CN117444968A - Method for planning scanning measurement path of large-scale barrel part local characteristic robot - Google Patents
Method for planning scanning measurement path of large-scale barrel part local characteristic robot Download PDFInfo
- Publication number
- CN117444968A CN117444968A CN202311448811.3A CN202311448811A CN117444968A CN 117444968 A CN117444968 A CN 117444968A CN 202311448811 A CN202311448811 A CN 202311448811A CN 117444968 A CN117444968 A CN 117444968A
- Authority
- CN
- China
- Prior art keywords
- scanning
- point
- points
- interference
- planning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000005259 measurement Methods 0.000 title claims abstract description 84
- 238000000034 method Methods 0.000 title claims abstract description 48
- 239000013598 vector Substances 0.000 claims abstract description 92
- 238000012545 processing Methods 0.000 claims abstract description 30
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 16
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000009826 distribution Methods 0.000 claims description 7
- 229910000831 Steel Inorganic materials 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 239000010959 steel Substances 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims 1
- 238000003754 machining Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 239000003016 pheromone Substances 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000000280 densification Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012625 in-situ measurement Methods 0.000 description 1
- 238000005305 interferometry Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 210000000707 wrist Anatomy 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1656—Programme controls characterised by programming, planning systems for manipulators
- B25J9/1664—Programme controls characterised by programming, planning systems for manipulators characterised by motion, path, trajectory planning
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Numerical Control (AREA)
Abstract
The invention discloses a method for planning a scanning measurement path of a large-scale barrel part local characteristic robot, which comprises the following steps: planning a scanning measurement viewpoint and planning a scanning measurement path; based on the model, identifying and dividing local features, establishing a scanning view field of each scanning measurement viewpoint according to the feature type, and completing planning of the scanning measurement viewpoints after interference judgment and processing; and forming a path by taking the planned scanning measurement viewpoint as a path point, solving the shortest path by using an ant colony algorithm, and adjusting by considering the unreachable point of the robot and the interference condition of the track and the workpiece to finally realize the scanning measurement path planning. The invention can effectively plan the scanning measurement view point of the local characteristic of the large-scale barrel part containing position and normal vector information according to the three-dimensional model of the measured part, and can plan the shortest scanning measurement path of the robot while avoiding the occurrence of the overrun condition and the interference condition of the robot.
Description
Technical Field
The invention relates to the technical field of three-dimensional scanning measurement of large-scale components, in particular to a method for planning a scanning measurement path of a large-scale barrel part local characteristic robot.
Background
The large-scale component assembly interface is a key machining feature for guaranteeing the final assembly quality of the aerospace large-scale industrial product, and the machining feature is effectively measured before final assembly so as to obtain machining allowance, so that the high-precision and high-efficiency finish machining of the assembly interface is an important guarantee.
Because of the special fixture or positioning equipment, the pose measurement or alignment can be directly carried out on the processing site, and the rest large cylinder components can only acquire pose data in a manual in-situ measurement mode, so that the efficiency of the mode is extremely low, the operation of workers is high, and the measurement precision and efficiency cannot be ensured. With the increase of annual output requirements of the barrel components, the production beats are continuously compressed, and the traditional gesture measurement and adjustment procedure is time-consuming and labor-consuming, so that the production bottleneck of the large barrel components is formed.
Disclosure of Invention
Aiming at the problems that the traditional processing mode of the large-sized barrel part is complex in process, strong in dependence of artificial experience and unstable in measurement result, the invention provides a method for planning a scanning measurement path of a large-sized barrel part local characteristic robot.
In order to achieve the above object, the present invention provides a method for planning a scanning measurement path of a large-scale drum local feature robot, comprising:
acquiring an original three-dimensional model of a barrel part, preprocessing the original three-dimensional model, and respectively acquiring triangular surface vertex and normal vector information in the original three-dimensional model and the preprocessed model;
characteristic identification and division are carried out based on triangular surface vertexes of the preprocessing model, and characteristics to be scanned are extracted, wherein the characteristics to be scanned comprise conventional structural characteristics and special structural characteristics, the conventional structural characteristics are structures which are directly planned according to triangular surface information of the model, and the special structural characteristics are structures which are required to be independently planned;
performing triangular surface normal vector clustering on each characteristic in the conventional structural characteristics, determining a scanning range of a single scanning measurement viewpoint, establishing a scanning view field of the single measurement viewpoint, and independently planning the measurement viewpoint according to triangular surface position information of the characteristic type by combining the characteristic type with the special structural characteristic;
based on the measurement view point, respectively processing and adjusting the interference condition of the scanner and the workpiece and the interference condition of the scanner and the scanning view field to obtain a planned scanning measurement view point;
and according to the planned scanning measurement view point, solving the pose of the robot through the kinematic inverse of the robot, judging whether the number of the unreachable points of the robot exceeds a preset threshold, respectively processing according to a judging result, obtaining a scanning measurement path, and solving the shortest path through an ant algorithm.
Preferably, preprocessing the original three-dimensional model to obtain triangular surface vertex and normal vector information in the original three-dimensional model and the preprocessed model, respectively, including:
preprocessing hanging holes to be scanned and measured in the original three-dimensional model, filling hole characteristics, and respectively reading triangular plane normal vector and vertex coordinate information of the original three-dimensional model and the preprocessed model.
