. Constructing non-conformal energy function of triangle , for the triangle , assuming that the two-dimensional coordinates of its three vertices are , , , its energy function for: ; in, , ; S35. Accumulate the non-conformal energy of all triangles to obtain the global energy function ; S36. Expand the global energy function and express it as a linear system ; in, , is a The sparse coefficient matrix of is related to the triangular gradient matrix, The dimension introduced by using the coordinates of all boundary points as constraints is A vector of S37, use the sparse matrix solution method to solve x, so that the global energy function As the minimum value, solve the two-dimensional coordinates of each vertex , That is, the vertices in three-dimensional space Mapped to a point in two-dimensional space.

[0009] Furthermore, step S4 specifically includes: S41, for any continuously scannable area point set , after step S3, we get The corresponding points of all points in two-dimensional space constitute a two-dimensional continuous scanning area point set ; S42, using the convex hull algorithm to obtain the two-dimensional continuously scannable area point set Find the two points farthest from each other and , and find out and Line segment between two points ; S43, for the two-dimensional continuously scannable area point set Medium and Project the point outside the On, find The two points farthest apart in a direction and , find the straight line segment passing through these two points ; S44, according to the scanning width set by the scanner Straight line segment Divide the straight line segment into equal parts Divide into segment, of which Indicates rounding up. For straight line segment Length; S45, extract the straight line segment The equal nodes ; S46, through the straight line segment The nodes of the equal division Parallel line segments ,by As the starting line segment, The parallel line segments are connected in sequence, and the point set is obtained. Two-dimensional scanning path ; S47, along the two-dimensional scanning path Sampling is performed with a fixed sampling step size s, with a point set The point closest to the sampling point is the two-dimensional scanning path point, and the point set is obtained. A two-dimensional scanning path point array ; S48. Repeat steps S41 to S47 for other continuously scannable area point sets to obtain a two-dimensional scanning path point array for each continuously scannable area.

[0010] Furthermore, step S5 specifically includes: S51. Find the two-dimensional scanning path point array with the most two-dimensional scanning path points as well as The last waypoint in ,calculate Find the distance from the first path point in the remaining two-dimensional scan path point array The path point with the smallest distance , and its corresponding two-dimensional scanning path point array ; S52, will The path points in the original order are added to At the end, we get a new two-dimensional scanning path point array , and and Mark as merged; S53, according to step S52, in the remaining unmerged array, find the distance of the first path point The array closest to the last path point is merged and updated ; S54, repeat step S53 until all arrays are marked as merged, and obtain a complete two-dimensional scanning path point array of the surface to be scanned .

[0011] Further, step S6 specifically includes: S61. Find the complete two-dimensional path point array of the surface to be scanned according to the index correspondence The path points in the image correspond to the vertices of the triangular mesh model of the surface in three-dimensional space, and a complete three-dimensional path point array of the surface to be scanned is obtained. ; S62. For a three-dimensional path point array A waypoint in , move the point along its normal Optimal scanning distance for offset scanners , get the offset path point ; S63. For a three-dimensional path point array Step S62 is performed on all path points in the scanned surface to obtain a complete three-dimensional scanning path point array. .

[0012] Furthermore, step S7 specifically includes: S71. For a complete 3D scanning path point array Offset path points in , its normal vector As the Z axis of the scanner tool coordinate system, that is: ; S72, according to the scanner tool coordinate system The X-axis of the tool coordinate system is determined by the X-axis of the coordinate system of the surface to be scanned. First, a reference vector of the X-axis of the coordinate system of the surface to be scanned is selected. , calculate the temporary vector ,Will Normalize to get the X-axis of the tool coordinate system; S73, determining the Y axis of the tool coordinate system according to the X axis and the Z axis of the tool coordinate system; S74, the three axis vectors of the tool coordinate system form a rotation matrix ; S75. According to the rotation matrix Calculate the scanner at the offset path point RPY static angle at and ; S76, will Coordinates and in The RPY static angle of the scanner at and Combined into a robotic arm Executable points at ; S77, complete three-dimensional scanning path point array Execute steps S71 to S76 for all offset path points in the to-be-scanned surface to obtain an array of executable scanning path points for the robot arm. , complete the robotic arm scanning path planning of the surface to be scanned.

