Method, system and equipment for automatically identifying typical structure of complex product based on instance library matching and medium

Abstract

The invention relates to the technical field of computer aided design, in particular to a method, a system and equipment for automatically identifying a typical structure of a complex product based on case library matching and a medium. According to the method, a predefined typical structure instance library is constructed, part shape information and space neighbor characteristics among parts are comprehensively considered, part matching is carried out in a product three-dimensional model by taking typical structure instances as targets, structures with similar shapes and structures are searched through depth-first search, and finally, typical structure identification is realized. The problem that the part connection relation in a complex product is difficult to obtain is effectively avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of computer-aided design, and in particular, to an automatic recognition method, system, device and medium for typical structures of complex products based on instance library matching. Background Art

[0002] A process plan is a productive process document compiled by the process department according to design requirements, process technical requirements and quality requirements. The process plan not only includes process information, but also production information and quality information, and is a specific work instruction for guiding workers to perform actual operations on a specified assembly process flow, including operation instructions, processes, assembly time sequences, change records and other information.

[0003] The content of process plan compilation is closely related to the structural characteristics of the product. Specifically, there is a strong reference value among the process plan compilation contents of some typical local structures. If such information can be obtained and reused, the compilation efficiency can be greatly improved. To achieve this goal, it is first necessary to effectively identify typical structures.

[0004] The Chinese invention patent with the name "Assembly body retrieval method based on spatial connection skeleton descriptor" and the authorization announcement number "CN108628965B" proposes an assembly body retrieval method based on spatial connection skeleton descriptor. A spatial connection skeleton is established according to the positions of the centers of parts in the assembly body, the center positions of mating surfaces, and the connection relationships between parts and mating surfaces. Based on the spatial connection skeleton, the spatial distance distribution of randomly sampled points on the surfaces of all two parts passing through the skeleton is statistically calculated as the descriptor input for assembly body retrieval. On this basis, the similarity comparison of the assembly body model is carried out by the method of optimal subsequence bijective matching to realize the comprehensive retrieval of the shape and connection relationship of the assembly body model. However, this method needs to consider the skeleton distances of any two pairs of parts in the assembly body, and for complex product models with too many parts, the description process of this method is relatively cumbersome.

[0005] The literature "Wang P, Zhang J, Li Y, et al. Reuse-oriented common structure discovery in assembly models[J]. Journal of Mechanical Science and Technology, 2017, 31:297-307" proposed a method for discovering typical structures in assemblies by integrating local differences. Based on the analysis of part similarity, similar parts and connection relationships are clustered, and the GSpan algorithm is used to discover the assembly graph descriptors. The obtained frequent subgraphs correspond to the typical structures in the assembly. This method relies on algorithms for self-discovery of typical structures and cannot effectively control the objects and scope of the recognition results. Therefore, there are still certain limitations in actual use.

[0006] Therefore, the existing recognition methods have the problem that it is difficult to effectively obtain typical structures in complex products. Summary of the Invention

[0007] In view of the problem that it is difficult to effectively obtain typical structures in complex products in the existing recognition methods, the present invention proposes an automatic recognition method, system, device and medium for typical structures of complex products based on instance library matching. By constructing a predefined typical structure instance library, considering the part shape information and the spatial proximity characteristics between parts, part matching is performed in the product three-dimensional model with the typical structure instance as the target, and structures with similar shapes and structures are found through depth-first search, and finally the recognition of typical structures is realized, effectively avoiding the problem of difficulty in obtaining part connection relationships in complex products.

[0008] The specific implementation content of the present invention is as follows: An automatic recognition method for typical structures of complex products based on instance library matching. First, a typical structure instance library is constructed according to the obtained STG model. Secondly, the part shape information and the spatial proximity characteristics between parts are obtained according to the constructed typical structure instance library. Then, according to the part shape information and the spatial proximity characteristics between parts, a matching candidate set of the target product three-dimensional model is obtained. Finally, according to the matching candidate set, the spatial distribution aggregation of the matching parts is obtained, and the typical structure is recognized.