Preferably, the extracting the features to be scanned includes:
and dividing the features to be scanned and measured and flange frames at two ends of the barrel by the coordinate distribution condition in the axis direction, removing the barrel part information irrelevant to the scanning and measuring according to the radius, and extracting the features to be scanned.
Preferably, performing triangular face normal vector clustering on each of the conventional structural features includes:
and clustering the conventional structural features by using a K-means algorithm, carrying out ordering treatment, respectively clustering each supporting leg and the suspended triangular surface normal vector in the conventional structural features according to the ordering treatment result, and demarcating the scanning range of a single measurement viewpoint in space based on the triangular surface vertex coordinates after each clustering.
Preferably, the demarcating the scanning range of the single measurement viewpoint in the space based on the triangular surface vertex coordinates after each clustering includes:
s1.1, creating and clustering a normal vector v in each class k Straight line L which coincides with and passes through origin k Projecting all points in the class to the line L k On the above, find out all the projection pointsIn straight line L k The two end points are taken as scanning heights, and the midpoint P of a line segment formed by the two end points is recorded m ；
S1.2 creating a plane S in each class k The plane S k Normal vector n of (2) k Normal vector v obtained after clustering with the data set k Coincident and in the same direction, the plane S k Passing through the midpoint P m Projecting all points in the class to said plane S k Obtaining a projection point setCalculating a set comprising said projection points +.>The minimum circumscribed rectangle of (2) is taken as a scanning plane S r ；
S1.3, in the scanning plane S r Respectively stretching the steel sheet upwards and downwards to serve as a reference surfaceThe formed space is the scanning range;
s1.4, clustering coordinates in the class if the scanning range is within the scannable range of the scanning field, classifying the coordinates into K' class, repeating S1.1-S1.3, and finally obtaining a scanning range set meeting the requirements.
Preferably, in S1.2, the method for solving the minimum bounding rectangle is:
s2.1, selecting a starting point P 0 At the saidProjection point setIf the point with the smallest y coordinate is found and contains a plurality of points with the same y coordinate, the point with the smallest x coordinate is taken as the starting point P 0 ；
S2.2, all the points in the point set are matched with the starting point P 0 The polar angles of the two layers are ranked from small to large according to the distance from the point to the starting point if the polar angles are the same;
s2.3, sequentially adding the ordered points into a convex hull, storing the points in the convex hull in a stack according to the adding sequence, carrying out cross product calculation on each point added with a vector formed by the point at the top of the stack and the previous point at the top of the stack, and if the cross product result is smaller than or equal to a preset value, removing the point at the top of the stack of the convex hull when the convex hull does not meet convexity;
s2.4, repeating the step S2.3 until all points are added into the convex hull;
s2.5, solving a minimum circumscribed rectangle by using a rotary cartoning method according to the convex hull, and finding a point pair P with the farthest distance on the convex hull i-1 、P i One side C as the smallest circumscribed rectangle l C i At C l C i Finding the furthest point P in the vertical direction k Drawing and C l C i Parallel side C of (2) j C k Intersecting at point P k At C l C i Finding the two points p furthest apart in the direction l 、P j Drawing edges thereby drawing and C l C i Two vertical sides C k C l 、C i C j Thereby forming an external rectangle, wherein, the included angle theta formed by each side of the rectangle and the convex hull i 、θ j 、θ k 、θ l The method comprises the steps of carrying out a first treatment on the surface of the The rectangle rotates anticlockwise along the edge of the convex hull, the convex hull is kept to be completely in the rectangular surrounding during the rotation, and the rotation angle and the circumscribed rectangular area S are recorded once per rotation i When the rotation angle reaches 90 degrees, finally obtaining the minimum value in the recorded area, wherein the rectangle corresponding to the minimum value is the minimum valueAnd (5) externally connecting a rectangle.
Preferably, the measuring viewpoint is individually planned according to the triangular face position information of the feature type, including:
for a single aperture feature, determining a measurement viewpoint from a conical field of view of the laser scanner based on the location and size of the aperture feature; clustering coordinate points of the distributed holes by using a DBSCAN algorithm for the distributed hole features, determining the positions of the hole features, and determining a measurement viewpoint according to a conical view field of a laser scanner;
wherein the feature types include: the outer surface, the inner surface and the assembly surface of the hanging hole and the flange frame.
Preferably, obtaining the planned scan measurement viewpoint includes:
the method comprises the steps of performing spatial interference processing by a rotation vector method, firstly processing the interference condition of a scanner and a workpiece, and then processing the interference condition of the scanner and a scanning view field until the planned scanning measurement view point is obtained;
the method for rotating the vector comprises the following steps:
s3.1, taking the minimum surrounding rectangular central point obtained in the scanning range as an origin, recording vectors of the origin and each interference point, and calculating an average value of the vectors to obtain an interference point reference vector u;
s3.2, performing cross multiplication on the interference point reference vector u and the normal vector v to obtain a rotation vector v R ；
S3.3, the normal vector v is along the rotation vector v R Performing rotation, performing space interference judgment once every rotation, stopping rotation if no interference occurs, and recording the rotated normal vector v' as the normal vector after interference treatment;
and S3.4, if interference conditions still exist after the maximum rotation angle is reached, replacing the rotation direction, repeating the step S3.3, and if the rotated normal vector v' meeting the conditions still cannot be obtained, clustering the points in the scanning range into two types through a K-means algorithm, and then repeating the judgment and the processing of spatial interference.