[0013] Based on the above technical solution, the present invention has the following beneficial effects: 1. The present invention uses the scanner cone angle as a constraint to segment and cluster the mesh vertices of the surface to be scanned, ensuring that the scanner can continuously scan the same scanning area without sudden posture changes, thereby reducing the noise and missing degree of the final scanning data.

[0014] 2. The present invention adopts a plane parameterization method based on least squares approximation quasi-conformal mapping to construct a mapping relationship between a three-dimensional mesh surface and a plane parameter domain. While ensuring that the flattened meshes do not overlap, the area of ​​the original mesh is maintained to the greatest extent possible, and the path planning of the mesh surface is transformed into path planning within the plane domain, which greatly reduces the difficulty of path planning for the three-dimensional space mesh surface.

[0015] 3. Convert the scanning path points into a data format that can be directly executed by the robot arm, avoiding multiple data conversions from planning to execution.

[0016] 4. The present invention provides an efficient robot arm path planning method for curved surface three-dimensional scanning while ensuring the quality reliability of curved surface three-dimensional scanning. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings: Figure 1 The present invention provides a flow chart of a method for planning a curved scanning path for a robotic arm based on a least squares conformal algorithm. DETAILED DESCRIPTION

[0018] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods, so that the implementation process of how the present application uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0019] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, so the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0020] Please refer to Figure 1 This embodiment proposes a robot surface scanning path planning method based on a least squares conformal algorithm, the method comprising the following steps: S1. Convert the design digital model of the surface to be scanned into a triangular mesh model to obtain the vertex set of the triangular mesh model of the surface to be scanned. V and the vertex triangle set T , and calculate the normal vector and curvature of the vertex; As a preferred implementation, step S1 specifically includes: S11. Import the design digital model of the surface to be scanned into the modeling software to extract the surface features and discretize it into a triangular mesh to obtain the vertex set of the triangular mesh model of the surface to be scanned. and the vertex triangle set ; Each triangle in ;in, Vertex Set The index of the vertex in ; S12. For each vertex , select its neighborhood radius The neighboring points within the local plane are used to calculate their normal vectors. ; S13. For each vertex , select its neighborhood radius The curvature of the neighboring points within is calculated using a method based on local surface fitting. .

[0021] S2, clustering the scannable points according to the curvature and normal vector calculated in step S1 to obtain multiple points sets in the continuously scannable area; As a preferred implementation, step S2 specifically includes: S21, the vertex set The points in are sorted in ascending order according to the curvature, and the vertex set is selected. The point with the smallest curvature As the starting point of clustering, let its curvature be , and its normal vector is ; S22, traversal Neighborhood radius , calculate its neighboring points k Neighboring points The normal vector With point The normal vector The angle between ,like Smaller than the set viewing cone angle , then and grouped into one category and and Marked as clustered; It should be noted that the cone angle is set in this step to consider the angle constraint of the scanning device; S23, when Judge as After clustering into one category, calculate Curvature With point Curvature The difference ,like Less than the set curvature change threshold , then Put in the cluster starting sequence middle; S24, traversal completed After the neighboring points of Reselect the point with the smallest curvature as the starting point of the new cluster, and repeat steps S22-S23 for the points that are not marked as clustered until the sequence is empty; points that are judged to be clustered together with the new cluster starting point are also considered as group into one category; And all its cluster points are put into the same point set, that is, a point set of a continuously scannable area is obtained. ; S25, set the vertex The points that are not marked as clustered in , for the set Repeat steps S21-S24 until the vertex set All points in are marked as clustered, and multiple different point sets of continuously scannable areas are obtained.

