Model processing method, device and storage medium based on singularity point topology alignment

Abstract

The application discloses a model processing method and device based on singular point topology alignment and a storage medium, and relates to the technical field of model processing. The model processing method based on singular point topology alignment comprises the following steps: constructing a first normal family structure corresponding to a half-edge structure of a model grid, wherein the first normal family structure comprises at least one first normal family, and the first normal family comprises at least one dual chord; searching for a helical dual chord in the model grid based on a first normal family with a width greater than a preset width threshold; searching for and eliminating a helical path of the helical dual chord to eliminate a helix in the model grid; constructing a second normal family structure of the model grid after the helix is eliminated; and processing the second normal family structure based on a chord extraction and / or chord complement to align non-helical dual chords of the model grid, so as to obtain a topologically aligned model grid. The application reduces singular point slicing of a full quadrilateral grid, improves the slicing size, and thus improves the structuring degree and grid quality of the grid.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of model processing, in particular to a model processing method and device based on singular point topology alignment and a storage medium. BACKGROUND

[0002] For a full quadrilateral mesh, in addition to the number of singular points, the size and number of singular point patches are also important indicators of its structured characteristics. When performing cell quality optimization on a full quadrilateral mesh, the structure of the original mesh is often maintained, and the size and number of patches are not changed. However, if the size of the patch is too small during the optimization process, the optimization effect will be poor. SUMMARY

[0003] The main purpose of the present application is to provide a model processing method and device based on singular point topology alignment and a storage medium, which aims to solve the technical problem of poor model mesh quality caused by too small patches in the mesh optimization process.

[0004] To achieve the above purpose, the present application provides a model processing method based on singular point topology alignment, which comprises: in response to the triggering operation of a model optimization control, acquiring a data file of a model mesh to be optimized, and constructing a first normal family structure corresponding to the half-edge structure of the model mesh, the first normal family structure comprising at least one first normal family, and the first normal family comprising at least one dual chord; based on the first normal family with a width greater than a preset width threshold, searching for a helical dual chord in the model mesh; searching for and eliminating the helical path of the helical dual chord to eliminate the helix in the model mesh; constructing a second normal family structure of the model mesh after eliminating the helix; based on the chord extraction and / or chord supplement processing the second normal family structure, aligning the non-helical dual chord of the model mesh, obtaining the topologically aligned model mesh, and rendering and displaying the topologically aligned model mesh.

[0005] In an embodiment, the step of searching and eliminating the spiral path of the spiral dual string to eliminate the spiral in the model mesh comprises: searching the spiral path of the spiral dual string; and eliminating the spiral path to eliminate the spiral in the model mesh. The step of searching the spiral path of the spiral dual string comprises: constructing a second undirected graph based on the half-edge structure of the model mesh, the second undirected graph being a three-fold graph containing each half-edge of a cell, and the nodes of the second undirected graph being the traveling directions of each half-edge, the traveling directions including a left turn, a right turn, and a straight walk; screening a starting node and a terminal node on the spiral dual string based on the first turn and the last turn of the spiral dual string, the second undirected graph, and a preset rule, to obtain a first path, wherein the preset rule comprises a node uniqueness rule corresponding to a surface cell, a legal node rule corresponding to a spiral direction, and an illegal node exclusion rule corresponding to a feature edge; searching a second path from the terminal node to the starting node based on a breadth-first algorithm and in combination with the preset rule; and constructing the spiral path according to the first path and the second path.

[0006] In an embodiment, after the step of eliminating the spiral path, the method further comprises: determining the total number of spirals and the total number of turns of the spiral dual string in the model mesh; and if the total number of spirals and the total number of turns are both greater than or equal to the total number of spirals and the total number of turns before the spiral path is eliminated, performing the step of searching and eliminating the spiral path of the spiral dual string to eliminate the spiral in the model mesh.

[0007] In addition, to achieve the above object, the present application further provides a model processing device based on singular point topology alignment, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the computer program is configured to implement the steps of the model processing method based on singular point topology alignment.

[0008] In addition, to achieve the above object, the present application further provides a storage medium, which is a computer readable storage medium, and a program for implementing the model processing method based on singular point topology alignment is stored on the computer readable storage medium, and the program for implementing the model processing method based on singular point topology alignment is executed by a processor to implement the steps of the model processing method based on singular point topology alignment.

[0009] The application provides a model processing method based on singular point topology alignment. The application constructs a first normal family structure corresponding to a half-edge structure of a model grid, the first normal family structure comprises at least one first normal family, and the first normal family comprises at least one dual chord. A helical dual chord in the model grid is searched based on a first normal family with a width greater than a preset width threshold. A helical path of the helical dual chord is searched and eliminated to eliminate a helix in the model grid. A second normal family structure of the model grid after the helix is eliminated is constructed. The second normal family structure is processed based on chord extraction and / or chord supplement to align non-helical dual chords of the model grid, and a topologically aligned model grid is obtained. That is, the application aligns non-helical singular points in the grid by operating on the dual chords, reduces singular point segmentation of the full quadrilateral grid, improves the segmentation size, and thus improves the structured degree and the grid quality of the grid. BRIEF DESCRIPTION OF DRAWINGS

[0010] The drawings incorporated in and forming a part of the specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the application.

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced as follows. Obviously, those skilled in the art can obtain other drawings from these drawings without any creative effort.

[0012] Figure 1 A flowchart is provided for the model processing method based on singular point topology alignment according to Embodiment One of the present application. Figure 2 A singular point segmentation diagram is provided for the model processing method based on singular point topology alignment according to Embodiment Two of the present application. Figure 3 A diagram of a Helic type model grid before optimization in Embodiment Nine of the model processing method based on singular point topology alignment according to the present application is provided. Figure 4 A diagram of a Helic type model grid after optimization in Embodiment Nine of the model processing method based on singular point topology alignment according to the present application is provided. Figure 5 A diagram of a non-rotating type model grid before optimization in Embodiment Nine of the model processing method based on singular point topology alignment according to the present application is provided. Figure 6 A diagram of a non-rotating type model grid after optimization in Embodiment Nine of the model processing method based on singular point topology alignment according to the present application is provided. Figure 7 A diagram of a model grid before optimization in Embodiment Nine of the model processing method based on singular point topology alignment according to the present application is provided. Figure 8 A diagram of a model grid after optimization in Embodiment Nine of the model processing method based on singular point topology alignment according to the present application is provided. Figure 9A schematic diagram of fairing processing of the model mesh after optimization in the model processing method based on singular point topology alignment of the present application in embodiment nine; Figure 10 A comparison schematic diagram of the model mesh before and after optimization in the model processing method based on singular point topology alignment of the present application in embodiment nine; Figure 11 A schematic diagram of the model mesh of an automobile part before optimization in the model processing method based on singular point topology alignment of the present application in embodiment nine; Figure 12 A schematic diagram of the model mesh of an automobile part after optimization in the model processing method based on singular point topology alignment of the present application in embodiment nine; Figure 13 A schematic diagram of the model mesh of another automobile part before optimization in the model processing method based on singular point topology alignment of the present application in embodiment nine; Figure 14 A schematic diagram of the model mesh of another automobile part after optimization in the model processing method based on singular point topology alignment of the present application in embodiment nine; Figure 15 A schematic diagram of the hardware structure involved in the model processing device based on singular point topology alignment of the present application. DETAILED DESCRIPTION

[0013] It should be understood that the specific embodiments described herein are merely intended to explain the technical solutions of the present application, and are not intended to limit the present application. In order to better understand the technical solutions of the present application, the specific embodiments will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0014] At present, when optimizing the unit quality of a full quadrilateral mesh, the structure of the original mesh is often maintained, and the size and number of patches will not be changed. However, if the size of the patch is too small during the optimization process, the optimization effect will be poor. The main solution of the present application is to construct a first regular family structure corresponding to the half-edge structure of the model mesh, wherein the first regular family structure comprises at least one first regular family, and the first regular family comprises at least one dual chord. Based on the first regular family with a width greater than a preset width threshold, a spiral dual chord in the model mesh is searched and obtained. The spiral path of the spiral dual chord is searched and eliminated to eliminate the spiral in the model mesh. A second regular family structure of the model mesh after eliminating the spiral is constructed. The second regular family structure is processed based on chord extraction and / or chord supplement to align the non-spiral dual chord of the model mesh, and a topologically aligned model mesh is obtained. The present application aligns the non-spiral singular points in the mesh by operating on the dual chord, reduces the singular point patches of the full quadrilateral mesh, improves the patch size, and thus improves the structured degree and mesh quality of the mesh.

[0015] It should be noted that the execution subject of the embodiment can be a terminal device, or a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, or a model processing device based on singular point topology alignment capable of realizing the above functions, and the embodiment does not make specific limitations. The following takes the model processing device based on singular point topology alignment as an example to illustrate the embodiment and the following embodiments.

[0016] Based on this, the first embodiment of the present application proposes a model processing method based on singular point topology alignment, please refer to Figure 1 , the model processing method based on singular point topology alignment includes steps S10-S50:

[0017] Step S10, in response to the triggering operation of the model optimization control, the data file of the model mesh to be optimized is obtained, and the first normal family structure corresponding to the half-edge structure of the model mesh is constructed, the first normal family structure contains at least one first normal family, and the first normal family contains at least one dual chord.