[0009] To better implement the present invention, further, the automatic recognition method for typical structures of complex products based on instance library matching specifically includes the following steps: Step S1: According to the obtained STG model, a typical structure instance library is constructed. The STG model is a model established with the local features of the three-dimensional model as vertices and the adjacency relationships between local features as edges. The instances in the typical structure instance library are assemblies. Step S2: According to the constructed typical structure instance library, obtain the part shape information and the spatial proximity characteristics between parts, convert the obtained part shape information into a k-dimensional vector, and calculate the set of proximity vectors according to the obtained spatial proximity characteristics between parts; Step S3: Calculate the similarity between parts according to the k-dimensional vectors of the parts, and obtain the initial set of similar parts according to the similarity; calculate the proximity set similarity between parts in the initial set according to the set of proximity vectors, and obtain the matching candidate set corresponding to each part according to the proximity set similarity; Step S4: Identify the matching parts according to the matching candidate set of each part in the typical structure instance and the spatial distribution aggregation of the matching parts, and obtain the typical structure composed of the matching parts.

[0010] To better implement the present invention, further, the step S2 specifically includes the following steps: Step S21: According to the constructed typical structure instance library, obtain the parts of the typical structure instance and the product model, and call the shape distribution algorithm to describe the shape information of the parts as a k-dimensional vector; Step S22: According to the constructed typical structure instance library, obtain the typical structure instance, calculate the proximity characteristics between parts, and obtain the set of proximity vectors.

[0011] To better implement the present invention, further, the step S22 specifically includes the following steps: Step S221: According to the constructed typical structure instance library, obtain the typical structure instance, and construct the OBB bounding box of the parts; the typical structure instance includes the maximum length, maximum width, and maximum height of the parts; Step S222: Judge the spatial distribution relationship between parts according to the bounding box interference situation, and obtain the set of proximity vectors; Step S223: Repeat steps S221 - S222 until each part in the product model and the typical structure instance library is traversed.

[0012] To better implement the present invention, further, the step S3 specifically includes the following steps: Step S31: Coarsely filter part p according to the set shape similarity threshold and the similarity between part p and part q to obtain the initial set; Step S32: Calculate the proximity set similarity between the parts in the initial set and part p according to the initial set; Step S33: Match the initial set according to the set proximity similarity threshold and the proximity similarity to obtain the matching candidate set.; To better implement the present invention, further, the step S4 specifically includes the following steps: Step S41: Construct an ordered set according to the size of the part neighbor set in the typical structure instance; Step S42: Initialize set list, set Ca, set Flag, and set Tar; Step S43: Assign the ordered set to set list. If part p is the first element of set list, assign the matching candidate set to set Ca; Step S44: Randomly select part q1 from set Ca and determine whether the set criterion conditions are met; Step S45: If not, assign Ca - {q1} to set Ca, and assign Ca p - {q1} to the matching candidate set Ca p of part p; if the set , then select the next part q2 in set Ca and return to Step S44; if the set , then end the search; Ca - {q1} and Ca p - {q1} are the sets obtained by removing element {q1} from Ca and Ca p respectively; Step S46: If the set criterion holds, add part p to set Flag, and at the same time establish an ordered set list p of the neighbor vector set Nei p of part p; Step S47: If , then use the part set corresponding to the output set Tar as the typical structure, and assign list ins — Flag to set list, assign to set Ca, and return to Step S44, list ins is the ordered set of the typical structure instance ins ; Step S48: If , then assign to set Tar, assign to set Ca, assign to set list, and return to Step S44, Nei q1 is the neighbor vector set of part q1; Step S49: Use the output set Tar as the matched typical structure.

[0013] To better implement the present invention, further, the criterion conditions set in Step S44 are: ; Among them, p is the first element of the initialized set list, and Nei p represents the set of neighbor vectors of the part p , and Nei q1 represents the set of neighbor vectors of the part q 1 . u and v are respectively a neighbor vector in Nei p and Nei q1 , and represents the matching candidate set of u.

[0014] To better implement the present invention, further, the establishment process of the STG model in step S1 is as follows: First, establish a vertex adjacency graph according to the topological relationship between the geometric elements of the obtained B-rep data; secondly, search for the maximum cliques in the adjacency graph and use the geometric regions corresponding to the maximum cliques as the local features of the solid model; then call a statistical method to transform the shape information of the local features into feature vectors, form a feature space, and use an unsupervised learning algorithm to suppress the local features differentially; finally, establish an STG model with the local features as vertices and the adjacency relationship between the local features as edges.