Preferably, the processing according to the determination result includes:
if the unreachable points exceed the preset threshold, dividing the unreachable points into two groups according to the coordinate distribution condition, rotating the cylinder piece along the axis, and completing scanning measurement in two times until the number of the unreachable points of the robot is smaller than the preset threshold;
if the unreachable point does not exceed the preset threshold, taking the scanning measurement viewpoint as a path point to form paths, and respectively solving each group of shortest paths by using an ant colony algorithm;
and for the unavoidable unreachable points, performing manual pose adjustment, establishing a conical scanning view field in space, rotating the scanning view field direction until the robot pose is reachable, and simultaneously meeting the condition that the cuboid scanning range is still in the conical scanning view field.
Preferably, the method further comprises the steps of judging whether a space interference condition occurs in the track executing process by using simulation software, adding path points to the path with interference to avoid interference, and simulating and verifying the correctness of the scanning measurement path.
Compared with the prior art, the invention has the following advantages and technical effects:
(1) According to the invention, the processing allowance information of the workpiece is obtained by carrying out three-dimensional scanning measurement on the workpiece, and the processing technology planned based on the theoretical model is adjusted or optimized according to the obtained actual processing allowance of the processing characteristics of the large-scale component, so that the finish machining quality of an assembly interface is effectively improved;
(2) The invention can effectively plan the scanning measurement view point of the local characteristic of the large-scale barrel part containing the position and the legal loss information according to the three-dimensional model of the measured part, and can plan the shortest scanning measurement path of the robot while avoiding the occurrence of the overrun condition and the interference condition of the robot.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application, illustrate and explain the application and are not to be construed as limiting the application. In the drawings:
FIG. 1 is a flow chart of a method for planning a scanning and measuring path of a large-scale drum part local characteristic robot in an embodiment of the invention;
FIG. 2 is a view of an object of investigation according to an embodiment of the present invention;
FIG. 3 is a schematic view of determining a scan height and a scan plane according to an embodiment of the present invention, wherein (a) is a schematic view projected onto a straight line and (b) is a schematic view projected onto a plane;
FIG. 4 is a schematic diagram of a rotary shell-and-card method according to an embodiment of the present invention, wherein (a) is a schematic diagram before rotation and (b) is a schematic diagram after rotation;
FIG. 5 is a schematic view of a scan field of view according to an embodiment of the present invention;
FIG. 6 is a schematic view illustrating a measurement view point of a hanging hole according to an embodiment of the present invention;
fig. 7 is a schematic view of flange frame measurement, in which (a) is a schematic view of outer surface planning and (b) is a schematic view of inner surface planning;
FIG. 8 is a schematic diagram of a scan-field unobstructed area setup of an embodiment of the invention;
FIG. 9 is a flow chart of a spatial interferometry process according to an embodiment of the present invention.
Detailed Description
It should be noted that, in the case of no conflict, the embodiments and features in the embodiments may be combined with each other. The present application will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.
The large-scale barrel is taken as a research object, the diameter of the barrel is 550mm, the length of the barrel is 2090mm, and the whole barrel mainly comprises a barrel section, two flange frames, four supporting legs and four hangers, as shown in figure 2. The research content is based on a three-dimensional model and is to a more complex characteristic structure on the barrel part: and carrying out three-dimensional scanning measurement planning on the support legs, the hanging and the flange frame. The support legs and the hanging assembly surface are mainly processed, processing allowance information is obtained through scanning measurement of the support legs and the hanging assembly surface, the flange frame is used as a positioning feature, scanning measurement is conducted on the flange frame, and a workpiece coordinate system is obtained through fitting according to a point cloud of a measurement result and is used as a processing reference.
The invention provides a method for planning a scanning measurement path of a large-scale barrel part local characteristic robot, which comprises the following steps of:
s101, preprocessing four hanging holes to be scanned and measured in a model, filling hole characteristics, and respectively reading triangular plane normal vector and vertex coordinate information of an original model and a preprocessed model;
s102, carrying out feature identification and division based on triangular surface information of a preprocessing model, dividing the two features needing scanning measurement and flange frames at two ends of a barrel part through the coordinate distribution condition of an axis direction, and eliminating information of a barrel part completely irrelevant to the scanning measurement according to the radius; so far, the characteristics of two parts to be scanned can be extracted: hanging, supporting legs and a flange frame;
the triangular surface vertexes of the hanging support leg parts are clustered by using a K-means algorithm, the total number of the supports and the hanging in the cylinder is 8, so that the K value is 8, ordered processing is carried out after the clustering is completed, the four hanging and the four support legs are numbered according to the vertex coordinate distribution condition of the triangular surface, namely, the hanging 1, the hanging 2, the hanging 3, the hanging 4, the support leg 1, the support leg 2, the support leg 3 and the support leg 4, the index value corresponding to the data set is 0 to 7, and each index value is guaranteed to correspond to a specific known characteristic structure.
S103, hanging holes and flange frames which can be planned directly according to the triangular surface information of the model (without holes) are of conventional structural characteristics, and hanging holes and flange frames which cannot be planned directly according to the triangular surface information of the model and need to be planned independently are of special structural characteristics.
S104, clustering each supporting leg and the triangular surface normal vector of the hanger respectively according to an ordering processing result, wherein the selection of the K value depends on the number of the surfaces with larger normal vector difference, and the number of the surfaces with larger normal vector difference is 7 for the hanger, so the K value takes 7; similarly, 8 faces with large normal vector differences are taken as 8 for the legs.