[0022] S3, projecting the vertices of the triangular network model from three dimensions to a two-dimensional plane using a least squares conformal algorithm; As a preferred implementation, step S3 specifically includes: S31, traverse the vertex triangle set , find all the edges that belong to only one triangle, take out the vertices of these edges, and get the surface boundary line point set ; S32, projecting the boundary line points onto a unit circle in a two-dimensional plane; for a boundary line point , and its projected coordinates are ,in, , , Boundary point exist The index in is the total number of boundary line points; S33, define a triangle gradient matrix for describing the relationship between vertex coordinate changes and two-dimensional gradients , for any vertex triangle , its triangular gradient matrix is: ; in, is the area of ​​the triangle, , , , The triangle The three vertices of and The coordinates of the axes; S34. Constructing non-conformal energy function of triangle , for the triangle , assuming that the two-dimensional coordinates of its three vertices are , , , its energy function for: ; in, , ; S35. Accumulate the non-conformal energy of all triangles to obtain the global energy function ; In this embodiment, in order to optimize the global conformality, the global energy function is minimized ,Right now: ; in , They are matrices composed of the coordinate components of all vertices of the triangular mesh in the two-dimensional coordinate system; S36. Expand the global energy function and express it as a linear system ; in, , is a The sparse coefficient matrix of is related to the triangular gradient matrix, assuming: For the matrix Middle OK The elements of the column are , The point indexed Connection is an edge in the triangular mesh model. , Yes The coordinates in the two-dimensional coordinate system are The expression is as follows: ; in express Is an included edge a triangle; if is not an edge in the triangular mesh model, then ; Representation Matrix For Elements Find partial differentials; and Points and The coordinates of the X and Y axes in three-dimensional space; The dimension introduced by using the coordinates of all boundary points as constraints is vector, assuming for No. r elements, The expression is as follows: ; S37, use the sparse matrix solution method to solve x, so that the global energy function is the minimum value, solve the two-dimensional coordinates of each vertex , That is, the vertices in three-dimensional space Mapped to a point in two-dimensional space.

[0023] S4, determining corresponding points of the points in each continuously scannable area point set on the two-dimensional plane, and then obtaining a two-dimensional scanning path point array through the corresponding points on the two-dimensional plane; As a preferred implementation, step S4 specifically includes: S41, for any continuously scannable area point set , after step S3, we get The corresponding points of all points in two-dimensional space constitute a two-dimensional continuous scanning area point set ; S42, using the convex hull algorithm to obtain the two-dimensional continuously scannable area point set Find the two points farthest from each other and , and find out and Line segment between two points ; S43, for the two-dimensional continuously scannable area point set Medium and Project the points outside the On, find The two points farthest apart in a direction and , find the straight line segment passing through these two points ; S44, according to the scanning width set by the scanner Straight line segment Divide the straight line segment into equal parts Divide into segment, of which Indicates rounding up. For straight line segment Length; S45, extract the straight line segment The equal nodes ; S46, through the straight line segment The nodes of the equal division Parallel line segments ,by As the starting line segment, The parallel line segments are connected in sequence, and the point set is obtained. Two-dimensional scanning path ; S47, along the two-dimensional scanning path Sampling is performed with a fixed sampling step size s, with a point set The point closest to the sampling point is the two-dimensional scanning path point, and the point set is obtained. A two-dimensional scanning path point array ; S48. Repeat steps S41 to S47 for other continuously scannable area point sets to obtain a two-dimensional scanning path point array for each continuously scannable area.