[0015] Based on the above-mentioned automatic recognition method for typical structures of complex products based on instance library matching, to better implement the present invention, further, an automatic recognition system for typical structures of complex products based on instance library matching is proposed, which is used to execute the above-mentioned automatic recognition method for typical structures of complex products based on instance library matching; it includes an instance library construction unit, a feature acquisition unit, a matching candidate unit, and an identification unit; The instance library construction unit is used to construct a typical structure instance library according to the obtained STG model; The feature acquisition unit is used to obtain the part shape information and the spatial neighbor characteristics between parts according to the constructed typical structure instance library; The matching candidate unit is used to obtain the matching candidate set of the three-dimensional model of the target product according to the part shape information and the spatial neighbor characteristics between parts; The identification unit is used to obtain the spatial distribution aggregation of the matching parts according to the matching candidate set, and identify the typical structure.

[0016] Based on the above-mentioned automatic recognition method for typical structures of complex products based on instance library matching, to better implement the present invention, further, an electronic device is proposed, including a memory and a processor; a computer program is stored on the memory; when the computer program is executed on the processor, the above-mentioned automatic recognition method for typical structures of complex products based on instance library matching is implemented.

[0017] Based on the above, an automatic recognition method for typical structures of complex products based on instance library matching is proposed. To better implement the present invention, further, a computer-readable storage medium is proposed. Computer instructions are stored on the computer-readable storage medium; when the computer instructions are executed on the above-mentioned electronic device, the automatic recognition method for typical structures of complex products based on instance library matching is implemented.

[0018] The present invention has the following beneficial effects: (1) The present invention replaces the part connection relationship in the existing method with the spatial distribution relationship between parts, which can effectively avoid the problem of difficulty in obtaining the part connection relationship in complex products.

[0019] (2) The present invention is implemented through a pre-constructed instance library, and the recognition object is controlled by adding, deleting, and modifying the instance library, which has good scalability. Description of the Drawings

[0020] Figure 1 It is the overall flowchart of the automatic recognition method for typical structures of complex products provided by the present invention.

[0021] Figure 2 It is an example diagram of the typical structure instance library constructed in the specific implementation manner of the method of the present invention.

[0022] Figure 3 It is an example diagram of the near neighbor set of the parts connecting the earpiece parts and the shape vectors of the parts in the first typical structure instance in the specific implementation manner of the method of the present invention.

[0023] Figure 4 It is an example diagram of the 3D model instance of the card board product in the specific implementation manner of the method of the present invention.

[0024] Figure 5 It is an example diagram of the matching candidate set of the rivet parts in the specific implementation manner of the method of the present invention.

[0025] Figure 6 It is the flowchart of using the matching parts to search for the typical structure in the specific implementation manner of the method of the present invention.

[0026] Figure 7 It is an example diagram of the part set corresponding to the typical structure identified from the card board product in the specific implementation manner of the method of the present invention. Specific Embodiments

[0027] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will combine the accompanying drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. It should be understood that the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments, and should not be regarded as a limitation of the protection scope. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work belong to the protection scope of the present invention.

[0028] In the description of the present invention, it should be noted that unless otherwise clearly defined and limited, the terms "set", "connected", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can also be directly connected, or indirectly connected through an intermediate medium, and can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0029] Embodiment 1: This embodiment proposes an automatic recognition method for typical structures of complex products based on instance library matching. First, according to the obtained STG model, a typical structure instance library is constructed; secondly, according to the constructed typical structure instance library, part shape information and the spatial proximity characteristics between parts are obtained; then, according to the part shape information and the spatial proximity characteristics between parts, a matching candidate set of the three-dimensional model of the target product is obtained; finally, according to the matching candidate set, the spatial distribution aggregation of the matching parts is obtained, and the typical structure is recognized.

[0030] Working principle: In this embodiment, by constructing a predefined typical structure instance library, comprehensively considering part shape information and the spatial proximity characteristics between parts, part matching is performed in the three-dimensional model of the product with the typical structure instance as the target, and a structure with similar shape and structure is found through depth-first search, and finally the recognition of the typical structure is realized; replacing the part connection relationship in the existing method with the spatial distribution relationship between parts can effectively avoid the problem of difficulty in obtaining the part connection relationship in complex products.