The scanning range of a single measurement viewpoint in space is defined based on the triangular plane vertex coordinates after each clustering, the scanning range is defined as a cuboid, the cuboid needs to contain all triangular plane vertices, and meanwhile, the volume of the cuboid needs to be ensured to be as small as possible, and the method specifically comprises the following steps:
(1) First, a straight line is projected. Creating a normal vector v obtained after clustering in each class k Straight line L which coincides with and passes through origin k As shown in fig. 3 (a). All points (P) 1 ，P 2 ，P 3 ，...，P i ) According to the formulaProjected to straight line L k On the one hand, find out these projection points +.>Two end points on the straight line are taken as the scanning height H, and the midpoint P of the line segment formed by the two end points is recorded m 。
(2) And then projected onto a plane. Creating a plane S in each class k Normal vector n of the plane k Normal vector v obtained after clustering with the data set k Coincide and co-rotate, the plane passing through P m As shown in fig. 3 (b). All points (P) 1 ，P 2 ，P 3 ，...，P i ) According to the formulaProjected to plane S k Obtaining a projection point setComputing contains the set of projection points +.>Is taken as a scanning plane S r 。
(3) To sweep byDrawing plane S r Respectively stretching the steel sheet upwards and downwards to serve as a reference surfaceThe space formed is the scanning range.
(4) If the scanning range is not satisfied, clustering the coordinates in the class, classifying the coordinates into K' class, and repeating the steps (1) to (3) to finally obtain a scanning range set meeting the requirements.
For solving the minimum circumscribed rectangle in the step (2), firstly, a Graham algorithm is selected to solve the convex hull, and the specific steps are as follows:
s1, selecting a starting point P 0 In the projection point setIf a plurality of points have the same y coordinates, taking the point with the smallest x coordinates as a starting point, wherein the reason for selecting the point with the smallest y coordinates as the starting point is that the point with smaller y coordinates is more likely to be a point on the convex hull as the point is more close to the lower part in the point set; the point with the smallest x coordinate is selected to ensure the uniqueness of the starting point in the point set;
s2, all the points in the point set are matched with the starting point P 0 The polar angles of the points are sequenced from small to large, if the polar angles are the same, the polar angles are sequenced from small to large according to the distance from the points to the starting point, so that the uniqueness of each point in the point set is ensured;
s3, sequentially adding the ordered points into the convex hull, storing the points in the convex hull in the stack according to the adding sequence, carrying out cross product calculation on each added point and a vector formed by the point at the top of the stack and the previous point at the top of the stack, and if the cross product result is less than or equal to 0, removing the point at the top of the stack of the convex hull when the current convex hull does not meet convexity;
s4, repeating the step S3 until all points are added into the convex hull;
according to the convex hull, the minimum circumscribed rectangle is solved by using a rotation clamping method, the rotation process of the rotation clamping method is schematically shown in figure 4, and after the convex hull P is obtained, the rotation clamping method is as followsFIG. 4 (a) shows that the two most distant point pairs P are found on the convex hull i-1 、P i These two points will become one side C of the smallest bounding rectangle l C i In straight line C l C i Finding the furthest point P in the vertical direction k Drawing and C l C i Parallel side C of (2) j C k Intersecting at point P k In straight line C l C i Finding the two points P furthest apart in the direction l 、P j Drawing edges thereby drawing and C l C i Two vertical sides C k C l 、C i C j Thereby forming an external rectangle, wherein, the included angle theta formed by each side of the rectangle and the convex hull i 、θ j 、θ k 、θ l The method comprises the steps of carrying out a first treatment on the surface of the The rectangle is rotated anticlockwise along the edge of the convex hull, the convex hull is required to be kept completely in the rectangular surrounding during the rotation, the situation after the first rotation is shown in fig. 4 (b), and the rotation angle and the circumscribed rectangular area S are recorded when the rectangle rotates once i When the rotation angle reaches 90 degrees, finally, the minimum value in the recorded area is obtained, and the rectangle corresponding to the minimum value is the minimum circumscribed rectangle;
because the existing convex hull solving methods are all methods in a two-dimensional space, if a plane normal vector is not parallel to any one of three coordinate axes of x, y and z in a three-dimensional space, the convex hull is difficult to directly find on the plane, so that the determination of the minimum circumscribed rectangle cannot be performed, the problem needs to be solved by projecting the minimum circumscribed rectangle to the two-dimensional plane, and the obtained minimum circumscribed rectangle is mapped back to the three-dimensional space through the intersection point of a straight line and the original plane.
The scan field is the range that the scanner can scan when fixing a certain angle of view, and the cone-shaped scan field is determined according to the field parameters of the scanner, as shown in figure 5, wherein P s To measure the viewpoint position, S 3 Is a deep plane of long view, P 3 For the surface S 3 Is defined by a center point of the lens. S is S 2 The data information can be acquired in the range of the far view depth on the basis of the reference distance of the view point, more accurate data information can be acquired in the reference plane, and P 2 For the surface S 2 Is defined by a center point of the lens. S is S 1 To avoid spatial interference during the movement of the scanner and to take the size of the scanning area into account, S will therefore be 1 The position of (2) is set to P s P 3 Midpoint, P of (1) 1 For the surface S 1 Is defined by a center point of (2); p in scan field of view s P 3 The normal vector formed coincides with and is inverted to the normal vector of the scan range, which is required to be within the scan field of view and is S 1 To S 3 The reference distance plane of the scan field is generally coincident with the plane of the minimum circumscribed rectangle, P 2 Coinciding with the midpoint of the smallest bounding rectangle.