[0024] S5, connecting adjacent scanning areas end to end according to the nearest neighbor relationship to obtain a complete two-dimensional scanning path point array for the entire surface to be scanned; As a preferred implementation, step S5 specifically includes: S51. Find the two-dimensional scanning path point array with the most two-dimensional scanning path points as well as The last waypoint in ,calculate Find the distance from the first path point in the remaining two-dimensional scan path point array The path point with the smallest distance , and its corresponding two-dimensional scanning path point array ; S52, will The path points in the original order are added to At the end, we get a new two-dimensional scanning path point array , and and Mark as merged; S53, according to step S52, in the remaining unmerged array, find the distance of the first path point The array closest to the last path point is merged and updated ; S54, repeat step S53 until all arrays are marked as merged, and obtain a complete two-dimensional scanning path point array of the surface to be scanned .

[0025] S6, inversely mapping the complete two-dimensional scanning path points obtained in step S5 back to three-dimensional space, and offsetting the path points according to the path point normal and the scanner parameters to obtain a complete three-dimensional scanning path point array; As a preferred implementation, step S6 specifically includes: S61. Find the complete two-dimensional path point array of the surface to be scanned according to the index correspondence The path points in the image correspond to the vertices of the triangular mesh model of the surface in three-dimensional space, and a complete three-dimensional path point array of the surface to be scanned is obtained. ; S62. For a three-dimensional path point array A waypoint in , move the point along its normal Optimal scanning distance for offset scanners , get the offset path point ; It should be noted that, in this embodiment, It can be regarded as the depth of field of the scanner. The offset is to allow the surface to be scanned to be scanned within the depth of field of the scanner. That is, the distance constraint of the scanning device is taken into account in this step. Only after the offset processing can the three-dimensional scanning path points desired by the present invention be obtained.

[0026] S63. For a three-dimensional path point array Step S62 is performed on all path points in the scanned surface to obtain a complete three-dimensional scanning path point array. .

[0027] S7, converting the complete three-dimensional scanning path point array into an executable scanning path point array of the robot arm, and completing the robot arm scanning path planning of the surface to be scanned; As a preferred implementation, step S7 specifically includes: S71. For a complete 3D scanning path point array Offset path points in , its normal vector As the Z axis of the scanner tool coordinate system, that is: ; S72, according to the scanner tool coordinate system The X-axis of the tool coordinate system is determined by the X-axis of the coordinate system of the surface to be scanned. First, a reference vector of the X-axis of the coordinate system of the surface to be scanned is selected. , calculate the temporary vector ,Will Normalize to get the X-axis of the tool coordinate system; S73, determining the Y axis of the tool coordinate system according to the X axis and the Z axis of the tool coordinate system; S74, the three axis vectors of the tool coordinate system form a rotation matrix ; S75. According to the rotation matrix Calculate the scanner at the offset path point RPY static angle at and ; S76, will Coordinates and in The RPY static angle of the scanner at and Combined into a robotic arm Executable points at ; S77, complete three-dimensional scanning path point array Execute steps S71 to S76 for all offset path points in the to-be-scanned surface to obtain an array of executable scanning path points for the robot arm. , complete the robotic arm scanning path planning of the surface to be scanned.

[0028] In summary, the method proposed in the present invention solves the problems of low efficiency of manual path planning, complex teaching process, inability to optimize the motion trajectory of the robot arm, and poor scanning coverage and continuity of complex surfaces by other methods in the prior art. It achieves the generation of an efficient, comprehensive and adaptable surface scanning path that can be executed by the robot arm.