[0031] Embodiment 2: This embodiment is described in detail in the form of steps on the basis of Embodiment 1 above.

[0032] The automatic recognition method for typical structures of complex products based on instance library matching specifically includes the following steps: Step S1: According to the obtained STG model, construct a typical structure instance library.

[0033] In the STG model described in step S1, local features are used as vertices and adjacency relationships are used as edges; the instances in the instance library are stored in the form of assemblies.

[0034] Step S2: According to the constructed typical structure instance library, convert the obtained part shape information into a k-dimensional vector, and calculate the set of neighbor vectors based on the obtained spatial neighbor characteristics between parts.

[0035] The specific steps of step S2 are as follows: Step S21: According to the constructed typical structure instance library, obtain the parts p of the typical structure instance library and the product model, and call the shape distribution algorithm to describe the shape information of the parts as a k-dimensional vector; Step S22: According to the constructed typical structure instance library, obtain the typical structure instances, calculate the neighbor characteristics between parts, and obtain the set of neighbor vectors.

[0036] To better implement the present invention, further, the specific steps of step S22 are as follows: Step S221: According to the constructed typical structure instance library, obtain the typical structure instances, and construct the OBB bounding box of the parts; the typical structure instances include the maximum length, maximum width, and maximum height of part p; Step S222: Judge the spatial distribution relationship between parts according to the interference situation of the bounding boxes, and obtain the set of neighbor vectors; Step S223: Repeat steps S221 - S222 until each part in the product model and the typical structure instance library is traversed.

[0037] Step S3: According to the k-dimensional vector and the set of neighbor vectors, search to obtain the matching candidate set of the three-dimensional model of the target product.

[0038] The specific steps of step S3 are as follows: Step S31: According to the set shape similarity threshold and the similarity between part p and part q, roughly filter part p to obtain the initial set; Step S32: According to the initial set, calculate the similarity between the parts in the initial set and the neighbor set of part p; Step S33: According to the set neighbor similarity threshold and the neighbor similarity, match the initial set to obtain the matching candidate set.

[0039] Step S4: According to the matching candidate set, obtain the spatial distribution aggregation of the matching parts, and identify the typical structure composed of the matching parts.

[0040] The specific steps of step S4 are as follows: Step S41: Construct an ordered set according to the size of the part neighbor set in the typical structure instance; Step S42: Initialize the set list, set Ca, set Flag, and set Tar; Step S43: Assign the ordered set to set list. If part p is the first element of set list, assign the matching candidate set to set Ca; Step S44: Randomly select part q1 from set Ca and determine whether the set criterion conditions are met; Step S45: If not, assign Ca - {q1} to set Ca and assign Ca - {q1} to the matching candidate set Ca; p -{q1} is the set obtained by removing the element {q1} from Ca and Ca; p ; If the set , then select the next part q2 in set Ca and return to Step S44; If the set , then end the search; Ca - {q1}, Ca p -{q1} are the sets obtained by removing the element {q1} from Ca and Ca respectively; p ; Step S46: If the set criterion holds, add part p to set Flag and simultaneously establish an ordered set list of the neighbor vector set Nei p ; p ; Step S47: If , then use the part set corresponding to the output set Tar as the typical structure, and assign list ins — Flag to set list, assign to set Ca, and return to Step S44, list ins is the ordered set of the typical structure instance ins ; Step S48: If , then assign to set Tar, assign to set Ca, assign list p to set list, and return to Step S44; Nei q1 is the neighbor vector set of part q1; Step S49: Use the output set Tar as the matching typical structure.

[0041] Working principle: In this embodiment, the process personnel select some typical structures in the product to construct an instance library according to work experience and business needs; then, taking the typical structure instances as inputs, the parts in the typical structures and the 3D model of the target product are preliminarily filtered by calculating the shape similarity, and further optimized by using the neighbor similarity between parts to obtain the candidate matching set of each part in the typical structure instances. Finally, using the depth-first search algorithm, according to the spatial distribution characteristics of the matching parts in the product, the structures identical or similar to the target structure are combined from all the matching parts and output to realize the recognition of the typical structure.