S105, scanning the hanging hole with 6 viewpoints, as shown in FIG. 6, P 1 To P 6 Six points on the inner ring of the chamfer at the inner side of the hole, the included angle formed by two adjacent points and the center of the inner ring of the chamfer is 60 degrees, and P 0 Is the center point on the outer side of the hole, P 0 And P 1 -P 6 The vectors formed by connection are six normal vectors, and the positions of the normal vectors are defined as the midpoints of six line segments;
the measurement plan of the flange frame can be divided into three parts: an outer surface, an inner surface, and an assembly surface; the flange frames at two sides of the cylinder part are different, 13 holes are formed at one side of the cylinder part, one hole is different from the other 12 holes, 18 holes are formed at the other side of the cylinder part, and the sizes of all the holes are consistent;
for the outer surface, the outer surface is divided into any 6 sections of circular arcs, the size of a line segment formed by connecting two end points of the circular arcs is 275mm, the end point of each section of circular arc is used as a normal vector position, 6 vectors pointing to the end points of the 6 sections of circular arcs from the center of the end face of the cylinder part are generated, one vector is used as an example, and as shown in fig. 7 (a), the vector is P 0 P s0 The new vector P obtained by rotating 45 DEG 0 P s Is normal vector, angle P s0 P 0 P s At 45 deg., the viewpoint is measured at a reference distance from the normal vector position in the normal vector direction, i.e. line segment P 0 P s The length is the reference distance;
because of the large size of scanners and robots, it is difficult to place them inside the cartridge to perform the measurement, so the inner surface requires the measurementThe measuring point is planned outside the barrel, as shown in FIG. 7 (b), P 1 、P 2 Is the two points with the longest distance after the outermost projection of the inner surface, P 0 Is the projection of the center of a circle, P s Is to measure the viewpoint position, line segment P 0 P 1 、P 0 P 2 Is 245mm in length, P is defined herein 0 P s The length of the line segment is 180mm, the normal vector of the inner surface also needs to refer to the method of dividing the outer surface into circular arcs, the number is 6, and the normal vector direction points to the scanning viewpoint P from the circular arc end points s ；
In the planning of the assembly surface, aiming at the characteristics of a three-dimensional model, the scanning measurement of the whole assembly surface can be completed while the hole is scanned, and the DBSCAN algorithm is used for clustering coordinate points of the hole and removing noise points, wherein the radius value is 22, the minimum sample number value is 100, the normal vector of each hole coincides with the axis of the hole, the direction points to the outer side, and the position of the measurement viewpoint is defined at the position where the distance of the midpoint of each cluster along the normal vector is the reference distance.
S106, in the measurement viewpoint planning stage, two cases of spatial interference occur: the scanner interferes with the space of the workpiece, and the scanning view field is blocked by the non-scanning area;
the judgment of both cases can be converted into the problem that whether the judged point set is in the cuboid or not, the problem can be converted into the problem that whether the judging points are on the same side of two parallel planes or not, the problem of the normal included angle between the judging points and the sides of the cuboid or not, and the problem of the symbol of the cosine value, namely the symbol of the vector inner product, is finally converted into the problem;
because the triangular surface area of some parts in the cylinder is larger, the vertex distribution is sparse, the dense treatment is needed, and points are added on the edges of the triangular surface with the length of more than 2mm, so that the distance between the middle point and the points of the cylinder is not more than 2mm; taking the triangular surface vertex subjected to the barrel part densification treatment as an interference judgment point set, taking a scanner or a scanning view field as an interference judgment cuboid, and realizing the judgment of spatial interference according to the mode that whether each point in the interference judgment point set is in the interference judgment cuboid or not; consider a scanner to robot tip connectionAnd the influence of the robot wrist in space is regarded as a cuboid of 600mm multiplied by 310mm in interference judgment according to the size of the scanner, a square of 310mm multiplied by 310mm is taken as the bottom surface of the cuboid, the central point of the bottom surface is a measurement viewpoint, and the direction of the longest side is the same as the direction of the normal vector; although the scan field of view is a rectangular pyramid, the actual scan target is a point in the scan range, so it is only necessary to consider that the scan range is not blocked, as shown in FIG. 8, P s For measuring the view point, the vector n is the scanning view field direction, the normal vector direction is opposite, a rectangle is established as the bottom surface on the farthest surface in the normal vector direction according to the cuboid size obtained by the scanning range, the rectangle is stretched along the normal vector direction until the rectangle is intersected with the measuring view point, and the cuboid is the non-shielding area in fig. 8.
