A nested intersection feature recognition method for complex structural parts

Abstract

A nested intersection feature recognition method for complex structure is proposed. Firstly, the structure model is preprocessed, and the extended attribute adjacency graph of the structure is constructed and stored by using adjacency array. Then, the extended attribute adjacency graph is decomposed and transition feature is suppressed, and the convex feature subgraph set and the concave feature subgraph set are obtained. Then, the intersection of the convex feature subgraph set and the concave feature subgraph set is judged, and the independent feature, convex nested feature and concave nested feature subgraph set are obtained. On this basis, the subgraphs in the convex nested intersection feature and concave nested intersection feature subgraph set are decomposed and reconstructed, and the new independent feature subgraph set is obtained. Finally, the graph isomorphism matching and rules are used to recognize the features in the independent feature subgraph set, and the feature recognition results are presented by visualization and saved in XML file. The method can accurately recognize the nested intersection features in complex structure, and has the advantage of low algorithm complexity.

Description

Technical Field

[0001] This invention belongs to the field of computer-aided manufacturing and relates to a method for recognizing nested intersecting features of complex structural parts. Background Technology

[0002] Currently, the industrial sector contains a large number of complex structural parts. Process engineers can program the machining of these parts using general-purpose computer-aided manufacturing (CAM) software such as UG, Catia, and Hypermill. However, due to the significant differences in part shape and structure, a large amount of manual operation is still required during the programming process. This is not only inefficient but also highly dependent on experience, making large-scale implementation difficult and hindering the improvement of production efficiency. Feature recognition can automatically associate part design information with manufacturing information, which is the key to solving the problem of programming automation and has attracted widespread attention from researchers in industry and academia.

[0003] Intersecting features are composed of at least two single features, and their recognition process involves hierarchical decomposition and reconstruction of the intersecting structure, making them far more difficult to recognize than single features. Nested features are a special case of intersecting features, formed by one or more features completely intersecting and overlapping with another feature in space. This complex, nested structure causes greater damage to the original features and has more complex relationships; algorithms suitable for recognizing general intersecting features often cannot effectively solve the problem of recognizing nested features. The literature "Based on..." Figure Two The algorithm for identifying manufacturing features by secondary decomposition [J]. Mechanical Design and Manufacturing, 2013(5):56-59” adopts the method of constructing the attribute adjacency graph of the part and obtains the minimum subgraph by secondary decomposition of the attribute adjacency graph, which realizes the effective identification of independent features. However, this method does not consider the decomposition of intersecting subgraphs and cannot realize the identification of intersecting features. The literature “Manufacturing Feature Identification Method in 3D Process Design [J]. Mechanical Design and Manufacturing, 2021(6):228-231,235” adopts the feature identification method based on extended attribute adjacency graph and graph isomorphism, which can effectively identify independent features. The adjacency graph representation of parts has been extended to effectively identify independent and partially intersecting features. However, this method still has limitations in identifying complex intersecting features, especially convex nested intersecting features. Invention patent CN101763069B discloses a method for identifying machining features of complex aircraft structural parts, achieving the identification of features such as generalized slots and openings in complex aircraft structural parts, but it still cannot process nested intersecting features. In summary, there is currently no clear method for identifying nested features in complex components, and accurate identification of nested features remains to be achieved. Summary of the Invention

[0004] In order to solve the problem of nested intersection feature recognition of complex structure parts, the application provides a nested intersection feature recognition method for complex structure parts.

[0005] The technical scheme of the application is as follows:

[0006] The nested intersection feature recognition method for complex structure parts comprises the following steps:

[0007] Step 1: import the structure part model for preprocessing, build the extended attribute adjacency graph (EAAG) of the part and store it in the extended attribute adjacency array; specifically:

[0008] Obtain the geometric data and topological data of the structure part model, represent one face in the structure part model by one node, represent the edge between two faces in the structure part model by the edge between the corresponding two nodes, and initialize the extended attribute adjacency graph of the structure part model;

[0009] Determine the concave-convex property and edge angle θ quadrant of the extended attribute adjacency graph edge, and classify the nodes into transition face nodes and non-transition face nodes according to the rules; the transition face is further divided into edge transition face and point transition face, wherein the edge transition face includes edge round corner transition face and edge chamfer transition face; the point transition face includes point round corner transition face and point chamfer transition face, so as to fill the face and edge attributes of the extended attribute adjacency graph;