[0018] In the embodiment, the model optimization control refers to the control that triggers the model mesh quality optimization in the user-oriented operation section of the model processing software. The model mesh is a structured mesh composed of multiple quadrilateral elements in a certain topological relationship, used to carry the geometric shape and topological information of the model. The half-edge structure is a basic data structure for describing the topology of the mesh, each physical edge corresponds to two half-edges with opposite directions, each half-edge is associated with a unique face element and adjacent half-edges, and the association index of vertices, edges and faces is established. The first normal family structure is a collection of dual chords after classification in the model mesh, used to integrate dual chords with the same topological characteristics, facilitate subsequent spiral search and topological alignment processing. The first normal family is the basic unit of the first normal family structure, composed of dual chords that meet certain topological conditions, and is the main target range of spiral search. The dual chord is to regard each quadrilateral element of the model mesh as a node of an undirected graph, and the connection relationship of adjacent elements as an edge of the undirected graph. After forming the undirected graph, the next step direction of each vertex is the path of walking straight in the last step.

[0019] As an optional implementation, the user imports the data file of the model mesh to be optimized, the model processing device based on singular point topology alignment reads the data file and renders and displays the model mesh on the display interface, displays the prompt information of the model mesh to be optimized on the display interface through singular point analysis, and the user clicks the model optimization control to construct the first normal family structure corresponding to the half-edge structure of the model mesh based on the data file.

[0020] As an optional implementation, a first undirected graph is first constructed based on all quadrilateral elements of the model mesh, each quadrilateral element being taken as a node of the first undirected graph, and an edge connection is established between corresponding nodes if two quadrilateral elements share a non-boundary edge. A plurality of continuous paths are tracked by traversing the first undirected graph from any node according to a rule that a next step direction of each vertex is straight walking from a previous step, each path being a dual chord. All dual chords are traversed, and a direction vector of each dual chord is calculated to determine that the dual chords with a direction vector included angle less than a preset angle threshold are mutually parallel dual chords. The dual chords emitted from mutually parallel and adjacent edge non-singular points are merged into a first normal family, and all first normal families are integrated to form a first normal family structure.

[0021] As another optional implementation, singular points are first identified from the model mesh, and a mesh line of each singular point is extended to a mesh boundary or another singular point to form a plurality of singular point patches. A first undirected graph corresponding to each singular point patch is constructed, each quadrilateral element in the singular point patch being taken as a node of the first undirected graph, and an edge connection is established between adjacent elements. Dual chords in each singular point patch are searched according to a straight walking rule, and parallel relationships of the dual chords in the singular point patch are determined to be merged into a first normal family in the singular point patch. All first normal families of the singular point patches are integrated to form a first normal family structure covering the entire model mesh.

[0022] Step S20, a helical dual chord in the model mesh is searched based on the first normal family with a width greater than a preset width threshold.

[0023] In the embodiment, the preset width threshold is a critical value for determining whether the first normal family has a condition for forming a helix, which is set by a user according to mesh accuracy requirements and engineering experience, and is usually 1 or 2. The helical dual chord refers to a dual chord whose element sequence is long enough to coincide with the original dual chord after the element is shifted to the left or right by a preset number of elements, which will cause a helical distortion of the mesh.

[0024] As an optional implementation, a width parameter of all first normal families in the first normal family structure is first obtained, the width parameter being a number of dual chords contained in each first normal family. A first normal family with a width parameter greater than a preset width threshold is marked as a candidate normal family, all dual chords in each candidate normal family are traversed to determine a current traversed dual chord as a target dual chord. A preset number is set to 1, all quadrilateral elements corresponding to the target dual chord are shifted to the left by 1 element to obtain a left element sequence, and shifted to the right by 1 element to obtain a right element sequence. The number of coincident elements of the left element sequence, the right element sequence and the target dual chord is counted respectively, and if the number of coincident elements of any sequence is greater than or equal to a preset coincidence threshold, the target dual chord is determined to be a helical dual chord.

[0025] As another optional implementation, the preset width threshold is set to 2, and the preset number is set to 1 and 2. First, a first normal family with a width greater than 2 is screened out as a candidate normal family. The parallelism parameter of the candidate normal family is used to calculate the distance between each pair of dual strings in the family. Based on the distance, a target region in which the dual string is translated by 1 or 2 units is determined, without globally traversing the grid. For each target dual string, left and right side unit sequences translated by 1 and 2 units are respectively generated, and the number of coincident units of each sequence with the original dual string is counted. If the number of coincident units of any sequence meets a preset coincidence threshold, the target dual string is marked as a spiral dual string, and the corresponding translation number and coincidence length are recorded.

[0026] In step S30, the spiral path of the spiral dual string is searched and eliminated to eliminate the spiral in the model grid.

[0027] In this embodiment, the spiral path refers to a path that can eliminate the spiral distortion corresponding to the spiral dual string, which is a legal loop in the second undirected graph. The second undirected graph is a triple graph constructed based on the half-edge structure of the model grid, used to carry the marching direction association relationship of the half-edge, and provides a topological basis for spiral path searching. The triple graph refers to a graph structure in which each half-edge corresponds to three nodes, corresponding to three marching directions. The marching direction includes left turn, right turn and straight walk, which is the turning selection when moving along the half-edge in the grid unit. The preset rule is a constraint condition for screening legal nodes and paths, including the face unit node uniqueness rule, the legal node rule corresponding to the spiral rotation direction, and the illegal node exclusion rule corresponding to the feature edge. The first path is an initial path from the starting node to the terminating node, which is screened based on the first and last turns of the spiral dual string. The second path is a path searched from the terminating node to the starting node based on the breadth-first algorithm, which together with the first path forms the spiral path.

[0028] As an optional implementation, based on the half-edge structure of the model grid, three nodes of left turn, right turn and straight walk are created for each half-edge, the connections between the nodes are established according to the grid topological relationship, and the second undirected graph is formed. For each spiral dual string, the half-edges corresponding to the first and last turns are extracted, and the legal starting node on the first turn and the legal terminating node on the last turn are screened out according to the preset rule. The first path is obtained by connecting the starting node and the terminating node. Based on the breadth-first algorithm, the legal nodes in the second undirected graph are traversed from the terminating node to the starting node direction, and the spiral path is searched. When the starting node is searched, the search is stopped, the second path is obtained, the first path and the second path are connected, and the closed spiral path is formed. The grid edges and face units corresponding to the spiral path are determined, and the spiral distortion is eliminated by topological adjustment of these edges and face units.

[0029] As another optional implementation, when constructing the second undirected graph, the illegal nodes corresponding to the hard edges are batch-labeled in combination with the characteristic edge distribution of the first normal family, so as to simplify the judgment of the legality of the nodes. The l-type corresponding nodes of the first circle and the last circle of the helix dual string are extracted, the first path is generated based on the parallelism parameter of the normal family, and it is ensured that the first path extends along the legal direction in the normal family. The optimized breadth-first algorithm is used to preferentially search the nodes in the same normal family, reduce invalid search across families, and in combination with the uniqueness constraint of the face unit node in the preset rule, the legality of the path is verified in real time. After the second path is searched, the helical path is formed by merging the first path, the half-edge and the face unit involved in the helical path are topologically reconstructed to eliminate the helix, and the half-edge structure of the mesh is updated synchronously.

[0030] Step S40, constructing the second normal family structure of the model mesh after the helix is eliminated.

[0031] In this embodiment, the second normal family structure is a collection of the dual strings in the model mesh after the helix distortion is eliminated and reclassified, which is used for subsequent topological alignment processing of the non-helical dual strings.

[0032] As an optional implementation, after the helix is eliminated, the half-edge structure of the model mesh is updated globally to ensure the association between the vertices, edges and faces. Based on the updated half-edge structure, a second undirected graph covering the entire model mesh is reconstructed, each quadrilateral unit is taken as a node, and an edge connection is established between adjacent units. All dual strings are searched again according to the straight rule, the direction vectors of the dual strings are calculated, the parallel relationship is determined, the parallel dual strings of the non-singular point emitting edge are merged, and multiple second normal families are formed. All the second normal families are integrated to form the second normal family structure, and the width, parallelism parameter and characteristic edge distribution of each second normal family are recorded.

[0033] As another optional implementation, after the helix is eliminated, the dual strings are not reconstructed globally, but are updated locally based on the first normal family structure. For the first normal family containing the dual strings eliminated by the helix, the dual strings affected by the helix elimination in the first normal family are modified, including adjusting the path and updating the characteristic edge label. For other first normal families not affected, the dual strings and the classification relationship thereof are directly retained. The parallelism and adjacency of the dual strings in all normal families are re-determined, the normal families are merged or split, and finally the second normal family structure is formed to reduce repeated calculation.

[0034] Step S50, processing the second normal family structure based on the extraction of the dual strings and / or the complementary strings, aligning the non-helical dual strings of the model mesh, obtaining the topologically aligned model mesh, and rendering and displaying the topologically aligned model mesh.

[0035] In this embodiment, string extraction refers to the operation of removing the dual string that meets the conditions in the second normal family structure and adjusting the positions of the left and right sides to achieve the alignment of non-helical dual strings. String supplement refers to the operation of adding a dual string to adjust the grid topology to achieve the alignment of non-helical dual strings in the second normal family structure. Non-helical dual string refers to a dual string in the model grid that does not form a helical structure. It may have topological dislocation due to singular points and needs to be aligned through string extraction or string supplement. The topologically aligned model grid is a model grid with regular topology and optimized aspect ratio after helical elimination and non-helical dual string alignment.