S107, recording a point set which is interfered in each scanning viewpoint space interference judgment in the space, namely, the point set in the cuboid, performing space interference processing by a rotating normal vector method, firstly processing the interference condition of a scanner and a workpiece, and then processing the interference condition of a scanning view field, wherein the processing ideas of the two are the same, and the specific flow is shown in a figure 9, wherein the rotating vector method comprises the following steps:
(1) Taking the minimum surrounding rectangular central point obtained in the scanning range as an origin, recording vectors of the origin and each interference point, and calculating an average value to obtain an interference point reference vector u;
(2) The interference point reference vector u and the normal vector v are subjected to cross multiplication to obtain a rotation vector v R ；
(3) The normal vector v is followed by the rotation vector v R Performing rotation, performing space interference judgment once every rotation, stopping rotation if no interference occurs, and recording the normal vector v' after the rotation at the moment as the normal vector after the interference treatment;
(4) And (3) if the interference condition still exists after the maximum rotation angle is reached, replacing the rotation direction, repeating the step (4), and if the normal vector v' meeting the condition still cannot be obtained, clustering the points in the scanning range into two types through a K-means algorithm, and repeating the judgment and the processing of the spatial interference.
S201, according to the planned scanning measurement viewpoint, solving the pose of the robot through inverse solution of the robot kinematics, judging whether the pose is reachable or not, and finally, scanning measurement is performed in four times, wherein the barrel rotates 90 degrees around the axis before the second to fourth times of measurement. The target points are stuck on the surface of the workpiece, and the tracker is used for capturing the target points in the measuring process, so that the measured point cloud data is ensured to have a uniform reference under a uniform coordinate system under the condition that the workpiece rotates.
For unavoidable unreachable points, manual pose adjustment is needed, a conical scanning view field is established in space, the direction of the scanning view field is rotated until the pose of the robot is reachable, and meanwhile the condition that the cuboid broken scanning range is still in the conical scanning view field is met.
S202, carrying out optimal path solving on paths formed by scanning measurement viewpoints of all groups one by using an ant colony algorithm, wherein the starting point of a first group is a robot HOME point, the starting point of a second group is an end point of the first group, the starting point of a third group is an end point of the second group, and the like; the starting point pheromone is set as 100, the values of the other elements are 0.1, the important degree factors of the pheromone are traversed within the range of 0.5-1.4, the important degree factors of the heuristic function are traversed within the range of 4.0-9.0, and meanwhile, the volatile factor of the pheromone is taken to be 0.999.
The foregoing is merely a preferred embodiment of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions easily conceivable by those skilled in the art within the technical scope of the present application should be covered in the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method for planning a scanning measurement path of a large-scale barrel part local characteristic robot is characterized by comprising the following steps:
acquiring an original three-dimensional model of a barrel part, preprocessing the original three-dimensional model, and respectively acquiring triangular surface vertex and normal vector information in the original three-dimensional model and the preprocessed model;
characteristic identification and division are carried out based on triangular surface vertexes of the preprocessing model, and characteristics to be scanned are extracted, wherein the characteristics to be scanned comprise conventional structural characteristics and special structural characteristics, the conventional structural characteristics are structures which are directly planned according to triangular surface information of the model, and the special structural characteristics are structures which are required to be independently planned;
performing triangular surface normal vector clustering on each characteristic in the conventional structural characteristics, determining a scanning range of a single scanning measurement viewpoint, establishing a scanning view field of the single measurement viewpoint, and independently planning the measurement viewpoint according to triangular surface position information of the characteristic type by combining the characteristic type with the special structural characteristic;
based on the measurement view point, respectively processing and adjusting the interference condition of the scanner and the workpiece and the interference condition of the scanner and the scanning view field to obtain a planned scanning measurement view point;
and according to the planned scanning measurement view point, solving the pose of the robot through the kinematic inverse of the robot, judging whether the number of the unreachable points of the robot exceeds a preset threshold, respectively processing according to a judging result, obtaining a scanning measurement path, and solving the shortest path through an ant algorithm.
2. The method for planning a scanning measurement path of a large-scale drum part local feature robot according to claim 1, wherein preprocessing the original three-dimensional model to obtain triangular surface vertex and normal vector information in the original three-dimensional model and the preprocessed model respectively comprises:
preprocessing hanging holes to be scanned and measured in the original three-dimensional model, filling hole characteristics, and respectively reading triangular plane normal vector and vertex coordinate information of the original three-dimensional model and the preprocessed model.
3. The method for planning a scanning measurement path of a large-scale cartridge local feature robot according to claim 2, wherein the extracting features to be scanned comprises:
and dividing the features to be scanned and measured and flange frames at two ends of the barrel by the coordinate distribution condition in the axis direction, removing the barrel part information irrelevant to the scanning and measuring according to the radius, and extracting the features to be scanned.
4. The method for planning a scanning measurement path of a large-scale cartridge local feature robot of claim 1, wherein performing triangular face normal vector clustering on each of the conventional structural features comprises:
and clustering the conventional structural features by using a K-means algorithm, carrying out ordering treatment, respectively clustering each supporting leg and the suspended triangular surface normal vector in the conventional structural features according to the ordering treatment result, and demarcating the scanning range of a single measurement viewpoint in space based on the triangular surface vertex coordinates after each clustering.