[0042] The other parts of this embodiment are the same as those of the above-mentioned Embodiment 1, so they will not be elaborated here.

[0043] Embodiment 3: On the basis of any one of the above-mentioned Embodiments 1-2, this embodiment is described with a specific embodiment as Figure 1 shown.

[0044] Step S1: According to the obtained STG model, construct an instance library of typical structures according to the process characteristics, and each instance is stored in the form of an assembly.

[0045] Step S21: For each part in the instance library and the product model p , describe the shape information of the part as a k dimensional vector ; Step S22: For the product model and each typical structure instance, calculate the neighbor characteristics between parts by using the bounding box. The specific steps are as follows: Step S221: Obtain the p maximum length, width and height of the part, and construct the OBB bounding box of the part; Step S222: Judge the spatial distribution relationship between parts according to the interference situation of the bounding boxes, and obtain the neighbor vector set of the part p : (1); where: A represents the instance or product model where the part p is located, q represents a part model in A other than p, , respectively represent the OBB bounding boxes of the parts p and q, represents the overlapping area of the bounding boxes in the three-dimensional space.

[0046] Step S223: Repeat Step S221-Step S222 until each part in the product and the instance library is traversed.

[0047] Step S3: For a product modelA and each typical structure instance Ins , according to Ins the part shape and neighbor similarity in A , search for its matching candidate set in the product, and the specific steps are as follows: Step S31: Set the shape similarity threshold , and perform rough filtering on the parts according to the size of the part similarity to obtain the initial set : (2); where s p represents the shape vector of part p, s q represents the shape vector of part q, k represents the dimension of the k-dimensional shape vector of the part, and l j p represents the j-th dimensional shape vector of part p, and l j q represents the j-th dimensional shape vector of part q.

[0048] Step S32: For the initial set after rough filtering, calculate the similarity between each part in it and the neighbor vector set of part p : (3); where represents the size of the neighbor vector set of part p .

[0049] Step S33: Set the shape neighbor similarity threshold , and perform fine matching on according to the size of the neighbor similarity to obtain the matching candidate set of part p ; (4).

[0050] Step S4: According to the matching candidate set of each part in Ins , realize typical structure recognition according to the spatial distribution aggregation of the matching parts: Step S41: Construct an ordered set according to the size of the neighbor set of each part in Ins : (5); Step S42: Initialize the sets list , Ca , Flag , Tar ; Step S43: Assign a value , assumingp If it is list the first element of, then perform assignment ; Step S44: Assume that list the first element is p , select Ca one part in , and make the following judgment: (6); Step S45: If this criterion does not hold, then perform assignment , , if the set , then select Ca the next part q2 in, and return to step S44; if the set , then end the search; Step S46: If this criterion holds, then add the part p to the set Flag , and at the same time establish an ordered set of : (7); Step S47: If , then output Tar the corresponding part set as the typical structure, and at the same time perform the following assignment and return to step S44; (8); Step S48: If , then perform the following assignment and return to step S44; (9); Step S49: Complete the above recursive process, and all the output Tar is the matching typical structure.

[0051] Working principle: The recognition of the typical structure in this embodiment is realized through a pre-defined instance library. Technicians can control the recognition object by adding, deleting, and modifying the instance library, which has good scalability; replacing the part connection relationship in the existing method with the spatial distribution relationship between parts can effectively avoid the problem of difficulty in obtaining the part connection relationship in complex products.

[0052] Other parts of this embodiment are the same as any one of the above Embodiment 1 - Embodiment 2, so they will not be elaborated here.

[0053] Embodiment 4: Based on any one of the above Embodiment 1 - Embodiment 3, such as Figure 2 , Figure 3 , Figure 4 , Figure 5 ,Figure 6 , Figure 7 As shown in Figure 7 , taking the clamp with a clamping function commonly used in the mechanical industry as an example, the specific steps of the automatic recognition method for the typical structure of complex products based on instance library matching are as follows: Step S1: The process personnel construct an instance library of typical structures according to the process characteristics, and each instance is stored in the form of an assembly. In this embodiment, two typical structure instances are defined, as Figure 2 shown; Step S2: For each part in the instance library and the product model, the shape distribution algorithm is used to describe the shape information of the part as a 50-dimensional vector , and the description results of some parts are as Figure 3 shown; For the product model and each typical structure instance, the nearest neighbor characteristics between parts are calculated using the bounding box. The specific steps are as follows: Step S221: Obtain the maximum length, width, and height of a part p in the model, and construct the OBB bounding box of the part; Step S222: Judge the spatial distribution relationship between parts according to the bounding box interference situation, and obtain the nearest neighbor vector set: (10); Taking the connecting lug in the first typical structure instance as an example, the parts included in its nearest neighbor vector set are as Figure 3 shown.