[0010] Store the extended attribute adjacency graph in the computer by using the extended attribute adjacency array;

[0011] Step 2: obtain the convex feature subgraph set and concave feature subgraph set by decomposing and suppressing the transition features of the extended attribute adjacency graph; specifically:

[0012] Remove the convex inner ring edge of the extended attribute adjacency graph, decompose the extended attribute adjacency graph into a plurality of new subgraphs and store them in the primary decomposition subgraph set;

[0013] Traverse the subgraphs in the primary decomposition subgraph set, check whether the subgraph has a concave inner ring edge, if yes, continue to decompose the subgraph, store the subgraph containing the convex face obtained after the decomposition into the convex feature subgraph set, and store the remaining subgraph back into the primary decomposition subgraph set;

[0014] Traverse the subgraphs in the primary decomposition subgraph set and the convex feature subgraph set, and remove the transition face nodes by classification; specifically, delete the face nodes and related edges for the point transition face; delete the face nodes and related edges for the edge transition face, and then create new edges between the support faces, wherein the support face refers to the non-transition face adjacent to the transition face;

[0015] Check the subgraphs in the initial decomposition subgraph set and the convex feature subgraph set. If there is a convex face, delete the corresponding nodes and edges, and store the new subgraph in the concave feature subgraph set. If there is no convex face, directly store it in the concave feature subgraph set.

[0016] Step three: Perform intersection feature discrimination on the convex feature subgraph set and the concave feature subgraph set to obtain an independent feature subgraph set, a convex nested intersection feature subgraph set, and a concave nested intersection feature subgraph set. Specifically:

[0017] Traverse the subgraphs in the convex feature subgraph set, and determine whether there is an N-1 convex edge face node, where N is the number of subgraph face nodes. If there is, store it in the independent feature subgraph set. Otherwise, store it in the convex nested intersection feature subgraph set.

[0018] Traverse the subgraphs in the concave feature subgraph set, and determine whether there is a convex edge in it. If there is, store the subgraph in the concave nested intersection feature subgraph set. Otherwise, store it in the independent feature subgraph set.

[0019] Step four: Decompose the subgraphs in the convex nested intersection feature subgraph set and the concave nested intersection feature subgraph set, and reconstruct the subgraphs obtained after decomposition and integrate them into the independent feature subgraph set. Specifically:

[0020] Extract the convex platform from the subgraphs in the convex nested intersection feature subgraph set to obtain new subgraphs and store them in the independent feature subgraph set. Then, delete these face nodes and their connected edges in the original subgraph to decompose the original subgraph. Determine whether the subgraph obtained after decomposition has a convex edge. If it does, store it in the concave intersection feature subgraph set. Otherwise, store it in the independent feature subgraph set.

[0021] Traverse the subgraphs in the concave nested intersection feature subgraph set, find a seed face as the starting face for decomposition, and extract the seed face and the face nodes connected to it with concave edges as expansion faces. Preserve the connection relationship and store it in the concave intersection decomposition subgraph set.

[0022] Delete the seed face node and the edges connected to it in the subgraph, and determine whether all the concave edges connected to the expansion face point to other expansion faces or the seed face. If so, delete the expansion face node and all edges connected to it from the subgraph. Otherwise, retain the expansion face node and delete the edges connecting it to other expansion faces. Decompose the subgraph.

[0023] After decomposition, determine whether the remaining subgraph meets the termination condition. If it does, stop the decomposition and store it in the concave intersection decomposition subgraph set. Otherwise, take the remaining subgraph as input again, find a new starting face, and perform iterative decomposition.