5. The method for planning a scanning measurement path of a large-scale cartridge local feature robot according to claim 4, wherein the defining the scanning range of the single measurement viewpoint in space based on the triangular surface vertex coordinates after each clustering comprises:
s1.1, creating and clustering a normal vector v in each class k Straight line L which coincides with and passes through origin k Projecting all points in the class to the line L k On the above, find out all the projection pointsIn straight line L k The two end points are taken as scanning heights, and the midpoint P of a line segment formed by the two end points is recorded m ；
S1.2 creating a plane S in each class k The plane S k Normal vector n of (2) k Normal vector v obtained after clustering with the data set k Coincident and in the same direction, the plane S k Passing through the midpoint P m Projecting all points in the class to said plane S k Obtaining a projection point setCalculating a set comprising said projection points +.>The minimum circumscribed rectangle of (2) is taken as a scanning plane S r ；
S1.3, in the scanning plane S r Respectively stretching the steel sheet upwards and downwards to serve as a reference surfaceThe formed space is the scanning range;
s1.4, clustering coordinates in the class if the scanning range is within the scannable range of the scanning field, classifying the coordinates into K' class, repeating S1.1-S1.3, and finally obtaining a scanning range set meeting the requirements.
6. The method for planning a scanning measurement path of a large-scale drum local feature robot according to claim 5, wherein in S1.2, the method for solving the minimum bounding rectangle is as follows:
s2.1, selecting a starting point P 0 At the projection point setIf the point with the smallest y coordinate is found and contains a plurality of points with the same y coordinate, the point with the smallest x coordinate is taken as the starting point P 0 ；
S2.2, all the points in the point set are matched with the starting point P 0 The polar angles of the two layers are ranked from small to large according to the distance from the point to the starting point if the polar angles are the same;
s2.3, sequentially adding the ordered points into a convex hull, storing the points in the convex hull in a stack according to the adding sequence, carrying out cross product calculation on each point added with a vector formed by the point at the top of the stack and the previous point at the top of the stack, and if the cross product result is smaller than or equal to a preset value, removing the point at the top of the stack of the convex hull when the convex hull does not meet convexity;
s2.4, repeating the step S2.3 until all points are added into the convex hull;
s2.5, solving a minimum circumscribed rectangle by using a rotary cartoning method according to the convex hull, and finding a point pair P with the farthest distance on the convex hull i-1 、P i One side C as the smallest circumscribed rectangle l C i At C l C i Finding the furthest point P in the vertical direction k Drawing and C l C i Parallel side C of (2) j C k Intersecting at point P k At C l C i Finding the two points P furthest apart in the direction l 、P j Drawing edges thereby drawing and C l C i Two vertical sides C k C l 、C i C j Thereby forming an external rectangle, wherein, the included angle theta formed by each side of the rectangle and the convex hull i 、θ j 、θ k 、θ l The method comprises the steps of carrying out a first treatment on the surface of the The rectangle rotates anticlockwise along the edge of the convex hull, the convex hull is kept to be completely in the rectangular surrounding during the rotation, and the rotation angle and the circumscribed rectangular area S are recorded once per rotation i And when the rotation angle reaches 90 degrees, finally obtaining the minimum value in the recorded area, wherein the rectangle corresponding to the minimum value is the minimum circumscribed rectangle.
7. The method for planning a scanning measurement path of a large-scale cartridge local feature robot according to claim 1, wherein the step of individually planning measurement viewpoints according to the triangular face position information of the feature type comprises the steps of:
for a single aperture feature, determining a measurement viewpoint from a conical field of view of the laser scanner based on the location and size of the aperture feature; clustering coordinate points of the distributed holes by using a DBSCAN algorithm for the distributed hole features, determining the positions of the hole features, and determining a measurement viewpoint according to a conical view field of a laser scanner;
wherein the feature types include: the outer surface, the inner surface and the assembly surface of the hanging hole and the flange frame.
8. The method for planning a scan measurement path of a large-scale cartridge local feature robot of claim 1, wherein obtaining the planned scan measurement viewpoint comprises:
the method comprises the steps of performing spatial interference processing by a rotation vector method, firstly processing the interference condition of a scanner and a workpiece, and then processing the interference condition of the scanner and a scanning view field until the planned scanning measurement view point is obtained;
the method for rotating the vector comprises the following steps:
s3.1, taking the minimum surrounding rectangular central point obtained in the scanning range as an origin, recording vectors of the origin and each interference point, and calculating an average value of the vectors to obtain an interference point reference vector u;
s3.2, performing cross multiplication on the interference point reference vector u and the normal vector v to obtain a rotation vector v R ；
S3.3, the normal vector v is along the rotation vector v R Performing rotation, performing space interference judgment once every rotation, stopping rotation if no interference occurs, and recording the rotated normal vector v' as the normal vector after interference treatment;
and S3.4, if interference conditions still exist after the maximum rotation angle is reached, replacing the rotation direction, repeating the step S3.3, and if the rotated normal vector v' meeting the conditions still cannot be obtained, clustering the points in the scanning range into two types through a K-means algorithm, and then repeating the judgment and the processing of spatial interference.
9. The method for planning a scanning measurement path of a large-scale drum local feature robot according to claim 1, wherein the steps of processing according to the judgment result respectively include:
if the unreachable points exceed the preset threshold, dividing the unreachable points into two groups according to the coordinate distribution condition, rotating the cylinder piece along the axis, and completing scanning measurement in two times until the number of the unreachable points of the robot is smaller than the preset threshold;
if the unreachable point does not exceed the preset threshold, taking the scanning measurement viewpoint as a path point to form paths, and respectively solving each group of shortest paths by using an ant colony algorithm;
and for the unavoidable unreachable points, performing manual pose adjustment, establishing a conical scanning view field in space, rotating the scanning view field direction until the robot pose is reachable, and simultaneously meeting the condition that the cuboid scanning range is still in the conical scanning view field.