[0054] Step S223: Repeat Step S221 - Step S222 until each part in the product and the instance library is traversed; Step S3: Given a pallet product model as Figure 4 shown A , for Figure 2 each typical structure instance Ins in Ins , search for its matching candidate set in the product A according to the part shape and nearest neighbor similarity in Step S31: Set the shape similarity threshold to 0.95, and perform rough filtering on the parts according to the size of the part similarity to obtain the initial set : (11); Step S32: For the initial set after rough filtering, calculate the nearest neighbor set similarity between each part in it and the part p : (12); Among them, Indicates parts p Size of the neighbor set.

[0055] Step S33: Set the shape neighbor similarity threshold to 0.80, and perform fine matching on according to the size of the neighbor similarity to obtain a matching candidate set ; ; (13); Taking the rivet of the first typical structure as an example, the matching candidate set after fine matching is as Figure 5 shown.

[0056] Step S4: Based on the matching candidate sets of each part in Ins , consider the spatial distribution aggregation of the matching parts to achieve typical structure recognition. The specific process of the recognition process is as Figure 6 shown, specifically including the following steps: Step S41: Construct an ordered set according to the size of the neighbor vector sets of each part in Ins : (14); Step S42: Initialize the sets list , Ca , Flag , Tar ; Step S43: Assign a value to , assuming p is the first element of list , then perform the assignment ; Step S44: Select a part Ca in , and make the following judgments: (15); Step S45: If this criterion is not met, then perform the assignment , . If the set , then select the next part in , and return to Step S44; if the set , then end the search; Step S46: If this criterion is met, then add the part p to the set Flag , and at the same time establish an ordered set of : (16); Step S47: If , then outputTar The corresponding set of parts is used as a typical structure, and the following assignments are made simultaneously, then return to step S44; (17); Step S48: If , then make the following assignments and return to step S44; (18); Step S49: Complete the above recursive process. At this time, the output Tar set is output a total of 1 time, which contains 13 part models, as Figure 7 shown. These parts represent a typical structure in the product model A .

[0057] This embodiment shows that the automatic recognition method for typical structures of complex products proposed by the present invention can be used for the automatic recognition of predefined structures and can achieve good results.

[0058] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention without departing from the principles and purposes of this embodiment.

[0059] Other parts of this embodiment are the same as any one of the above Embodiment 1 - Embodiment 3, so they will not be described again.

[0060] Embodiment 5: Based on any one of the above Embodiment 1 - Embodiment 4, this embodiment illustrates the establishment process of the STG model with a specific example.

[0061] The establishment process of the STG model is as follows: First, establish a vertex adjacency graph according to the topological relationship between geometric elements of the obtained B - rep data; secondly, search for the maximum cliques in the adjacency graph and use the geometric regions corresponding to the maximum cliques as local features of the solid model; then call statistical methods to transform the shape information of the local features into feature vectors to form a feature space, and use an unsupervised learning algorithm to suppress the local features differently; finally, establish an STG model with local features as vertices and adjacency relationships as edges.

[0062] Other parts of this embodiment are the same as any one of the above Embodiment 1 - Embodiment 4, so they will not be described again.

[0063] Embodiment 6: Based on any one of the above-mentioned Embodiment 1 - Embodiment 5, this embodiment proposes an automatic recognition system for typical structures of complex products based on instance library matching, which is used to execute the above-mentioned automatic recognition method for typical structures of complex products based on instance library matching; it includes an instance library construction unit, a feature acquisition unit, a matching candidate unit, and an identification unit; The instance library construction unit is used to construct a typical structure instance library according to the obtained STG model; The feature acquisition unit is used to obtain part shape information and the spatial proximity characteristics between parts according to the constructed typical structure instance library; The matching candidate unit is used to obtain a matching candidate set of the 3D model of the target product according to the part shape information and the spatial proximity characteristics between parts; The identification unit is used to obtain the spatial distribution aggregation of the matching parts according to the matching candidate set, and identify the typical structure.

[0064] This embodiment also proposes an electronic device, including a memory and a processor; a computer program is stored on the memory; when the computer program is executed on the processor, the above-mentioned automatic recognition method for typical structures of complex products based on instance library matching is implemented.

[0065] This embodiment also proposes a computer-readable storage medium, on which a computer instruction is stored; when the computer instruction is executed on the above-mentioned electronic device, the above-mentioned automatic recognition method for typical structures of complex products based on instance library matching is implemented.

[0066] Other parts of this embodiment are the same as any one of the above-mentioned Embodiment 1 - Embodiment 5, so they will not be elaborated here.

[0067] The processor involved in the embodiments of this application may be a chip. For example, it may be a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on chip (SoC), a central processing unit (CPU), a network processor (NP), a digital signal processing circuit (DSP), a microcontroller unit (MCU), a programmable logic device (PLD), or other integrated chips.

[0068] The memory involved in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchlink dynamic random access memory (SLDRAM), and direct rambus random access memory (DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include but not be limited to these and any other suitable types of memory.

[0069] It should be understood that in various embodiments of the present application, the sequence numbers of the above processes do not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0070] Those of ordinary skill in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0071] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and modules described above can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0072] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling, direct coupling, or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of devices or modules can be in electrical, mechanical, or other forms.

[0073] The modules described as separate components may or may not be physically separated. The components displayed as modules may or may not be physical modules, that is, they can be located in one device, or they can be distributed to multiple devices. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0074] In addition, in each embodiment of the present application, the functional modules can be integrated in one device, or each module can exist physically alone, or two or more modules can be integrated in one device.

[0075] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using a software program, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website, computer, server, or data center to another website, computer, server, or data center in a wired manner (such as coaxial cable, optical fiber, Digital Subscriber Line (DSL)) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that contains one or more integrated media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a Solid State Disk (SSD)), etc.

[0076] As described above, this is only the specific implementation manner of the present application. However, the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. An automatic recognition method for typical structures of complex products based on instance library matching, characterized in that, It includes the following steps: Step S1: According to the obtained STG model, construct a typical structure instance library. The STG model is a model established with local features of a 3D model as vertices and adjacency relationships between local features as edges. Instances in the typical structure instance library are assemblies; Step S2: According to the constructed typical structure instance library, obtain part shape information and spatial proximity characteristics between parts. Convert the obtained part shape information into a k-dimensional vector, and calculate a set of proximity vectors according to the obtained spatial proximity characteristics between parts; Step S3: Calculate the similarity between parts according to the k-dimensional vectors of parts, and obtain an initial set of similar parts according to the similarity; Calculate the proximity set similarity between parts in the initial set according to the set of proximity vectors, and obtain a matching candidate set corresponding to each part according to the proximity set similarity; Step S4: Identify matching parts according to the matching candidate set of each part in the typical structure instance and the spatial distribution aggregation of matching parts, and obtain a typical structure composed of matching parts.

2. The automatic recognition method for the typical structure of complex products based on instance library matching according to claim 1, wherein The specific steps of step S2 include the following steps: Step S21: According to the constructed typical structure instance library, obtain parts of the typical structure instance and the product model, and call the shape distribution algorithm to describe the shape information of the parts as a k-dimensional vector; Step S22: According to the constructed typical structure instance library, obtain the typical structure instance, calculate the proximity characteristics between parts, and obtain a set of proximity vectors.

3. The automatic recognition method for the typical structure of complex products based on case library matching according to claim 2, wherein The specific steps of step S22 include the following steps: Step S221: According to the constructed typical structure instance library, obtain the typical structure instance, and construct an OBB bounding box for the parts; The typical structure instance includes the maximum length, maximum width, and maximum height of the parts; Step S222: Judge the spatial distribution relationship between parts according to the interference situation of the bounding boxes, and obtain a set of proximity vectors; Step S223: Repeat steps S221 - S222 until each part in the product model and the typical structure instance library is traversed.

4. The automatic recognition method for typical structures of complex products based on instance library matching according to claim 1, characterized in that The specific steps of step S3 include the following steps: Step S31: Coarsely filter part p according to the set shape similarity threshold and the similarity between part p and part q to obtain an initial set; Step S32: According to the initial set, calculate the proximity set similarity between the parts in the initial set and part p; Step S33: Match the initial set according to the set proximity similarity threshold and proximity similarity to obtain a matching candidate set.

5. The automatic recognition method for typical structures of complex products based on instance library matching according to claim 4, characterized in that The specific steps of step S4 include the following steps: Step S41: Construct an ordered set according to the size of the part proximity set in the typical structure instance; Step S42: Initialize set list, set Ca, set Flag, and set Tar; Step S43: Assign the ordered set to set list. If part p is the first element of set list, assign the matching candidate set to set Ca; Step S44: Randomly select part q1 from set Ca and judge whether it meets the set criterion conditions; Step S45: If not satisfied, assign Ca - {q1} to set Ca, and assign Ca p - {q1} to the matching candidate set Ca of part p p ; If the set , then select the next part q2 in set Ca and return to step S44; If the set , then end the search; Ca - {q1}, Ca p - {q1} are the sets obtained by removing element {q1} from Ca and Ca p respectively; Step S46: If the set criterion holds, add part p to the set Flag, and at the same time establish an ordered set list of the neighbor vector set Nei of part p p of the ordered set list p ; Step S47: If , then take the part set corresponding to the output set Tar as the typical structure, and assign list ins — Flag to the set list, assign to the set Ca, and return to step S44, list ins is an ordered set of typical structure instances ins ; Step S48: If , then assign to the set Tar, assign to the set Ca, assign list p to the set list, and return to step S44; it is the set of neighbor vectors of part q1; Step S49: Use the output set Tar as the matching typical structure.

6. The automatic recognition method for typical structures of complex products based on instance library matching according to claim 5, characterized in that The criterion conditions set in step S44 are: ； Among them, p is the first element of the initialized set list, Nei p represents the set of neighbor vectors of the part p , Nei q1 represents the set of neighbor vectors of the part q 1 . u and v are respectively a neighbor vector in Nei p and Nei q1 , and represents the matching candidate set of u.

7. A method for automatically identifying typical structures of complex products based on instance library matching according to claim 1, characterized in that The establishment process of the STG model described in step S1 is as follows: First, establish a vertex adjacency graph according to the topological relationship between the geometric elements of the obtained B-rep data; secondly, search for the maximum cliques in the adjacency graph and use the geometric regions corresponding to the maximum cliques as the local features of the solid model; then call the statistical method to transform the shape information of the local features into feature vectors, form a feature space, and use the unsupervised learning algorithm to suppress the local features differentially; finally, establish an STG model with the local features as vertices and the adjacency relationship between the local features as edges.

8. A complex product typical structure automatic recognition system based on case library matching, which is used to execute the complex product typical structure automatic recognition method based on case library matching as described in claim 1; characterized in that, It includes an instance library construction unit, a feature acquisition unit, a matching candidate unit, and an identification unit; The instance library construction unit is used to construct a typical structure instance library according to the obtained STG model; The feature acquisition unit is used to obtain the part shape information and the spatial proximity characteristics between parts according to the constructed typical structure instance library; The matching candidate unit is used to obtain a matching candidate set of the three-dimensional model of the target product according to the part shape information and the spatial proximity characteristics between parts; The identification unit is used to obtain the spatial distribution aggregation of the matching parts according to the matching candidate set and identify the typical structure.

9. An electronic device, characterized in that, It includes a memory and a processor; a computer program is stored on the memory; when the computer program is executed on the processor, the automatic recognition method for the typical structure of a complex product based on instance library matching as described in any one of claims 1-7 is implemented.

10. A computer-readable storage medium, characterized in that, A computer instruction is stored on the computer-readable storage medium; when the computer instruction is executed on the electronic device as described in claim 9, the automatic recognition method for the typical structure of a complex product based on instance library matching as described in any one of claims 1-7 is implemented.

Images