[0024] Traverse the subgraphs in the concave intersection decomposition subgraph set, check whether there are face nodes with the same type, normal vector and face equation in any two subgraphs, if there are, merge the two subgraphs into, and store the new subgraph and the remaining subgraphs that do not meet the merging condition in the independent feature subgraph set;

[0025] Step five: feature recognition is performed on the independent feature subgraph set by using graph isomorphism matching and rules, specifically:

[0026] The independent feature subgraph set and the feature subgraph in the pre-defined feature library are matched by using a graph isomorphism algorithm to identify fixed features such as through holes, blind holes, through grooves, through steps, blind steps and other pre-defined topological structures;

[0027] The independent feature subgraph set that is not recognized by the graph isomorphism matching is matched with the established feature rules to identify features with non-fixed topological structures but with regular patterns such as closed grooves, open grooves and bosses;

[0028] Step six: the feature recognition results of the structural part are visualized and saved in an XML file; specifically:

[0029] A user interface (UI) is created, including a tree list, which ensures that selecting a specific list item will highlight the corresponding face in the three-dimensional model, ensuring that the recognition results are immediately reflected on the UI. Through user interaction mechanisms such as mouse clicks, users can select and highlight features. A right-click menu is added to the UI to allow users to delete selected features.

[0030] The feature recognition results of the structural part are sorted and saved as an XML file, with the root element named "FeatureRecognitionResults"; each feature category is presented as a sub-element "FeatureCategory", which contains detailed information about each specific feature, such as a unique identifier, type, location, size and shape.

[0031] The beneficial effects of the present application lie in its fine recognition and processing steps, which successfully and effectively decompose the nested intersection features in the structural part. This innovation not only significantly improves the accuracy of feature recognition, but also greatly improves the efficiency of the entire recognition process, making the recognition of nested intersection features in complex structural parts more reliable and efficient. Through the application of the present application, designers can more accurately obtain feature information of the structural part, thereby providing strong technical support for subsequent product design, manufacturing and testing. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 is a flowchart of the nested intersection feature recognition method for complex structural parts in the method of the present application.

[0033] Figure 2 is a schematic diagram of the part topology in the method of the present invention.

[0034] Figure 3 is a schematic diagram of the edge angle calculation in the method of the present invention

[0035] Figure 4 is a schematic diagram of the concave edge, convex edge, concave cut edge, convex cut edge in the part in the method of the present invention.

[0036] Figure 5 is a schematic diagram of the edge fillet transition surface, point fillet transition surface, edge chamfer transition surface, point chamfer transition surface in the part in the method of the present invention.

[0037] Figure 6 is a schematic diagram of a three-dimensional model containing nested intersecting features in the method of the present invention, wherein numbers 1 to 36 represent the identification of each face of the model.

[0038] Figure 7 is an extended attribute adjacency graph of a part containing nested intersecting features in the method of the present invention, wherein 1-36 corresponds to Figure 6 face 1 to face 36 in the figure.

[0039] Figure 8 is a schematic diagram of several machining features in the feature library and their feature subgraphs in the method of the present invention, wherein (a) is a schematic diagram of a through hole feature and its feature subgraph, (b) is a schematic diagram of a blind hole feature and its feature subgraph, (c) is a schematic diagram of a blind step feature and its feature subgraph, (d) is a schematic diagram of a through slot feature and its feature subgraph, (e) is a schematic diagram of a through step feature and its feature subgraph.

[0040] Figure 9 is a schematic diagram of a boss feature in the method of the present invention.

[0041] Figure 10 is a schematic diagram of an open slot feature in the method of the present invention.

[0042] Figure 11 is a schematic diagram of a closed slot feature in the method of the present invention.

[0043] Figure 12 is a schematic diagram of a channel feature in the method of the present invention. DETAILED DESCRIPTION

[0044] The specific embodiments of the present invention are further illustrated in conjunction with the accompanying drawings and technical solutions.

[0045] Figure 1The specific process of the complex structure nested intersection feature recognition method is shown, and the implementation steps are as follows: 1) import the structure model for pretreatment, build the extended attribute adjacency graph of the structure model and store it by the extended attribute adjacency array; 2) obtain the convex feature subgraph set and the concave feature subgraph set by decomposing and transition feature restraining the extended attribute adjacency graph; 3) discriminate the intersection features of the convex feature subgraph set and the concave feature subgraph set to obtain the independent feature subgraph set, the convex nested intersection feature subgraph set and the concave nested intersection feature subgraph set; 4) decompose and reconstruct the independent feature subgraph set from the convex nested intersection feature subgraph set and the concave nested intersection feature subgraph set; 5) recognize the features of the independent feature subgraph set by graph isomorphism matching and rules; 6) visualize the structure feature recognition result and save it in an XML file. The above steps and principles will be described in detail as follows:

[0046] Step 1: import the structure model for pretreatment, build the extended attribute adjacency graph EAAG of the structure model and store it by the extended attribute adjacency array;

[0047] (1) Import the structure model, and the encapsulated function provided by the secondary development interface of the software can obtain the topological body of the part model. Then, the corresponding topological body is obtained layer by layer in the order of solid, shell, face, loop, fin, edge and vertex as shown in Figure 2 . Then, the B-rep data of each layer is extracted from the topological body of each layer. Finally, the extended attribute adjacency graph is initialized according to the relationship between the face and the edge. In the initialized extended attribute adjacency graph, each node represents a face in the part model, and the edge between each two nodes corresponds to an edge in the part model. The attribute adjacency graph is filled with part and edge attribute values according to the obtained B-rep data.

[0048] (2) The extended attribute adjacency graph construction also needs to judge the concave-convex property of the edge. The included angle between two adjacent faces is denoted as θ, and the calculation method of θ is as follows:

[0049] As shown in Figure 3 , the normal vectors of the two planes are N1 and N2 respectively, and the included angle θ is calculated by formula (1):

[0050]

[0051] where • denotes the dot product of vectors, and ||N1|| and ||N2|| denote the modulus of the two normal vectors, respectively. According to the range of θ, the concave-convex of the edge is divided into four types: concave edge (θ > π), convex edge (θ < π), concave tangent edge (θ gradually approaches π-), and convex tangent edge (θ gradually approaches π+). For the case of θ = π, the normal plane of the edge at point P is established, and the intersection line of the two planes is obtained. By judging the concave-convex of the intersection line at point P, if there is a concave intersection line at point P, it is judged as a concave tangent edge, otherwise, it is a convex tangent edge. As shown in FIG. 1, e1 is a concave edge, e2 is a convex edge, e3 is a concave tangent edge, and e4 is a convex tangent edge. Figure 4

[0052] (3) Add the quadrant information of the edge included angle θ. In order to make the machining feature more accurately described in shape, the quadrant information of the edge included angle θ is encoded as 0 to 7 in turn, where 0 indicates that the edge included angle is located in the first quadrant, 1 indicates that the edge included angle is π / 2, and so on, until 7 indicates that the edge included angle is located in the fourth quadrant, corresponding to 2π:

[0053] (4) Divide the face node into transition face and non-transition face, and divide the transition face into four types for identification, namely edge fillet transition face, point fillet transition face, edge chamfer transition face, and point chamfer transition face. The characteristics of the fillet transition face are summarized as follows: the face type is a cylindrical surface, a conical surface, a torus, a spherical surface, or a B-spline surface; it is smooth by itself, without sharp points or sharp edges inside; there is at least one smooth edge; any two faces adjacent to the smooth edge are not parallel; the maximum curvature radius is less than a given threshold; and the area ratio to the adjacent plane needs to be less than a set threshold. Based on the identification of the fillet transition face, if the fillet transition face also meets the condition of being connected to two non-transition faces through smooth edges, it is determined to be an edge fillet transition face; otherwise, it is determined to be a point fillet transition face. For the edge fillet transition face, the two non-transition faces need to be recorded as the supporting faces of the edge fillet transition face.

[0054] The characteristics of the edge chamfer transition face are summarized as follows: the face type is a plane or a conical surface, and the number of edges is usually 4, with two parallel edges, and the quotient of the area and the average of the two longest edges is less than 5. In addition, the two longest edges are not parallel to the two adjacent faces, and the area ratio to the adjacent faces is less than a certain threshold, which can be determined by machining experience, and the included angle with the adjacent faces is greater than 90° and less than 270°.

[0055] The characteristics of the point chamfer face are as follows: the type of the point chamfer face is a plane, and the number of edges is usually 3, with all edge lengths less than 3. In addition, the area ratio to the adjacent faces is less than a certain threshold, which is determined by machining experience, and the included angle with the adjacent faces is greater than 90° and less than 270°.