[0036] As an optional implementation, the second normal family structure is processed based on string extraction and / or string supplement. There are three processing methods. The first method is to extract the second normal family structure to align the non-helical dual strings of the model grid, obtain the topologically aligned model grid, and then render and display the topologically aligned model grid on the display interface. The second method is to supplement the second normal family structure to align the non-helical dual strings of the model grid, obtain the topologically aligned model grid, and then render and display the topologically aligned model grid on the display interface. The third method is to first extract the second normal family structure, then supplement the second normal family structure, align the non-helical dual strings of the model grid, obtain the topologically aligned model grid, and then render and display the topologically aligned model grid on the display interface.

[0037] As an optional implementation, all second normal families in the second normal family structure are traversed to identify the second normal family with a width of 1, which corresponds to the dual string to be processed. For each dual string to be processed, all left and right sides are traversed to determine whether each side is a characteristic edge. If at least one side in each edge pair is a non-characteristic edge, the string extraction action is performed. If both left and right sides are non-characteristic edges, the middle position of the left and right sides is taken as a new edge, the topology of the adjacent unit is adjusted, and the original dual string to be processed is removed. If only the left side is a non-characteristic edge, the left side is translated to the position of the right side. If only the right side is a non-characteristic edge, the right side is translated to the position of the left side. All dual strings that are transversely intersected with the extracted dual string are traversed to check whether the characteristic edge has changed due to string extraction. If it has changed, the characteristic edge information is updated. For the dual string to be processed that cannot be extracted, i.e., there is an edge pair in which both left and right sides are characteristic edges, based on the position information of the dual string to be processed in the model grid, a suitable string supplement position is selected, a new dual string is added, the local grid topology is adjusted, and alignment is achieved.

[0038] As another optional implementation, the second normal family structure is classified into a drawable string normal family and a non-drawable string normal family. The drawable string normal family is a normal family with a width of 1 and satisfying all left and right edge pairs of at least one non-characteristic edge, and is classified and processed according to the left and right edge mobility. For the dual string with both left and right mobility, a weighted average method is used to determine the new edge position, and the weight is set based on the length and topological importance of the edge. For the dual string with only one side mobility, the movable edge is translated by a fixed offset, to ensure that the aspect ratio of the grid cell after translation is within a reasonable range. The non-drawable string normal family is a normal family with a width of 1 and having left and right edges that are both characteristic edges, and a complementary string method of local grid reconstruction is used. According to the distribution of the cells around the dual string, one or more dual strings parallel to the dual string are added, the original normal family with a width of 1 is split, the connection relationship of the adjacent cells is adjusted, and the alignment of the non-helical dual string is realized.

[0039] Optionally, after the string drawing and / or complementary string operation, the model grid is subjected to fairing processing, and the overall grid optimization is performed after the fairing processing. Optionally, the fairing processing can adopt a spring fairing method, a diffusion fairing method, a linear elasticity fairing method, a Laplace fairing method, an energy minimization fairing method, a curvature-driven fairing, etc. Other fairing processing methods can also be used, and the present embodiment does not make specific limitations thereto.

[0040] For example, in the optimization of the structured grid of an aero-engine blade cooling passage, a user clicks the model grid optimization control, and the device responds to the control to obtain the original model grid of the blade cooling passage, which is composed of thousands of quadrilateral elements and has multiple topological distortions caused by singular points. The half-edge structure of the model grid is constructed, the grid is split into three generalized rectangular patches without internal singular points through singular point slicing, a first undirected graph is constructed in each patch, and 12 dual chords are searched and merged to form four first regular families, two of which have a width of 3, which is greater than the preset width threshold 2, and are marked as candidate regular families. The translation search is performed on the 6 dual chords in the candidate regular family, the preset number is set to 1 and 2, and the preset coincidence threshold is 3 elements. Finally, 3 spiral dual chords are searched, 2 of which are k = 1, and 1 is k = 2. The second undirected graph is constructed based on the half-edge structure, each half-edge corresponds to 3 nodes, a total of 4800 nodes are generated, the legal nodes of the first and last circles of the spiral dual chord are screened according to the preset rule to generate the first path, the second path is searched using the breadth-first algorithm to form 3 spiral paths, and the topological adjustment is performed on each spiral path to eliminate the corresponding spiral distortion. After the spiral is eliminated, the second regular family structure is reconstructed to obtain 5 second regular families, 3 of which are second regular families with a width of 1 corresponding to unaligned non-spiral dual chords. For the two width-1 regular families that meet the chord extraction condition, the chord extraction operation is performed, one of which has a movable left and right edge, and the other of which has a movable right edge. The left edge is translated to the position of the right edge. For the width-1 regular family that cannot be extracted, a new dual chord is added for chord supplementing to adjust the local grid topology. Finally, the topologically aligned model grid of the blade cooling passage is obtained, the aspect ratio of the grid elements is controlled to be less than 5, and there is no obvious spiral distortion and topological dislocation.

[0041] The embodiment realizes the ordered management of the dual chord by constructing the regular family structure, narrows the spiral search range, and improves the efficiency and accuracy of spiral identification. The spiral path search based on the triple graph and the breadth-first algorithm ensures the accuracy of spiral elimination and avoids introducing new topological distortions. Through the combination of chord extraction and chord supplementing, the effective alignment of non-spiral dual chords is realized, and the topological structure of the model grid is optimized. The entire process is carried out around the singular point topological alignment, from preprocessing to spiral elimination and then to topological optimization, forming a complete technical closed loop, which significantly improves the aspect ratio distribution of the structured grid and improves the grid quality. It provides a high-quality grid basis for subsequent numerical simulation, engineering analysis and other applications, while reducing the computational cost of grid optimization and improving the processing efficiency.

[0042] Based on the above embodiment one, in the second embodiment of the present application, before step S10, the following steps are included:

[0043] Step S01, determine the half-edge structure of the model grid.

[0044] In this embodiment, the singular point refers to a vertex in the model mesh whose number of connected edges is not equal to 4, which will cause the mesh topology distortion and is the core object of subsequent slicing processing. The half-edge structure is a basic data structure for describing the mesh topology. Each physical edge corresponds to two opposite direction half-edges. Each half-edge is associated with a unique face element and an adjacent half-edge, and the association index of the vertex, edge and face is established.

[0045] As an optional implementation, the original data file of the model mesh is read, the vertex coordinates, edge vertex association information and face edge association information are extracted, two opposite direction half-edge records are generated for each physical edge, each half-edge record contains the face element index, adjacent half-edge index and vertex index, a complete half-edge structure data table is constructed, and fast association query of the vertex, edge and face is realized.

[0046] As another optional implementation, the half-edge structure is constructed in real time based on the model mesh generation algorithm. During the mesh element generation process, a half-edge corresponding to each edge of a quadrilateral face element is generated synchronously after the creation of the face element. The correct association of the two half-edges of each physical edge is ensured through topology relationship verification, and the adjacent relationship and the face information of the half-edge are updated in real time to generate a dynamically maintained half-edge structure.

[0047] Step S02, the mesh line of the singular point in the model mesh is extended towards the mesh boundary direction so that the mesh line passes through all the singular points and the end point of the mesh line is located at the mesh boundary.

[0048] In this embodiment, the mesh line refers to a continuous line segment formed by sequentially connecting vertices in the model mesh, including the boundary mesh line and the internal mesh line, which is the skeleton structure of the mesh element. Referring to Figure 2 , Figure 2 An example of singular point slicing is shown. The purple line is the mesh line derived from the singular point.

[0049] As an optional implementation, all vertices of the model mesh are traversed to identify all singular points. For each singular point, the mesh line corresponding to all edges associated with the singular point is obtained, and vertices and edges are added along the extension direction of each mesh line until the extended mesh line reaches the boundary of the model mesh or intersects with the mesh line extended by other singular points. The newly added vertex, edge and half-edge information in the extension process are recorded.

[0050] As another optional implementation, for each singular point, the adjacent face elements and edges are queried through the half-edge structure to determine the extension direction vector of the mesh line. Based on the direction vector, an extension path is generated, and it is detected whether there is another singular point on the extension path. If there is, the extension is performed to the singular point. If not, the extension is performed to the mesh boundary. The topological continuity of the newly added mesh line and the original mesh line is ensured during the extension process.

[0051] Step S03, dividing the model mesh into a plurality of singular point patches according to the extended mesh lines, so as to take each singular point patch as an updated model mesh.

[0052] In the embodiment, the singular point patch refers to an independent mesh region obtained by dividing the original model mesh by the extended mesh lines, each patch does not contain an inner singular point, only singular points can exist on the boundary, and has a generalized rectangular structure. As an optional implementation, all the extended mesh lines are collected, the coordinates of the intersection points of the mesh lines are determined, the intersection points are taken as the vertices of the patch boundary, adjacent intersection points are connected in a clockwise or counterclockwise order to form a closed patch boundary, and the mesh region surrounded by each closed boundary is a singular point patch. The vertex, edge, face element and half-edge information in each patch are extracted to form independent patch mesh data.

[0053] As another optional implementation, the region growing method is used for patch segmentation, the extended mesh lines are taken as boundary constraints, and from an arbitrary non-boundary face element of the original model mesh, adjacent face elements are sequentially included in the current patch until expansion is impossible (the extended mesh line is encountered), the segmentation of one patch is completed, and the process is repeated until all face elements are assigned to the corresponding singular point patch. Each patch forms independent model mesh data.

[0054] In the embodiment, the topology alignment can be performed on the entire model mesh, or the model mesh can be determined based on the singular point patch, and then each singular point patch is taken as a model mesh. Each patch can be processed in parallel to improve the processing efficiency.