[0029] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0030] The above-mentioned embodiments only express several implementation modes of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A method for planning a curved scanning path for a robotic arm based on a least squares conformal algorithm, characterized in that: The specific steps include: S1. Convert the design digital model of the surface to be scanned into a triangular mesh model to obtain the vertex set of the triangular mesh model of the surface to be scanned. V and the vertex triangle set T , and calculate the normal vector and curvature of the vertex; S2, clustering the continuously scannable triangular mesh vertices according to the curvature and normal vector calculated in step S1 to obtain a plurality of different continuously scannable area point sets; S3, projecting the vertices of the triangular network model from three dimensions to a two-dimensional plane using a least squares conformal algorithm; S4, determining corresponding points of the points in each continuously scannable area point set on the two-dimensional plane, and then obtaining a two-dimensional scanning path point array through the corresponding points on the two-dimensional plane; S5, connecting adjacent scanning areas end to end according to the nearest neighbor relationship to obtain a complete two-dimensional scanning path point array for the entire surface to be scanned; S6, inversely mapping the complete two-dimensional scanning path points obtained in step S5 back to three-dimensional space, and offsetting the path points according to the path point normal and the scanner parameters to obtain a complete three-dimensional scanning path point array; S7, converting the complete three-dimensional scanning path point array into an executable scanning path point array of the robotic arm, and completing the robotic arm scanning path planning of the surface to be scanned.

2. The method for planning a curved surface scanning path of a robot arm according to claim 1, characterized in that: Step S1 specifically includes: S11. Import the design digital model of the surface to be scanned into the modeling software to extract the surface features and discretize it into a triangular mesh to obtain the vertex set of the triangular mesh model of the surface to be scanned. and the vertex triangle set ; Each triangle in ;in, Vertex Set The index of the vertex in ; S12. For each vertex , select its neighborhood radius The neighboring points within the local plane are used to calculate their normal vectors. ; S13. For each vertex , select its neighborhood radius The curvature of the neighboring points within is calculated using a method based on local surface fitting. .

3. The method for planning a curved surface scanning path of a robot arm according to claim 1, characterized in that: Step S2 specifically includes: S21, the vertex set The points in are sorted in ascending order according to the curvature, and the vertex set is selected. The point with the smallest curvature As the starting point of clustering, let its curvature be , and its normal vector is ; S22, traversal Neighborhood radius , calculate its neighboring points k Neighboring points The normal vector With point The normal vector The angle between ,like Smaller than the set viewing cone angle , then and grouped into one category and and Marked as clustered; S23, when Judge as After clustering into one category, calculate Curvature With point Curvature The difference ,like Less than the set curvature change threshold , then Put in the cluster starting sequence middle; S24, traversal completed After the neighboring points of Reselect the point with the smallest curvature as the starting point of the new cluster, and repeat steps S22-S23 for the points that are not marked as clustered until the sequence is empty; points that are judged to be clustered together with the new cluster starting point are also considered as group into one category; And all its cluster points are put into the same point set, that is, a point set of a continuously scannable area is obtained. ; S25, set the vertex The points that are not marked as clustered in , for the set Repeat steps S21-S24 until the vertex set All points in are marked as clustered, and multiple different point sets of continuously scannable areas are obtained.

4. The method for planning a curved surface scanning path of a robot arm according to claim 2, characterized in that: Step S3 specifically includes: S31, traverse the vertex triangle set , find all the edges that belong to only one triangle, take out the vertices of these edges, and get the surface boundary line point set ; S32, projecting the boundary line points onto a unit circle in a two-dimensional plane; for a boundary line point , and its projected coordinates are ,in, , , Boundary point exist The index in is the total number of boundary line points; S33, define a triangle gradient matrix for describing the relationship between vertex coordinate changes and two-dimensional gradients , for any vertex triangle , its triangular gradient matrix is: ； in, is the area of ​​the triangle, , , , The triangle The three vertices of and The coordinates of the axes; S34. Constructing non-conformal energy function of triangle , for the triangle , assuming that the two-dimensional coordinates of its three vertices are , , , its energy function for: ； in, , ; S35. Accumulate the non-conformal energy of all triangles to obtain the global energy function ; S36. Expand the global energy function and express it as a linear system ; in, , is a The sparse coefficient matrix of is related to the triangular gradient matrix, The dimension introduced by using the coordinates of all boundary points as constraints is A vector of S37, use the sparse matrix solution method to solve x, so that the global energy function As the minimum value, solve the two-dimensional coordinates of each vertex , That is, the vertices in three-dimensional space Mapped to a point in two-dimensional space.

5. The method for planning a curved surface scanning path of a robot arm according to claim 1, characterized in that: Step S4 specifically includes: S41, for any continuously scannable area point set , after step S3, we get The corresponding points of all points in two-dimensional space constitute a two-dimensional continuous scanning area point set ; S42, using the convex hull algorithm to obtain the two-dimensional continuously scannable area point set Find the two points farthest from each other and , and find out and Line segment between two points ; S43, for the two-dimensional continuously scannable area point set Medium and Project the point outside the On, find The two points farthest apart in a direction and , find the straight line segment passing through these two points ; S44, according to the scanning width set by the scanner Straight line segment Divide the straight line segment into equal parts Divide into segment, of which Indicates rounding up. For straight line segment Length; S45, extract the straight line segment The equal nodes ; S46, through the straight line segment The nodes of the equal division Parallel line segments ,by As the starting line segment, The parallel line segments are connected in sequence, and the point set is obtained. Two-dimensional scanning path ; S47, along the two-dimensional scanning path Sampling is performed with a fixed sampling step size s, with a point set The point closest to the sampling point is the two-dimensional scanning path point, and the point set is obtained. A two-dimensional scanning path point array ; S48. Repeat steps S41 to S47 for other continuously scannable area point sets to obtain a two-dimensional scanning path point array for each continuously scannable area.

6. The method for planning a curved surface scanning path of a robot arm according to claim 5, characterized in that: Step S5 specifically includes: S51. Find the two-dimensional scanning path point array with the most two-dimensional scanning path points as well as The last waypoint in ,calculate Find the distance from the first path point in the remaining two-dimensional scan path point array The path point with the smallest distance , and its corresponding two-dimensional scanning path point array ; S52, will The path points in the original order are added to At the end, we get a new two-dimensional scanning path point array , and and Mark as merged; S53, according to step S52, in the remaining unmerged array, find the distance of the first path point The array closest to the last path point is merged and updated ; S54, repeat step S53 until all arrays are marked as merged, and obtain a complete two-dimensional scanning path point array of the surface to be scanned .

7. The method for planning a curved surface scanning path of a robot arm according to claim 1, characterized in that: Step S6 specifically includes: S61. Find the complete two-dimensional path point array of the surface to be scanned according to the index correspondence The path points in the image correspond to the vertices of the triangular mesh model of the surface in three-dimensional space, and a complete three-dimensional path point array of the surface to be scanned is obtained. ; S62. For a three-dimensional path point array A waypoint in , move the point along its normal Optimal scanning distance for offset scanners , get the offset path point ; S63. For a three-dimensional path point array Step S62 is performed on all path points in the scanned surface to obtain a complete three-dimensional scanning path point array. .

8. The method for planning a curved surface scanning path of a robot arm according to claim 7, characterized in that: Step S7 specifically includes: S71. For a complete 3D scanning path point array Offset path points in , its normal vector As the Z axis of the scanner tool coordinate system, that is: ; S72, according to the scanner tool coordinate system The X-axis of the tool coordinate system is determined by the X-axis of the coordinate system of the surface to be scanned. First, a reference vector of the X-axis of the coordinate system of the surface to be scanned is selected. , calculate the temporary vector ,Will Normalize to get the X-axis of the tool coordinate system; S73, determining the Y axis of the tool coordinate system according to the X axis and the Z axis of the tool coordinate system; S74, the three axis vectors of the tool coordinate system form a rotation matrix ; S75. According to the rotation matrix Calculate the scanner at the offset path point RPY static angle at and ; S76, will Coordinates and in The RPY static angle of the scanner at and Combined into a robotic arm Executable points at ; S77, complete three-dimensional scanning path point array Execute steps S71 to S76 for all offset path points in the to-be-scanned surface to obtain an array of executable scanning path points for the robot arm. , complete the robotic arm scanning path planning of the surface to be scanned.

Images