[0056] ​According to the four transition surface matching rules, the face nodes in the extended attribute adjacency graph are traversed and matched. In this process, the matching rules are applied to each face node. If a face node meets a rule, i.e. meets the conditions of a transition surface, it is marked as this type of transition surface. For example, as shown in Figure 5 f1 is an edge fillet transition surface, f2 is a point fillet transition surface, f3 is an edge chamfer transition surface, and f4 is a point chamfer transition surface.

[0057] (5) After obtaining the information of the above faces and edges, the face and edge node attributes of the extended attribute adjacency graph are filled again to construct a complete extended attribute adjacency graph of the part model. Figure 6 For a three-dimensional model of a part, where numbers 1 to 36 represent the identification of each face of the model, the extended attribute adjacency graph is as shown in Figure 7 where 1-36 correspond to faces 1 to 36 in Figure 6 . The extended attribute adjacency graph is stored in a computer using an extended attribute adjacency array. An extended attribute adjacency array A of size n x n (n is the number of face nodes of the extended attribute adjacency graph) is created. The diagonal elements A[i][i] (0≤i≤n) of the extended attribute adjacency array are face node objects of the inherited face class, which have attributes such as face type, concave-convex property of the face, transition surface type of the face, etc. The non-diagonal elements A[i][j] (0≤i≤n, 0≤j≤n, i≠j) of the extended attribute adjacency array are edge node objects of the inherited edge class, which have attributes such as edge type, concave-convex property of the edge, quadrant to which the included angle θ of the edge belongs, etc.

[0058] Step 2: Obtain the convex feature subgraph set and the concave feature subgraph set by decomposing and suppressing the transition features of the extended attribute adjacency graph;

[0059] (1) Remove the convex inner loop edges. Traverse the face nodes in the extended attribute adjacency graph and check their attribute information. If the number of convex inner loops of a face node is greater than 1 (the attribute query function can query whether a loop is an inner loop or an outer loop, and if it is an inner loop and the concave-convex properties of the edges constituting the loop are both concave edges or concave tangent edges, then the inner loop is a concave inner loop, and the judgment method of convex inner loop in the following is similar), delete the convex inner loop edges in the face in the extended attribute adjacency graph, and decompose the extended attribute adjacency graph to obtain a set of first decomposed subgraphs (FDSS).

[0060] (2) Remove the concave inner loop edges. Traverse the FDSS once, and determine whether there is a face node with more than 0 concave inner loop edges in the subgraph. If there is, remove the concave inner loop edges in the face in the subgraph, realize the decomposition of the subgraph, and if the new subgraph obtained after decomposition has a face node with the concave inner loop attribute true, store it as the set of convex feature subgraphs (CVFSS). Store the new subgraph remaining after decomposition in the FDSS.

[0061] (3) Transition feature suppression. First traverse the subgraphs in the FDSS and the set of concave feature subgraphs (CVFSS), and if there is a transition face node in the subgraph, further process it; the transition face is a point transition face, and the face node and all edges connected thereto are deleted. If it is an edge transition face, further determine whether the two support faces of the face are both in the subgraph. If so, delete the face node and all edges connected thereto, and create a new edge between the two support faces; if not, only delete the transition face and all edges connected thereto.

[0062] (4) Delete convex faces. Traverse the subgraphs in the FDSS, and determine whether there is a convex face (the definition of a convex face is that the edges surrounding a face are all convex edges or convex cut edges, and the face is called a convex face) in the subgraph. If so, delete the convex face node and the edges connected thereto in the subgraph, and store the new subgraph obtained in the set of concave feature subgraphs (CCFSS). If there is no convex face, directly store the subgraph in the CCFSS.

[0063] Step 3: Perform intersection feature discrimination on the convex feature subgraph set CVFSS and the concave feature subgraph set CCFSS to obtain an independent feature subgraph set and a nested intersecting feature subgraph set;

[0064] (1) Convex intersecting feature subgraph discrimination. Traverse the subgraphs in the convex feature subgraph set CVFSS, and determine whether there is a face node with N-1 edges and all convex edges in the subgraph, where N is the number of face nodes in the subgraph. If there is such a face node, determine that the subgraph is an independent convex feature, and store it in the set of independent feature subgraphs (IFSS). Otherwise, store the subgraph in the set of convex intersecting feature subgraphs (CVIFSS).

[0065] (2) Concave Intersecting Feature Subgraph Discrimination. Traverse the subgraphs in the set of concave feature subgraphs CCFSS, and determine whether there is a convex edge in the subgraph. If not, store it in IFSS. Otherwise, store it as the set of concave intersecting feature subgraphs (ConcaveIntersectingFeatureSubgraph) CCIFSS.

[0066] Step 4: Decompose and reconstruct the subgraphs of the set of convex nested intersecting feature subgraphs and the set of concave nested intersecting feature subgraphs to obtain the set of independent feature subgraphs;

[0067] (1) Decompose the subgraphs in the set of convex nested intersecting feature subgraphs. Traverse the subgraphs in CVIFSS, extract the convex faces and concave inner ring face attributes with true, and the concave faces connected by convex edges or having concave edges connected only to the concave inner ring face attributes with true. Preserve the edge connection relationship between these face nodes, extract the new subgraph, and store it in IFSS. Then, delete these face nodes and the edges connected to these face nodes, decompose the original subgraph, and determine whether the decomposed subgraph has a convex edge. If so, store the obtained subgraph in the set of concave intersecting feature subgraphs CCIFSS, and if not, store the obtained subgraph in the set of independent feature subgraphs IFSS.

[0068] (2) Decompose the subgraphs in the set of concave nested intersecting feature subgraphs. Traverse the subgraphs in the set of concave nested intersecting feature subgraphs CCIFSS, find all face nodes with only concave edges as the starting face for concave intersecting subgraph decomposition. If there is no such face node, select the face node with the most concave edges; if the face node with the most concave edges is not unique, select the one with the largest area. Treat the face nodes connected to the starting face with concave edges as expansion faces, extract the seed face and all expansion faces, preserve the connection relationship between these face nodes, and store it as the set of concave intersecting decomposition subgraphs (ConcaveIntersectingDecompositionSubgraph) CCIDSS; delete the seed face node and all edges connected to it from the subgraph. Determine whether the directions of the concave edges connected to the expansion faces are all other expansion faces, if so, delete the expansion face node and all edges connected to it from the subgraph; otherwise, preserve the expansion face node and delete the edges connected to other expansion faces. Decompose the subgraph. Determine whether the remaining subgraph after decomposition meets one of the following termination conditions: (1) there is a face node in the subgraph with n-1 concave edges, where n is the number of face nodes in the subgraph; (2) all face nodes in the subgraph have and only have two edges. If one of the conditions is met, stop the decomposition and store the remaining subgraph after decomposition in CCIDSS. If none of the termination conditions is met, take the remaining subgraph as input again and iterate the decomposition with a new starting face.

[0069] (3) Subgraph reconstruction of CCIDSS subgraphs: For subgraphs in CCIDSS, there may be cases where what was originally a single feature is divided into several sub-features due to feature intersection. Therefore, it is necessary to cluster sub-features belonging to the same processing feature in CCIDSS into composite processing features. The specific implementation method is as follows: traverse the subgraphs in CCIDSS and check whether there are face nodes with the same type, normal vector, and face equation in any two subgraphs. If such a case exists, merge the two subgraphs into a new subgraph, and store the new subgraph and the remaining subgraphs in CCIDSS that do not meet the merging conditions as an Independent Feature Subgraph Set (IFSS).

[0070] Step 5: Use graph isomorphism matching and rules to perform feature recognition on the set of independent feature subgraphs;

[0071] (1) Combine the set of independent feature subgraphs with, for example Figure 8 The feature subgraphs in the predefined feature library are matched using a graph isomorphism algorithm to identify fixed features of predefined topological structures such as through holes, blind holes, through slots, through steps, and blind steps. Successfully identified independent feature subgraphs of the same category are stored as a new feature set. Then, an identified feature set is created, storing feature sets of different categories for further reference and management. Finally, successfully identified feature subgraphs are removed from the independent feature subgraph set.

[0072] (2) Match the set of independent feature subgraphs with the established feature rules. This is to identify features with non-fixed but regular topological structures, such as closed slots, open slots, and bosses. The identified independent feature subgraphs of the same type are stored as a new feature set. Then, they are added to the already identified feature set. The following are three typical processing feature identification rules:

[0073] ① If an Independent Feature Subgraph (IFS) contains a base surface node with M = N-1 edges, all of which are convex, and N is the number of surface nodes in the IFS; and all other surface nodes have 3 edges, then the IFS can be considered a boss feature. For example... Figure 9 As shown, f1 is a base face node with N-1 convex edges, and the other face nodes all have 3 edges, thus exhibiting a boss feature.

[0074] ② If an Independent Feature Subgraph (IFS) contains a base surface node with M = N-2 edges, all of which are concave; besides this base surface node, there are N-3 other base surface nodes with 3 edges each; and furthermore, there are 2 other base surface nodes with 2 edges each, then the IFS can be determined to be an open slot feature. For example... Figure 10As shown, f1 is a base node with N-2 convex edges, and there are N-3 nodes with 3 edges and 2 nodes with 2 edges, thus it is an open slot feature.

[0075] ③If there is a base node in the independent feature subgraph IFS, the number of edges M = N-1, and all are concave edges. In addition to this node, the number of edges of the remaining nodes M is 3. Then it can be judged that the IFS is a pocket feature. For example, Figure 11 As shown, f1 is a base node with N-1 concave edges, and the number of edges of the remaining nodes is 3, thus it is a pocket feature.

[0076] ④If the independent feature subgraph IFS has 3 or more nodes, and each node has two edges, then it can be judged that the IFS is a passage feature. For example, Figure 12 As shown, f1 is a base node with N-1 nodes, and each node has two edges, thus it is a passage feature.

[0077] Step 6, visualize the structure feature recognition results and save them in an XML file;

[0078] Create a user interface (UI) including a tree list, the first level list displays different categories of features, and the second level list displays each specific feature. Through the association between the interface and the three-dimensional model, it is ensured that selecting a specific list item will highlight the corresponding surface in the three-dimensional model. Real-time updating is realized to ensure that the recognition results are immediately reflected on the UI. Through user interaction mechanisms such as mouse clicks, users can select and highlight features. Finally, add a right-click menu to the UI to allow users to delete selected features.

[0079] The structure feature recognition results are saved in XML format to meet the use requirements of subsequent computer-aided process planning systems. The specific steps include: first, identify the machining features of the structure to generate structured data; second, organize the recognition results into an XML file, with the root element named "FeatureRecognitionResults"; in the XML file, each feature category is presented as a sub-element "FeatureCategory", which contains detailed information of each specific feature in the structure, such as unique identifier, type, location, size, shape, etc.; finally, save the generated XML file to ensure that its format meets the requirements of subsequent CAPP systems.

Claims

1. A method for identifying nested intersecting features of complex structural components, characterized in that, Includes the following steps: Step 1: Import the structural component model for preprocessing, construct the extended attribute adjacency graph EAAG of the part, and store it in the extended attribute adjacency array; Specifically: Obtain the geometric and topological data of the structural component model. One face in the structural component model is represented by one node, and the edge between two faces in the structural component model is represented by the edge between the corresponding two nodes. Initialize the extended attribute adjacency graph of the structural component model. Determine the concavity / convexity of edges in the extended attribute adjacency graph and the quadrant of the edge angle θ, and classify nodes into transition surface nodes and non-transition surface nodes according to rules; transition surfaces are further divided into edge transition surfaces and point transition surfaces, where edge transition surfaces include edge rounded transition surfaces and edge chamfered transition surfaces; point transition surfaces include point rounded transition surfaces and point chamfered transition surfaces, in order to fill the face and edge attributes of the extended attribute adjacency graph. An extended attribute adjacency array is used to store the extended attribute adjacency graph in the computer; Step 2: By decomposing the extended attribute adjacency graph and suppressing transition features, we obtain a set of convex feature subgraphs and a set of concave feature subgraphs; specifically: Remove the convex inner loop edges of the extended attribute adjacency graph, decompose the extended attribute adjacency graph into several new subgraphs and store them in the initial decomposed subgraph set; Traverse the subgraphs in the initial decomposed subgraph set, check if there are concave inner loop edges in the subgraphs. If there are, continue to decompose the subgraphs. Store the subgraphs with convex surfaces obtained after decomposition as convex feature subgraphs into the convex feature subgraph set. Store the remaining subgraphs back into the initial decomposed subgraph set. Traverse the subgraphs in the initial decomposition subgraph set and the convex feature subgraph set, and classify and remove the transition surface nodes. Specifically, delete the surface nodes and related edges for point transition surfaces; delete the surface nodes and related edges for edge transition surfaces, and then create new edges between supporting surfaces, where supporting surfaces refer to those non-transition surfaces that are adjacent to the transition surface. Check the subgraphs in the initial decomposition subgraph set and the convex feature subgraph set. If there are convex surfaces, delete the corresponding nodes and edges, and store the new subgraph in the concave feature subgraph set. If there is no convex surface, it is directly stored in the set of concave feature subgraphs; Step 3: Perform intersection feature discrimination on the convex feature subgraph set and the concave feature subgraph set to obtain the independent feature subgraph set, the convex nested intersecting feature subgraph set, and the concave nested intersecting feature subgraph set; Specifically: Traverse the subgraphs in the convex feature subgraph set, and determine whether there are N-1 convex edge face nodes, where N is the number of face nodes in the subgraph. If they exist, store them in the independent feature subgraph set; otherwise, store them in the convex nested intersecting feature subgraph set. Traverse the subgraphs in the concave feature subgraph set, and determine whether there are convex edges. If there are, store the subgraph in the concave nested intersecting feature subgraph set; otherwise, store it in the independent feature subgraph set. Step 4: Decompose the subgraphs in the convex nested intersecting feature subgraph set and the concave nested intersecting feature subgraph set, reconstruct the subgraphs obtained after decomposition, and merge them into the independent feature subgraph set, specifically as follows: Extract the convex facets from the subgraphs in the set of nested intersecting feature subgraphs to obtain new subgraphs, and store them in the set of independent feature subgraphs; Subsequently, these face nodes and their connected edges in the atomic graph are deleted to decompose the atomic graph. It is then determined whether the subgraphs obtained after decomposition have convex edges. If they do, they are stored in the set of concave intersecting feature subgraphs; otherwise, they are stored in the set of independent feature subgraphs. Traverse the subgraphs in the set of concave nested intersecting feature subgraphs, find the seed face as the starting face for decomposition, extract the seed face and the face nodes connected to it by concave edges as the expansion face, preserve the connection relationship, and store them in the set of concave intersecting decomposition subgraphs. Delete the seed face node and its connected edges in the subgraph. Determine whether all the concave edges connected to the extended face point to other extended faces or seed faces. If so, delete the extended face node and all its connected edges from the subgraph. Otherwise, keep the extended face node, delete the edges connecting it to other extended faces, and decompose the subgraph. After decomposition, determine whether the remaining subgraph satisfies the termination condition; if it does, stop the decomposition and store it in the set of concave intersecting decomposition subgraphs; otherwise, use the remaining subgraph as input again to find a new starting face for iterative decomposition. Traverse the subgraphs in the concave intersection decomposition subgraph set, check if there are face nodes with the same type, normal vector, and face equation in any two subgraphs. If so, merge the two subgraphs into a new subgraph and store the new subgraph and the remaining subgraphs that do not meet the merging conditions into the independent feature subgraph set. Step 5: Perform feature identification on the set of independent feature subgraphs using graph isomorphism matching and rules, specifically as follows: The set of independent feature sub-maps is matched with feature sub-maps in a predefined feature library using a graph isomorphism algorithm to identify fixed features of through holes, blind holes, through slots, through steps, and blind steps. The set of independent feature subgraphs that were not identified by graph isomorphism matching is matched with the established feature rules in order to identify features such as closed slots, open slots, and bosses that are not fixed but have regularity. Step 6: Visualize the feature recognition results of the structural components and save them in an XML file.

2. The nested intersection feature recognition method for complex structural components according to claim 1, characterized in that, Create a user interface, including a tree-like list, ensuring that selected list items will have their corresponding faces highlighted in the 3D model, and ensuring that the recognition results are reflected on the UI in real time; implement user interaction mechanisms to allow users to select and highlight features; and add a right-click menu to the UI to allow users to delete selected features; The structural component feature recognition results are organized and saved as an XML file, with the root element named "FeatureRecognitionResults"; each feature category is presented as a child element "FeatureCategory", which contains detailed information about each specific feature.

Images