[0055] Based on any of the above embodiments, in the third embodiment of the present application, the step of constructing the first regular family structure corresponding to the half-edge structure of the model mesh includes: step S11, constructing a first undirected graph based on the model mesh, the elements of the model mesh being taken as the nodes of the first undirected graph, and the element connection relationship being taken as the edges of the first undirected graph. Step S12, searching a plurality of dual chords according to the first undirected graph. Step S14, constructing a first regular family based on at least one first regular family.

[0056] In the embodiment, the first undirected graph is a graph structure representing the topological relationship between the model grid cells, the nodes correspond one-to-one to the quadrilateral cells of the model grid, and the edges represent the adjacent relationship between the cells. The cell connection relationship refers to the adjacent relationship formed by two quadrilateral cells sharing a non-boundary edge, which is the core basis for constructing the undirected graph edge. The dual chord is a continuous node path formed in the first undirected graph by following the rule that the next step direction of each vertex is straight from the previous step, reflecting the ordered topological trajectory of the grid cell. Parallel to each other means that the angle between the direction vectors of two dual chords is less than a preset threshold (usually set to 5°), and the dual chords that meet this condition have the same topological direction characteristic. The first normal family structure is a global topological set formed by integrating all first normal families, which is used to centrally manage dual chords with the same direction characteristic and provide classification support for subsequent spiral search.

[0057] As an optional implementation, all quadrilateral cells of the model grid are traversed, each cell is assigned a unique node identifier and a mapping relationship is established, the adjacent cells of each cell sharing a non-boundary edge are obtained through a grid topology query tool, undirected edges are established between the corresponding nodes, and finally a complete first undirected graph is formed. Starting from any node of the first undirected graph, the adjacent nodes are traced in turn according to the "straight" rule, i.e., each time only the adjacent node consistent with the direction of the previous step is selected, until the path reaches the boundary of the undirected graph or forms a closed loop, the tracing is stopped and the path is recorded as a dual chord, and the process is repeated until all nodes are traversed, and all dual chords are collected. The direction vector of each dual chord is calculated, the parallel relationship of any two dual chords is determined through a vector angle calculation tool, all mutually parallel dual chords are merged into a set, each set is a first normal family, and finally at least one first normal family is formed. All formed first normal families are collected, each normal family is assigned a unique identifier, the core attributes of each normal family including the dual chord index, direction vector, and width are recorded, all normal family information is integrated according to a preset data structure, and a complete first normal family structure is constructed.

[0058] Exemplarily, in the optimization scenario of the structured mesh of an aero-engine blade cooling passage, the model mesh is a quadrilateral structured mesh of the cooling passage, containing 2000 quadrilateral elements. All 2000 elements are traversed and assigned with unique node identifiers, the adjacent elements of each element are retrieved through a topology query tool (each element is associated with an average of 4 adjacent elements), undirected edges are established between the corresponding nodes, and finally a first undirected graph containing 2000 nodes and 4000 edges is constructed. Starting from the node 10 of the undirected graph (corresponding to the first element at the entrance of the cooling passage), the adjacent nodes are continuously tracked according to the "straight walk" rule to form the first pair of chords with a length of 45, and the process is repeated until all nodes are traversed, and a total of 60 pairs of chords are collected. The direction vector of each pair of chords is obtained through a vector calculation tool, and the pairs of chords with an included angle less than 5° are merged to finally form 8 first normal families, of which 4 normal families each contain 8 pairs of chords, and 4 normal families each contain 7 pairs of chords. Each normal family is assigned with an identifier F1-F8, and the core attributes of each normal family such as the index of the pair of chords, the direction vector (such as the direction vector of F3 is (0.99, 0.12)), and the width are recorded, all normal family information is integrated according to a tree-shaped data structure, and a complete first normal family structure is constructed.

[0059] It should be noted that the number of mesh elements, the number of pairs of chords and the number of normal families in the above example are only exemplary expressions, which can be flexibly adjusted according to the complexity and size of the model mesh in actual application. The core logic of steps S11-S14 is not affected by the size of the mesh, and the effective classification and management of the pairs of chords can be realized.

[0060] In the embodiment, the constructed first undirected graph, the dual chord and the first normal family structure provide a core topological infrastructure for the overall scheme, convert the scattered model grid cells into an ordered topological set, solve the problem of "no classification basis" in the subsequent spiral dual chord search, and make the overall scheme form an effective link from grid preprocessing to spiral elimination. Secondly, the first undirected graph clearly defines the adjacent relationship of the grid cells, ensures the accuracy of the dual chord search, avoids the path disorder caused by the ambiguous topological relationship, provides accurate classification basis for the subsequent width-based screening of candidate normal families, significantly reduces the search range of the spiral dual chord, and improves the efficiency of the subsequent spiral identification. Thirdly, the centralized management of parallel dual chords by the first normal family structure enables accurate positioning of the topological area to which the spiral dual chord belongs during the subsequent spiral elimination process, reduces invalid searches across regions, provides data support for fast search and elimination of spiral paths, and reduces the complexity of spiral elimination. Finally, the topological structure formed by the steps provides a regular topological foundation throughout the overall scheme, lays a regular topological foundation for the construction of the second normal family structure after the elimination of the spiral and the subsequent topological alignment of the extracted and supplemented chords, ensures the topological consistency of each step in the overall scheme, and ultimately improves the overall quality and efficiency of the model grid topological alignment, providing high-quality grid support for subsequent numerical simulation, engineering analysis and other applications.

[0061] Based on any of the above embodiments, in the fourth embodiment of the present application, step S20 comprises:

[0062] Step S21, for the first normal family with a width greater than the preset width threshold, determining a target dual chord currently traversed.

[0063] In the embodiment, the preset width threshold is a critical value for judging whether the first normal family has the condition of forming a spiral, which is set by the user according to the grid accuracy requirement and engineering experience, and is usually 2. The target dual chord refers to the dual chord currently being traversed and to be determined whether it is a spiral structure in the first normal family with a width meeting the requirements.

[0064] As an optional implementation, first read the width attribute of each first normal family in the first normal family structure, select the normal family with a width greater than the preset width threshold as a candidate normal family, traverse each candidate normal family in a preset traversal order (such as ascending order of normal family identifier), determine the dual chord currently traversed in the normal family as the target dual chord, and repeat until all dual chords in the candidate normal family are traversed.

[0065] Step S22, offsetting the cells of the target dual chord by a preset number of cells to the left and to the right respectively to obtain two cell sequences.

[0066] In the embodiment, the preset number refers to the number of unit offsets of the dual string unit, which is used to generate a comparison sequence, and is usually 1 or 2. The unit sequence refers to a continuous grid unit set formed by offsetting the units of the target dual string in a preset direction, and the length of the unit sequence of the target dual string is consistent.

[0067] As an optional implementation, the preset number is set to 1, all units contained in the target dual string and the unit arrangement order are obtained, based on the grid topology relationship, the left adjacent unit (left side along the direction of the dual string) of each unit is determined, the left adjacent units are arranged in turn according to the unit arrangement order of the target dual string, and a left unit sequence is formed. Similarly, the right adjacent unit of each unit is determined, and a right unit sequence is formed according to the same arrangement order, so as to ensure that the two unit sequences are consistent with the number of units of the target dual string.

[0068] Step S23: If the number of coincident units of at least one of the unit sequences and the target dual string meets a preset threshold, the target dual string is determined to be a spiral dual string.

[0069] In the embodiment, the preset threshold is a critical number for determining whether the unit sequence and the target dual string exist spiral coincidence, which is usually set according to the length of the dual string and is 3-5 units. The number of coincident units refers to the number of grid units that are completely the same in the unit sequence and the target dual string, which reflects the degree of overlap of the two.

[0070] As an optional implementation, the left unit sequence, the right unit sequence and the units of the target dual string are traversed respectively, the number of coincident units of the two sequences and the target dual string is counted through unit unique identifier comparison, and if the number of coincident units of any sequence is greater than or equal to the preset threshold, the target dual string is determined to be a spiral dual string, and key information such as the number of coincident units and the offset direction is recorded.

[0071] For example, in the optimization scenario of the structured grid of the cooling channel of the aero-engine blade, based on the first normal family structure (8 normal families F1-F8) constructed before, the preset width threshold is set to 2, and the candidate normal family is 6 normal families (F1-F6) with a width greater than or equal to 2. F1-F6 are traversed in ascending order of identification, and the first dual string in F1 is determined as the target dual string. The preset number is set to 1, the target dual string contains 45 units, the left and right adjacent units of each unit are obtained in turn, and the left unit sequence and the right unit sequence each containing 45 units are formed. The preset threshold is set to 3, and through unit identifier comparison, it is found that the number of coincident units of the left unit sequence and the target dual string is 4, which meets the preset threshold, and the target dual string is determined to be a spiral dual string. Continue to traverse the dual strings in all candidate normal families, and finally 5 spiral dual strings are identified, all of which meet the condition that the number of coincident units is greater than or equal to 3.

[0072] It should be noted that the preset width threshold, the preset number, and the preset threshold value can be adjusted according to the grid complexity. For example, the preset threshold value can be increased to 5 for a complex grid to ensure the accuracy of the spiral dual string recognition. The core determination logic is not affected by the value. In this embodiment, the preset width threshold is used to screen the candidate normal family, and the "offset-coincidence determination" logic is used to accurately identify the spiral dual string. This not only avoids invalid traversal and improves search efficiency, but also provides a targeted basis for subsequent spiral elimination, ensuring the smoothness and accuracy of the overall topology alignment process.

[0073] Based on any of the above embodiments, in the fifth embodiment of the present application, step S30 comprises:

[0074] Step S31 determines the traversal order of the spiral dual string based on the number of turns of each spiral dual string.

[0075] In this embodiment, the number of turns refers to the number of times the spiral dual string revolves around the center to close, which is a core indicator for measuring the influence range and distortion degree of the spiral. The traversal order refers to the order of searching and eliminating the identified spiral dual string, which directly affects the overall spiral elimination effect.

[0076] As an optional implementation, the number of turns of each spiral dual string is counted, and the traversal order is determined in descending order of the number of turns. The spiral dual string with more turns is processed first. If the number of turns is the same, the width of the normal family to which it belongs is used for supplementary sorting in descending order to ensure that the spiral with a larger influence range is eliminated first.

[0077] Step S32, according to the traversal order, for each spiral dual string, executes the steps of searching and eliminating the spiral path of the spiral dual string to eliminate the spiral in the model grid until the traversal of each spiral dual string is completed.

[0078] In this embodiment, the spiral path refers to a closed topology path that can accurately eliminate the corresponding spiral distortion. The spiral is eliminated by adjusting the grid cells involved in the path.

[0079] As an optional implementation, according to the determined traversal order, each spiral dual string is selected in turn, a triple graph is constructed based on the half-edge structure of the model grid, the spiral path of the spiral dual string is searched according to a preset rule, the grid edges and face cells involved in the path are topologically adjusted to eliminate the spiral, the half-edge structure and dual string information of the grid are updated synchronously, and the next spiral dual string is processed after the current spiral dual string is processed, until all spiral dual strings are processed.

[0080] Exemplarily, in the optimization scenario of the structured grid of the cooling channel of the aero-engine blade, 5 spiral dual strings are previously identified, and the number of turns thereof is 3 turns, 2 turns, 3 turns, 2 turns and 1 turn respectively. In descending order of the number of turns, the spiral dual strings with the same number of turns are sorted according to the width of the normal group to which they belong, and the final traversal order is 3 turns (width of the normal group is 8), 3 turns (width of the normal group is 7), 2 turns (width of the normal group is 8), 2 turns (width of the normal group is 7) and 1 turn. In this order, each spiral dual string is processed in turn, the triple graph is first constructed to search for the path of the first 3-turn spiral, the related grid cells are adjusted to eliminate the spiral, and after the grid structure is updated, the next one is processed, until all 5 spiral dual strings are processed, and no new spiral distortion occurs during the processing.

[0081] It should be noted that the sorting rule of the traversal order can be adjusted according to actual optimization requirements, and the core is to preferentially process the spiral dual string that has a greater impact on the grid quality, to ensure the optimization effect.

[0082] In this embodiment, the traversal order is determined according to the number of turns of the spiral dual string, and the spiral with a wide influence range and serious distortion is preferentially eliminated by processing in turn, which not only improves the pertinence and effect of the overall grid optimization, but also ensures the orderliness of the spiral elimination process and avoids topological confusion caused by disordered processing.

[0083] Based on any of the above embodiments, in the sixth embodiment of the present application, step S30 comprises:

[0084] Step S33, searching for a spiral path of the spiral dual string.

[0085] In this embodiment, the spiral path refers to a closed topological path that can accurately match the distortion characteristics of the spiral dual string, and is the core basis for performing the spiral elimination operation.

[0086] As an optional implementation, the process of “constructing a second undirected graph → screening start and end nodes → searching for a path → integrating the path” is performed, and based on the model grid topological relationship and the preset rules, a closed path capable of eliminating the spiral distortion is gradually constructed.

[0087] Step S34, eliminating the spiral path to eliminate the spiral in the model grid.

[0088] In this embodiment, eliminating the spiral path refers to topologically reconstructing the grid edges and surface elements involved in the spiral path, correcting the distorted topological relationship, and restoring the grid to normal.

[0089] As an optional implementation, all grid edges and surface elements corresponding to the spiral path are identified, the topological connection relationship of these elements is adjusted, the spiral distortion structure in the path is removed, the legal connection of adjacent elements is re-established, the half-edge structure and dual chord information of the model grid are synchronously updated, and the continuity and regularity of the grid topology after elimination are ensured.

[0090] Optionally, the step S33 comprises: a step S331 of constructing a second undirected graph based on the half-edge structure of the model grid, the second undirected graph being a three-fold graph containing each half-edge of the elements, and the nodes of the second undirected graph being the traveling directions of each half-edge, the traveling directions including left turn, right turn and straight walk.

[0091] In this embodiment, the second undirected graph is a topological graph for bearing the association relationship of the half-edge traveling directions, and provides a node connection basis for path searching. The three-fold graph refers to a graph structure in which each half-edge corresponds to three nodes, which respectively map the three traveling directions.

[0092] As an optional implementation, all half-edges of the model grid are traversed, three nodes of left turn, right turn and straight walk are respectively created for each half-edge, and the undirected edges between the nodes corresponding to adjacent half-edges are established according to the topological connection relationship of the grid elements, so as to ensure that the node connection conforms to the actual topological logic of the grid and form a complete second undirected graph.

[0093] The step S332 comprises: screening the start node and the end node on the spiral dual chord based on the first circle and the tail circle of the spiral dual chord, the second undirected graph and a preset rule, to obtain a first path, wherein the preset rule comprises a node uniqueness rule corresponding to a surface element, a legal node rule corresponding to a spiral direction, and an illegal node exclusion rule corresponding to a feature edge.

[0094] In this embodiment, the first circle refers to the outermost closed element circle of the spiral dual chord, and the tail circle refers to the innermost closed element circle of the spiral dual chord. The start node is the starting point of the first path, and the end node is the terminal point of the first path, both of which are legal nodes in the second undirected graph. The first path is an initial path connecting the start node and the end node, and conforms to the topological characteristics of the first and tail circles of the spiral dual chord. The node uniqueness rule of the surface element means that the node corresponding to each surface element in the path appears only once; the legal node rule of the spiral direction means that the node traveling direction needs to be consistent with the spiral direction; and the illegal node exclusion rule of the feature edge means that the node corresponding to the feature edge cannot be used as a path node.

[0095] As an optional implementation, the half-edges corresponding to the first circle and the tail circle of the helical dual string are extracted, the travel direction nodes of the half-edges are matched in the second undirected graph, the illegal nodes are excluded in combination with the preset rule, the nodes corresponding to the non-characteristic edge and meeting the rotation requirement on the first circle are screened as the starting node, the nodes corresponding to the non-characteristic edge on the tail circle are screened as the terminal node, and the first path is formed by connecting the starting node and the terminal node.

[0096] In step S333, the second path is searched from the terminal node to the starting node based on the breadth-first algorithm in combination with the preset rule.

[0097] In the embodiment, the breadth-first algorithm refers to a path search algorithm that starts from the terminal node, traverses adjacent legal nodes layer by layer until the starting node is found, and can ensure the shortest and legal of the search path.

[0098] As an optional implementation, the breadth-first algorithm is enabled to traverse the nodes in the second undirected graph starting from the terminal node, only the legal nodes are reserved and the path is recorded in the traversal process in strict compliance with the preset rule, and the traversal is stopped when the starting node is searched, thereby obtaining the second path from the terminal node to the starting node.

[0099] In step S334, the helical path is constructed according to the first path and the second path.

[0100] In the embodiment, the helical path is a closed path formed by connecting the first path and the second path at the tail and the head, and covers the distortion region of the helical dual string.

[0101] As an optional implementation, the first path (starting node→terminal node) and the second path (terminal node→starting node) are connected at the tail and the head to form a closed helical path, and the path is ensured to have no node conflict and no topological fracture through topological verification, thereby finally obtaining a complete helical path.

[0102] For example, in the scenario of optimizing the structured grid of the cooling channel of an aero-engine blade, for a helical dual string with 3 circles, a second undirected graph is constructed based on the grid half-edge structure, 9600 nodes (4800 half-edges x 2 direction nodes) are generated, and legal node connections are established. The half-edges corresponding to the first circle and the tail circle of the helical dual string are extracted, the starting node (a straight node of a half-edge on the first circle) and the terminal node (a straight node of a half-edge on the tail circle) are screened in combination with the preset rule, and the first path with a length of 12 is obtained by connection. The second path with a length of 10 is searched by traversing the second undirected graph starting from the terminal node through the breadth-first algorithm and complying with the preset rule. When step S37 is performed, the two paths are connected to form a closed helical path. The topological reconstruction is performed on the 32 face units and 64 edges involved in the path to eliminate the helical distortion, and the grid structure is updated synchronously.

[0103] The embodiment constructs a triplex map, combines a breadth-first algorithm to accurately search a spiral path, and then eliminates the spiral through topological reconstruction, thereby ensuring the accuracy and efficiency of the path search, completely correcting the spiral distortion, and providing a regular grid basis for subsequent topological alignment.

[0104] Based on any of the above embodiments, in the seventh embodiment of the present application, after step S34, the following steps are included:

[0105] In step S35, the total number of spiral dual edges and the total number of spiral turns in the model grid are determined.

[0106] In the present embodiment, the total number of spirals refers to the total number of spiral dual edges remaining in the model grid after the current spiral path is eliminated. The total number of spiral turns refers to the sum of the number of turns of all remaining spiral dual edges, which is a comprehensive indicator for measuring the degree of overall spiral distortion.

[0107] As an optional implementation, based on the half-edge structure and normal family information of the updated model grid after the spiral path is eliminated, all normal families are re-traversed to identify and count the number of remaining spiral dual edges as the total number of spirals, and the number of turns of each remaining spiral dual edge is accumulated to obtain the total number of spiral turns.

[0108] In step S36, if the total number of spirals and the total number of spiral turns are both greater than or equal to the total number of spirals and the total number of spiral turns before the spiral path is eliminated, the step of searching and eliminating the spiral path of the spiral dual edge is performed to eliminate the spiral in the model grid.

[0109] In the present embodiment, the total number of spirals and the total number of spiral turns before elimination refer to the number of spiral dual edges and the sum of the number of turns in the model grid recorded before the current spiral path elimination operation is performed, which is used to compare with the indicators after elimination to determine the effectiveness of the elimination operation.

[0110] As an optional implementation, the total number of spirals and the total number of spiral turns after elimination counted in step S39 are compared with the corresponding indicators recorded before elimination. If both are greater than or equal to the indicators before elimination, it indicates that the current elimination operation has not effectively reduced the spiral distortion, and the search and elimination process of steps S33 to S38 needs to be re-executed. If any indicator is less than the indicator before elimination, it is determined that the elimination is effective and the current iteration is stopped.

[0111] For example, for a certain aircraft engine blade cooling channel grid, after the first spiral elimination is completed, the total number of remaining spirals is 3 and the total number of spiral turns is 5, while the total number of spirals before elimination is 3 and the total number of turns is 4. Since the total number of spirals and the total number of turns after elimination are not less than the indicators before elimination, the search and elimination process of the spiral path is restarted, and after the related operations are performed again, the total number of remaining spirals is reduced to 1 and the total number of turns is reduced to 2, which meets the optimization requirements.

[0112] It should be noted that the number of iterations can be set to an upper limit (such as 3 times) to avoid infinite loop, and if the index requirement is not met after reaching the iteration upper limit, the preset rule or threshold can be adjusted and then processed.

[0113] The embodiment compares the total number of spirals before and after elimination to determine the necessity of iteration, ensures the effectiveness of spiral elimination operation, avoids waste of computing resources caused by invalid processing, and continuously reduces spiral distortion through iteration optimization to improve the regularity of grid topology.

[0114] Based on any of the above embodiments, in the eighth embodiment of the present application, the steps of aligning the non-spiral dual strings of the model grid based on the second normal family structure to obtain the topologically aligned model grid, comprising:

[0115] Step S51, determine the first dual string in the second normal family structure, at least one edge is not a characteristic edge. Step S52, based on the target edge in the first dual string which is not a characteristic edge, execute the string extraction action. Step S53, determine the second dual string which transversely intersects with the first dual string. Step S54, if the characteristic edge of the second dual string changes before string extraction, modify the characteristic edge of the second dual string to align the non-spiral dual strings of the model grid, and obtain the topologically aligned model grid.

[0116] In the present embodiment, the first dual string is the dual string in the second normal family structure that satisfies the condition of "at least one edge is not a characteristic edge", which is the target object of the string extraction action. The target edge is the non-characteristic edge of the first dual string, which is the execution carrier of the string extraction action. The string extraction action refers to an operation of adjusting the position of the target edge or removing the original dual string to align the topology of adjacent grid cells. The second dual string refers to the dual string that transversely intersects with the first dual string, and its characteristic edge may change due to the string extraction action of the first dual string. The characteristic edge change refers to the change of the original characteristic edge attribute (whether it is a characteristic edge) or position of the second dual string caused by the string extraction action.

[0117] As an optional implementation, all dual strings in the second normal family structure are traversed, and the left side and the right side of each dual string are checked one by one to determine whether there is at least one side that does not belong to the characteristic side. The dual string that meets the condition is marked as the first dual string, and the first dual string list is formed by aggregation. For each first dual string, a non-characteristic side with stronger mobility is preferentially selected as a target side (such as the middle position when both sides are non-characteristic sides, or the side when one side is a non-characteristic side), and the target side is translated to a position aligned with the other side by a topological adjustment tool, while the original first dual string is removed, and the connection relationship of the adjacent grid cells is reconstructed. Based on the topological correlation information of the second normal family structure, all dual strings that have intersection points with the first dual string path are queried, and the dual strings that are transversely intersected (with an intersection angle of 80°-100°) are screened out as second dual strings and their intersection positions with the first dual string are recorded. By comparing the characteristic side information of the second dual string before and after the string extraction, if it is found that the characteristic side attribute or position changes, the characteristic side label of the second dual string is adjusted based on the grid topology regularization principle, and the characteristic side attribute is supplemented or removed to ensure the topological alignment of the second dual string and the surrounding dual strings, and finally a regular model grid with completely aligned non-helical dual strings is obtained.

[0118] For example, in the optimization of an aero-engine blade cooling channel grid, 4 first dual strings are screened out from the second normal family structure, of which 2 are single-sided non-characteristic sides and 2 are double-sided non-characteristic sides. For the dual string with a single non-characteristic side, the non-characteristic side is translated to the characteristic side position; for the dual string with double non-characteristic sides, a new side is constructed at the middle position and the original dual string is removed. Then 6 second dual strings that are transversely intersected with the 4 first dual strings are queried, and it is found by comparison that the characteristic sides of 3 of them change in attribute due to string extraction. By adjusting the characteristic side labels of these second dual strings, the precise alignment of all non-helical dual strings is finally achieved, and the grid cells are arranged regularly.

[0119] It should be noted that during the string extraction process, it is necessary to ensure that the translation of the target side does not destroy the quadrilateral structure of the grid cell, and the modification of the characteristic side needs to follow the principle of overall topological consistency. In this embodiment, the string extraction action is performed on the dual strings that meet the conditions, and the characteristic sides of the associated dual strings are simultaneously modified, which effectively aligns the non-helical dual strings and improves the topological regularity of the model grid, providing high-quality grid support for subsequent engineering applications.

[0120] Based on any of the above embodiments, in the ninth embodiment of the present application, the steps of aligning the non-helical dual strings of the model grid based on the second normal family structure after the string supplementing processing include:

[0121] Step S55, determine the third dual chord in the second normal family structure, both sides of which are characteristic edges. Step S56, based on the position information of the third dual chord in the model grid, perform the chord supplement action to align the non-helical dual chord of the model grid, and obtain the topologically aligned model grid.

[0122] In this embodiment, the third dual chord is a dual chord in the second normal family structure, both sides of which belong to the characteristic edge. Since the characteristic edge cannot be moved, it cannot be aligned by the chord extraction action and needs to be handled by the chord supplement method. The position information refers to the coordinate range, adjacent cell distribution and belonging normal family of the third dual chord in the model grid, which provides the basis for the determination of the chord supplement position. The chord supplement action refers to adding a new dual chord at a reasonable position of the third dual chord, adjusting the local grid topology, and realizing the alignment of the non-helical dual chord.

[0123] As an optional implementation, all dual chords in the second normal family structure are traversed, and the attributes of the left side edge and the right side edge of each dual chord are checked one by one. The dual chord whose two edges are both marked as characteristic edges is determined as the third dual chord, a list of third dual chords is formed, and the position coordinates of each chord are recorded. The position information of the third dual chord is obtained, the distribution density and topology direction of the surrounding grid cells are analyzed, and one or more dual chords are added in the gap between the third dual chord and the adjacent dual chord in the direction parallel to the third dual chord. The original normal family with a width of 1 is split, the connection relationship of the adjacent cells is re-established, the non-helical dual chord is regularly arranged in the same direction, and finally the topologically aligned model grid is obtained.

[0124] Illustratively, in the optimization of the cooling channel grid of an aero-engine blade, two third dual chords are identified from the second normal family structure, both of which are located in the turning area of the cooling channel and both of which have characteristic edges on both sides. According to the position coordinates and the surrounding cell distribution of the two third dual chords, a parallel dual chord is added on the inside of each chord, the original normal family is split into two new normal families with a width of 1, the topology connection of the adjacent cells is adjusted, the non-helical dual chord in the turning area is regularly aligned along the channel direction, and the length-width ratio of the grid cells is controlled within a reasonable range.

[0125] It should be noted that the number and position of the chord supplement need to be adjusted according to the actual topology of the grid to ensure that the newly added dual chord does not overlap with the original chord and does not introduce new singular points. In this embodiment, the chord supplement action is performed on the third dual chord with characteristic edges on both sides, which solves the topology alignment problem that cannot be handled by the chord extraction action, realizes the overall regularity of the non-helical dual chord, and further improves the topology quality and consistency of the model grid.

[0126] Based on this, the execution logic of the model processing method based on singular point topology alignment is described in combination with the overall process. For a surface mesh, a half-edge structure is used for reading operations. That is, each edge of a mesh element corresponds to two half-edges, which represent different directions, and each half-edge corresponds to a different surface element. In order to optimize the aspect ratio of a structured quadrilateral mesh, the structure of the mesh is first analyzed and calculated. For a structured mesh, the singular point slicing method is used. All singular points in the mesh are selected in turn, and then all directions extending from the singular points are selected in turn. The corresponding mesh lines are selected and extended until they hit other singular points or boundary points. In this way, the entire mesh is divided into several different slices by the selected mesh lines, and there are no singular points inside each slice. And each slice is a generalized rectangle. Next, the dual chord of the entire mesh is established. The so-called dual chord is that each element in the quadrilateral mesh is regarded as a node in an undirected graph, and then the connection relationship between adjacent elements is regarded as an edge in the undirected graph. In this way, an undirected graph is obtained, and since all elements are quadrilaterals, for non-boundary quadrilateral elements, four elements are connected, so the degree of the corresponding node in the undirected graph is 4. The dual chord refers to a path in this graph, and the next step direction at each fixed point is the straight direction of the previous step. For a certain dual chord, whether it is a loop is recorded, and the edges on the dual chord are recorded as the characteristic edges of the dual chord, and whether it is self-intersecting and the position of the self-intersection are recorded. At the same time, the normal family structure of the entire mesh is established. If two dual chords are adjacent (or parallel), and the adjacent edges are not the edges emitted by the singular points, then the two edges can be merged into the same normal family. For the normal family, record its width (consisting of several dual chords), whether it is a ring, whether it is self-intersecting and the position of the self-intersection. The next step is to search for the spiral dual chord in the original mesh. For a dual chord, for a positive integer k, shift all elements of the dual chord to the left and right by k elements, respectively, to obtain two element sequences. If one of the two sequences has a coincident part with the original dual chord, and the length of the coincident part is long enough, then a dual chord that spans a spiral with a width of k is found. Preferably, the dual chord with k = 1, 2 is searched.

[0127] The width of each helix, i.e. the length of each helix for one turn, and the number of turns of the helix are calculated, only considering helices with at least two turns. Then all the helices are sorted according to the number of turns from more to less. Next, the search and elimination helix path algorithm is performed in turn according to the order of the number of turns of the helices. If a path is searched, the path is eliminated. After elimination, the number of helices, the width, and the number of turns in the graph are recalculated. If the total number of helices and the total number of turns of the helices do not decrease, the entire elimination helix algorithm ends. Otherwise, the search helix is continued for all the new helices in turn. If no path is searched, the path is searched for the next helix. If a helix is searched for an elimination path, the graph is recalculated in the manner of the previous paragraph. If all the helices are not searched for a path, the entire elimination helix algorithm also ends. For the search of the helix path, first, the half-edge structure of the entire grid is already available, so for a quadrilateral face unit, there are corresponding four half-edges. Next, an undirected graph G is established, which is a triple graph of all half-edges, that is, for each half-edge h, there are three nodes hl, hm, hr in G, which represent left turn, right turn, and straight walk, respectively. Therefore, if the number of quadrilateral face units in the original grid is n, the number of half-edges is 4n, and the number of nodes of G is 12n.

[0128] Next, the connection relationship in G is established. Before the elimination path is performed, the elimination helix path is defined. The path is a loop in the graph G. Of course, not any loop is legal, and the entire loop needs to satisfy the following conditions: 1. For each face unit f, according to counterclockwise, the four half-edges are h1, h2, h3, h4, which correspond to 12 nodes in the graph G. The 12 nodes in the loop can only appear at most once at the same time. Except for two cases, two nodes can appear once: 1.1, (h1m, h2m), (h1m, h4m), etc. That is, the m of the two vertical half-edges appears once. 1.2, (h1l, h2m), (h1r, h4m), etc. That is, the first appearing half-edge of the two vertical half-edges is left turn and the second appearing half-edge is right turn. 2. For each helix and a half-edge on the helix, if the helix is right-handed rotation relative to the half-edge, only the r node of the three nodes corresponding to the half-edge is a legal node. If it is left-handed rotation, only the l node is a legal node. For the same edge, corresponding to two half-edges, two legal nodes are l-type and r-type. 3. Taking h1 as an example, if h2 and h4 are both hard edges (characteristic edges), then the three nodes corresponding to h1 are not legal nodes.

[0129] Under such rules, for a helix, for its first and last two circles, the number of edges in each circle is n. Then the first and last two circles have 2n edges, which correspond to 4n legal nodes (2n of each type of r and l), and then the 2n l-type nodes are pairwise corresponding. Selecting one pair, there is a natural path path from the point on the first circle to the point on the second circle. Then, using the breadth-first search algorithm (BFS), search for a path from the end of the path to the start, thereby forming a loop with the previous path. Note that when searching, consider the above restrictions 2 and 3. When a shortest path is searched, calculate whether restriction 1 is satisfied. If satisfied, a reasonable path is found. If not satisfied, start from the starting point of the path, find the node that causes the path to be illegal after being added, and then temporarily remove the node from the graph and start searching again. Until a legal loop is found or a loop cannot be found. Next, re-establish the entire graph of dual chords and regular families. Then align the non-helix type. First find all regular families of width 1, each of which corresponds to some unaligned singular points. Chord extraction or chord supplementation can be used for alignment.

[0130] Chord extraction: for each regular family of width 1 (dual chord). Traverse all left and right edges to determine whether at least one of the left or right is not a hard edge (characteristic edge). If at least one of the left and right edges is not a hard edge for all left and right edge pairs, the dual chord can be extracted. If the left edge can be moved and the right edge cannot be moved, move the left edge to the right. Similarly, if the right edge can be moved and the left edge cannot be moved, move the right edge to the left. During the extraction of the chord, since the data structure of the dual chord has been established before, for each dual chord that intersects the extracted dual chord horizontally, it is necessary to detect whether the information of the chord needs to be modified, for example, if the characteristic edge is eliminated, the characteristic edge of the dual chord needs to be modified. Preferably, additional processing is performed for self-intersection cases.

[0131] Chord supplementation: for dual chords that cannot be extracted or that will cause the grid to be too wide after extraction, chord supplementation is used to determine the location of the dual chord to be supplemented in the model grid, and the pre-set local movement and chord supplementation are detected in turn. After the operation, the grid is smoothed, and then the overall grid optimization is performed.

[0132] For example, referring to Figure 3 and Figure 4 , Figure 3 shows a schematic diagram of a Helic (rotation) type model grid before optimization, Figure 4The model mesh of Helic type is shown in the schematic diagram after optimization. In the generation of structured quadrilateral, some patches are easily generated by repeatedly rotating around a certain direction, resulting in too fine patches and affecting the subsequent optimization quality. Through the method of the embodiment, for example Figure 3 and Figure 4 , the color blocks other than black are the patches forming the spiral. In one spiral, the patches of different turns can be adjacent, or the patches of different turns are separated by other spirals. Therefore, according to the color blocks Figure 3 , there are 50 spirals before processing, and after processing, there are 4 spirals as shown in Figure 4 . It is not described in detail in Figure 3 , in Figure 4 , the blue color in the lower left is the first spiral, the purple color below is the second spiral, and the purple and gray colors in the upper right are separated by two spirals. For example, refer to Figure 5 and Figure 6 , Figure 5 shows the schematic diagram of the model mesh before optimization of the non-rotating type, Figure 6 shows the schematic diagram of the model mesh after optimization of the non-rotating type. For patches of the non-Helic type, alignment is performed based on singular points. For example Figure 5 to Figure 6 , by eliminating the dual string, the number of patches is reduced from 2080 to 606. To facilitate subsequent quality optimization. And Figure 5 and Figure 6 , the purple lines in Figure 6 are the mesh lines derived from the singular points. It can be seen that the number of lines in Figure 6 is reduced compared to Figure 5 .

[0133] For example, refer to Figure 7 , Figure 8 and Figure 9 , Figure 7 shows the schematic diagram of the model before optimization, and the aspect ratio of the patch is 4.0. It should be noted that the aspect ratio is a core index for measuring the quality of mesh elements in CAE (Computer Aided Engineering) finite element analysis, defined as the ratio of the longest side length to the shortest side length (or the shortest characteristic size). High aspect ratio elements will increase the inaccuracy of finite element representation, leading to an enlarged error in the calculation result. In areas where the derivative changes dramatically (such as stress concentration), high aspect ratio will significantly reduce the result accuracy. Excessive aspect ratio will hinder the convergence of nonlinear analysis, and even lead to failure to solve serious element distortion, which may cause calculation errors and interruptions. Poor quality mesh (high aspect ratio) may increase the single iteration calculation time, and the total calculation cost may not decrease but increase, and the result reliability is greatly reduced. For example, refer to Figure 8 , Figure 8A schematic diagram of a model for singular point alignment and spiral elimination using the technical solution described in the above embodiments is shown, but without smoothing processing; the aspect ratio of the optimized model's segments is 3.5. Furthermore, referring to... Figure 9 , Figure 9 The diagram shows the model after singular point alignment and smoothing, with the aspect ratio of its segments further reduced to 3.0.

[0134] For example, refer to Figure 10 , Figure 10 for Figure 7 A local grid diagram. Figure 10 The grid in the medium-colored section has a poor aspect ratio. Figure 10 In the image, 'a' represents the optimized mesh. Figure 10 In Figure 'b', the mesh is shown before optimization. Without altering the overall mesh structure, the aspect ratio of the structured quadrilateral mesh was optimized, reducing the number of poor-quality elements (colored elements) with unfavorable aspect ratios and improving the Jacobian ratio. For example, this application can be applied to the field of automotive design and manufacturing, see reference... Figure 11 and Figure 12 , Figure 11 and Figure 12 This is a mesh diagram of automotive parts, which is visible. Figure 11 The excessive number of singular points in the medium-sized mesh leads to poor mesh quality. The solution proposed in this application improves this, as shown below. Figure 12 The number of singularities is reduced, improving mesh quality. For example, refer to... Figure 13 and Figure 14 , Figure 13 and Figure 14 This is a mesh diagram of another automotive component, which is visible. Figure 13 The excessive number of singular points in the medium-sized mesh leads to poor mesh quality. The solution proposed in this application improves this, as shown below. Figure 14 The number of singularities is reduced, which improves the mesh quality.

[0135] This application provides a model processing device based on singular point topology alignment. The device includes: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, which, when executed, enable the at least one processor to perform the singular point topology alignment-based model processing method described in Embodiment 1. References are made below. Figure 15This document illustrates a schematic diagram of a model processing device suitable for implementing the singular point topology alignment embodiments of this application. The singular point topology alignment-based model processing device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, personal digital assistants (PDAs), tablets, and in-vehicle terminals, as well as fixed terminals such as digital TVs and desktop computers. Figure 15 The model processing device based on singular point topology alignment shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments in this application. Figure 15 As shown, the singularity-based topology alignment model processing device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 1002 or a program loaded from storage device 1003 into random access memory (RAM) 1004. The random access memory 1004 also stores various programs and data required for the operation of the singularity-based topology alignment model processing device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the singularity-based topology alignment model processing device to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows a singularity-based topology alignment model processing device with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0136] In particular, according to the embodiments of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments of the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing program code for executing the method shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network through a communication device, or installed from the storage device 1003, or installed from the ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiments of the present application are executed. The model processing device based on singular point topology alignment provided by the present application adopts the model processing method based on singular point topology alignment in the above-mentioned embodiments, and can solve the technical problem that the grid optimization process exists small fragments, resulting in poor model grid quality. Compared with the prior art, the beneficial effects of the model processing device based on singular point topology alignment provided by the present application are the same as those of the model processing device based on singular point topology alignment provided by the above-mentioned embodiments, and other technical features in the model processing device based on singular point topology alignment are the same as those disclosed in the above-mentioned embodiment method, and will not be repeated here.

[0137] It should be understood that parts of the present application can be realized in hardware, software, firmware, or a combination thereof. In the description of the above-mentioned embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner. The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0138] The application provides a computer readable storage medium having computer readable program instructions (i.e. computer programs) stored thereon for performing the model processing method based on singularity point topology alignment in the above embodiment. The computer readable storage medium provided by the application may, for example, be a U disk, but is not limited to an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, system or device, or any combination of the above. More specific examples of the computer readable storage medium may include, but are not limited to, an electrical connection having one or more conductive wires, a portable computer disk, a hard disk, a random access memory (RAM), a read only memory (ROM), an erasable programmable read only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the embodiment, the computer readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system or device. The program code contained on the computer readable storage medium can be transmitted by any suitable medium, including but not limited to an electrical wire, an optical cable, a radio frequency (RF), or any suitable combination of the above. The above computer readable storage medium may be contained in the model processing device based on singularity point topology alignment; or may exist separately and not be assembled into the model processing device based on singularity point topology alignment. The above computer readable storage medium carries one or more programs, which, when executed by the model processing device based on singularity point topology alignment, cause the model processing device based on singularity point topology alignment to: construct a first normal family structure corresponding to a half edge structure of a model mesh, the first normal family structure containing at least one first normal family, the first normal family containing at least one dual chord; search for a spiral dual chord in the model mesh based on a first normal family with a width greater than a preset width threshold; search for and eliminate a spiral path of the spiral dual chord to eliminate a spiral in the model mesh; construct a second normal family structure of the model mesh after the spiral is eliminated; and process the second normal family structure based on a chord extraction and / or a complementary chord to align non-spiral dual chords of the model mesh to obtain a topologically aligned model mesh. The computer program code for performing the operations of the application may be written in one or more programming languages or combinations thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as the "C" language or similar programming languages.The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider). The flow diagrams and block diagrams in the drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flow diagrams and block diagrams can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical functions ("instructions"). It should also be noted that in some alternative implementations, the functions noted in the block can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks can sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flowchart illustrations, and combinations thereof, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or combinations of hardware and software. The modules described in the embodiments of the present application can be implemented by software, or by hardware. In some cases, the name of the module does not constitute a limitation on the module itself. The computer readable storage medium provided by the present application is a computer readable storage medium, which stores computer readable program instructions (i.e., computer programs) for executing the model processing method based on singular point topology alignment described above, and can solve the technical problem that the grid optimization process has too small fragments, resulting in poor model grid quality. Compared with the prior art, the computer readable storage medium provided by the present application has the same beneficial effects as the model processing method based on singular point topology alignment provided by the above embodiments, and will not be described here.

[0139] The embodiment of the present application provides a computer program product, comprising a computer program, which realizes the steps of the model processing method based on singular point topology alignment when executed by a processor. The computer program product provided by the present application can solve the technical problem that the grid optimization process exists small fragments, resulting in poor model grid quality. Compared with the prior art, the beneficial effects of the computer program product provided by the embodiment of the present application are the same as those of the model processing method based on singular point topology alignment provided by the above-mentioned embodiment, and are not described here. The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent flow transformation using the content of the specification and drawings of the present application, or direct or indirect application in other related technical fields, are also included in the patent processing scope of the present application.

Claims

1. A model processing method based on singular point topological alignment, characterized in that, The model processing method based on singular point topology alignment includes: In response to the triggering operation of the model optimization control, the data file of the model mesh to be optimized is obtained, and the first normal family structure corresponding to the half-side structure of the model mesh is constructed. The first normal family structure contains at least one first normal family, and the first normal family contains at least one dual chord. Based on the first normal family with a width greater than a preset width threshold, the spiral dual chords in the model mesh are searched to obtain the chords. Search for and eliminate the spiral paths of the spiral dual chords to eliminate spirals in the model mesh; Construct a second normal family structure for the model mesh after eliminating spirals; Based on the second normal family structure, the non-spiral dual chords of the model mesh are aligned to obtain the topologically aligned model mesh for rendering and display. The step of constructing the first normal family structure corresponding to the half-side structure of the model mesh includes: A first undirected graph is constructed based on the model mesh, where the cells of the model mesh serve as nodes of the first undirected graph, and the cell connections serve as edges of the first undirected graph. Multiple dual chords are obtained by searching the first undirected graph; Parallel dual strings are combined to form the first normal family, so as to form at least one first normal family. Construct the first normal family structure based on at least one first normal family; For the first normal family whose width is greater than a preset width threshold, determine the target dual chord currently being traversed; The units of the target dual chord are shifted to the left and to the right by a predetermined number of units respectively to obtain two unit sequences; If the number of overlapping units between at least one of the unit sequences and the target dual chord meets a preset threshold, the target dual chord is determined to be a spiral dual chord. The step of searching for and eliminating the spiral paths of the spiral dual chords to eliminate spirals in the model mesh includes: Search for the spiral path of the spiral dual chord; Eliminate the spiral path to remove spirals from the model mesh; The step of searching for the helical path of the helical dual chord includes: A second undirected graph is constructed based on the half-side structure of the model mesh. The second undirected graph is a triple graph containing each half-side of the unit. The nodes of the second undirected graph are the travel directions of each half-side, including left turn, right turn, and straight ahead. Based on the first and last loops of the spiral dual chord, the second undirected graph, and preset rules, the starting and ending nodes on the spiral dual chord are filtered to obtain the first path. The preset rules include the node uniqueness rule corresponding to the surface unit, the legal node rule corresponding to the spiral direction, and the illegal node exclusion rule corresponding to the feature edge. The second path is obtained by searching from the end node to the start node using the breadth-first search algorithm, combined with the preset rules. The spiral path is constructed based on the first path and the second path.

2. The model processing method based on singular point topological alignment as described in claim 1, characterized in that, Before the step of constructing the first normal family structure corresponding to the half-side structure of the model mesh, the following steps are included: Determine the half-side structure of the model mesh; Extend the grid lines of the singular points in the model mesh toward the grid boundary so that the grid lines pass through all the singular points and the endpoints of the grid lines are located at the grid boundary. The model mesh is divided into multiple singularity pieces based on the extended mesh lines, and each singularity piece is used as the updated model mesh.

3. The model processing method based on singular point topological alignment as described in claim 1, characterized in that, After the step of eliminating the spiral path, the following is included: Determine the total number of helices and the total number of turns of the helical dual chord in the model mesh; If the total number of spirals and the total number of spiral turns are both greater than or equal to the total number of spirals and the total number of spiral turns before eliminating the spiral path, then perform the step of searching for and eliminating the spiral path of the spiral dual chord to eliminate the spirals in the model mesh.

4. The model processing method based on singular point topological alignment as described in claim 1, characterized in that, The steps for obtaining a topologically aligned model mesh by aligning the non-spiral dual chords of the model mesh based on the second normal family structure through string-drawing processing include: In the second normal family structure, determine the first dual chord in which at least one edge is not a characteristic edge; Based on the target edge that is not a characteristic edge in the first dual chord, perform the string-pulling action; Determine the second dual chord that intersects laterally with the first dual chord; If the characteristic edges of the second dual chord have changed from those before the string was drawn, modify the characteristic edges of the second dual chord to align with the non-spiral dual chords of the model mesh, thus obtaining the topologically aligned model mesh.

5. The model processing method based on singular point topological alignment as described in claim 1, characterized in that, The steps for obtaining a topologically aligned model mesh by aligning the non-spiral dual chords of the model mesh based on the second normal family structure using chord completion processing include: In the second normal family structure, determine the third dual chord where both sides are characteristic edges; Based on the position information of the third dual chord in the model mesh, a chord-complementing action is performed to align the non-spiral dual chords of the model mesh, resulting in a topologically aligned model mesh.

6. A model processing device based on singular point topology alignment, characterized in that, The singularity-based topology alignment model processing device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the singularity-based topology alignment model processing method as described in any one of claims 1 to 5.

7. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the model processing method based on singular point topology alignment as described in any one of claims 1 to 5.

Images