10. The method of claim 9, further comprising determining whether a spatial interference condition occurs during the execution of the trajectory using simulation software, adding path point avoidance interference to the path where the interference occurs, and simulating verification of the correctness of the scanned measurement path.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311448811.3A CN117444968A (en) | 2023-11-02 | 2023-11-02 | Method for planning scanning measurement path of large-scale barrel part local characteristic robot |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311448811.3A CN117444968A (en) | 2023-11-02 | 2023-11-02 | Method for planning scanning measurement path of large-scale barrel part local characteristic robot |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117444968A true CN117444968A (en) | 2024-01-26 |
Family
ID=89583212
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311448811.3A Pending CN117444968A (en) | 2023-11-02 | 2023-11-02 | Method for planning scanning measurement path of large-scale barrel part local characteristic robot |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117444968A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019157924A1 (en) * | 2018-02-13 | 2019-08-22 | 视辰信息科技(上海)有限公司 | Real-time detection method and system for three-dimensional object |
| CN113155054A (en) * | 2021-04-15 | 2021-07-23 | 西安交通大学 | Automatic three-dimensional scanning planning method for surface structured light |
| CN115325962A (en) * | 2022-08-26 | 2022-11-11 | 中国科学院长春光学精密机械与物理研究所 | Automatic laser three-dimensional scanning track planning method |
| CN116476070A (en) * | 2023-05-22 | 2023-07-25 | 北京航空航天大学 | Method for adjusting scanning measurement path of large-scale barrel part local characteristic robot |
| CN116512806A (en) * | 2023-04-28 | 2023-08-01 | 华侨大学 | Robot scanning viewpoint planning method and system for stone carving cutting workpiece measurement |
-
2023
- 2023-11-02 CN CN202311448811.3A patent/CN117444968A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019157924A1 (en) * | 2018-02-13 | 2019-08-22 | 视辰信息科技(上海)有限公司 | Real-time detection method and system for three-dimensional object |
| CN113155054A (en) * | 2021-04-15 | 2021-07-23 | 西安交通大学 | Automatic three-dimensional scanning planning method for surface structured light |
| CN115325962A (en) * | 2022-08-26 | 2022-11-11 | 中国科学院长春光学精密机械与物理研究所 | Automatic laser three-dimensional scanning track planning method |
| CN116512806A (en) * | 2023-04-28 | 2023-08-01 | 华侨大学 | Robot scanning viewpoint planning method and system for stone carving cutting workpiece measurement |
| CN116476070A (en) * | 2023-05-22 | 2023-07-25 | 北京航空航天大学 | Method for adjusting scanning measurement path of large-scale barrel part local characteristic robot |
Non-Patent Citations (1)
| Title |
|---|
| 张晓蕾;梁延德;何福本;王瑞锋;: "基于动态路径规划的三维扫描方法设计与应用", 组合机床与自动化加工技术, no. 03, 20 March 2018 (2018-03-20), pages 54 - 56 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106166750B (en) | A kind of modified D* mechanical arm dynamic obstacle avoidance paths planning method | |
| US5471541A (en) | System for determining the pose of an object which utilizes range profiles and synethic profiles derived from a model | |
| US10060857B1 (en) | Robotic feature mapping and motion control | |
| CN114571153B (en) | Weld joint identification and robot weld joint tracking method based on 3D point cloud | |
| CN109345620A (en) | Merge the improvement ICP object under test point cloud method of quick point feature histogram | |
| CN105469404B (en) | A kind of rotary body approximating method and device based on three dimensional point cloud | |
| CN113096094B (en) | Three-dimensional object surface defect detection method | |
| CN108356819A (en) | Based on the industrial machinery arm Collision Free Path Planning for improving A* algorithms | |
| CN108765489A (en) | A kind of pose computational methods, system, medium and equipment based on combination target | |
| CN113192054A (en) | Method and system for detecting and positioning complex parts based on 2-3D vision fusion | |
| WO2023103621A1 (en) | Ship weld joint feature recognition method | |
| CN112907683B (en) | Camera calibration method and device for dispensing platform and related equipment | |
| CN108955520B (en) | Structured light three-dimensional scanning accessibility analysis method and analysis system | |
| CN110455187A (en) | A kind of detection method of the box body workpiece weld seam based on 3D vision | |
| CN116402866A (en) | Point cloud-based part digital twin geometric modeling and error assessment method and system | |
| CN115325962A (en) | Automatic laser three-dimensional scanning track planning method | |
| Jin et al. | Computing tëichmuller shape space | |
| CN110431498A (en) | The adquisitiones and welding robot system of welding bead information | |
| JPH09212643A (en) | Method for recognition of three-dimensional object and device therefor | |
| CN117444968A (en) | Method for planning scanning measurement path of large-scale barrel part local characteristic robot | |
| JPH05104465A (en) | Work planning device for robot | |
| JP7093680B2 (en) | Structure difference extraction device, structure difference extraction method and program | |
| CN113592976B (en) | Map data processing method and device, household appliance and readable storage medium | |
| CN112818428B (en) | Light full-automatic scanning path planning method for CAD model surface structure | |
| Zhang et al. | Design and Research of Low‐Cost and Self‐Adaptive Terrestrial Laser Scanning for Indoor Measurement Based on Adaptive Indoor Measurement Scanning Strategy and Structural Characteristics Point Cloud Segmentation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |