Fuselage skin AO planning method, system and equipment based on typical characteristics and medium

Abstract

The invention relates to the technical field of aircraft manufacturing, in particular to a fuselage skin AO planning method, system and equipment based on typical characteristics and a medium. The method comprises the following steps: firstly, acquiring fuselage skin data according to acquired three-dimensional model data, and constructing a fuselage skin library; secondly, according to the aircraft three-dimensional model, extracting and constructing a fuselage skin typical planning feature library; analyzing a typical planning feature library of the fuselage skin, determining the relationship between the features and the skin in combination with a knowledge graph technology, reasoning the fuselage skin AO, and forming a whole-aircraft fuselage skin assembly content set; and finally, calling a graph traversal algorithm, and checking the integrity of the whole aircraft body skin assembly content set. The structural expression of the fuselage skin AO planning process is improved, and the planning efficiency and accuracy are effectively improved.

Description

Technical Field

[0001] The invention relates to the technical field of aircraft manufacturing, and in particular to a fuselage skin AO planning method, system, equipment and medium based on typical features. Background Art

[0002] With the rapid development of information technology such as knowledge graphs and machine learning, aircraft manufacturing has also transformed from two-dimensional drawings to a manufacturing model based on three-dimensional models. The fuselage skin, also known as the fuselage shell, is a very important part of the aircraft structure. It covers the aircraft skeleton to form the aircraft's appearance. It mainly has the functions of aircraft aerodynamic shape, protecting internal structures and systems, serving as a load-bearing structure, and shock absorption. The fuselage skin is spread throughout the aircraft, and its characteristics include a large number of fuselage skins and complex assembly conditions. The assembly content of the fuselage skin is mainly carried in the form of an assembly outline (AO), which is mainly based on manual experience combined with the characteristics of the fuselage skins of various parts of the aircraft to form the final assembly content collection. In the fuselage skin planning method based on manual experience, there are problems such as low efficiency, difficulty in ensuring completeness, and inability to accumulate process knowledge. Summary of the invention

[0003] Aiming at the problems of low efficiency, difficulty in ensuring completeness and inability to accumulate process knowledge of fuselage skin, the present invention proposes a fuselage skin AO planning method, system, equipment and medium based on typical features; the method constructs a fuselage skin typical structural feature library, a fuselage skin object library, and a fuselage skin typical planning feature library, and builds a graph network for fuselage skin planning based on various parameter libraries to perform structured graph expression on its planning process and results, thereby effectively improving the planning efficiency and accuracy.

[0004] A fuselage skin AO planning method based on typical features specifically comprises the following steps: Step S1: acquiring fuselage skin data according to the acquired three-dimensional model data, and constructing a fuselage skin library; Step S2: extracting and constructing a typical planning feature library of the fuselage skin according to the three-dimensional model of the aircraft; Step S3: Analyze the typical planning feature library of the fuselage skin, combine the knowledge graph technology, clarify the relationship between the features and the skin, infer the fuselage skin AO, and form the full fuselage skin assembly content set; Step S4: Call the graph traversal algorithm to check the completeness of the whole fuselage skin assembly content set.

[0005] In order to better implement the present invention, further, the specific operation of step S1 is: according to the appearance characteristics of the aircraft skin, the fuselage skin data is obtained, and the fuselage skin data set is defined as the full-aircraft fuselage skin library D{D1, D2, D3...Dt}, wherein Dt represents the tth specific fuselage skin object; the typical structural characteristics of the fuselage skin object are {T1, T2, T3...Tm}, wherein Tm represents the mth typical characteristic of the fuselage skin.

[0006] In order to better implement the present invention, further, the step S2 specifically includes the following steps: Step S21: extracting the fuselage skin AO planned by the recognition process experience by using semantic recognition, obtaining the main influencing factors affecting the fuselage skin AO, and recording them as Q{Q1, Q2, Q3...QN}, where QN is the Nth main influencing factor affecting the fuselage skin AO; Step S22: Obtain the frequency of occurrence of the impact factor QJ and record it as PJ; Step S23: The frequency of occurrence in the assembly content is regarded as the important frequency of occurrence and recorded as Ej; Step S24: taking the n features with the highest frequency as typical planning features of the fuselage skin and recording them as F{F1, F2, F3...Fn}; Step S25: Call the BERT model to extract the relationship type between typical planning features of the fuselage skin and generate triples<Fi,Rk,Fj> A typical planning feature graph network of the fuselage skin is constructed in the form of; where: Fi represents the i-th feature of the fuselage skin planning feature, Fj represents the j-th feature of the fuselage skin planning feature, and Rk represents the relationship type between the two.

[0007] In order to better implement the present invention, further, the step S3 specifically includes the following steps: Step S31: Based on the attention mechanism and weight mechanism, analyze the importance value T of each node and the importance weight of its neighboring nodes to it, recorded as W{W1,W2…W i}, where W i is the importance weight of the i-th neighboring node to it; Step S32: Taking the fuselage skin object as input, semantic matching and logic verification are performed on it in the fuselage skin typical planning feature graph network to form the final AO assembly content.

[0008] In order to better implement the present invention, further, the step S32 specifically includes the following steps: Step S321: Obtain the fuselage skin object Di and its typical structural features in the fuselage skin library; Step S322: semantically matching a local network whose structural features are satisfied in the typical planning feature network; Step S323: forming the final AO assembly content based on the screened local network; Step S324: loop through the fuselage skin library objects.

[0009] In order to better implement the present invention, further, the generation logic of the final AO assembly content is specifically as follows: using a deep traversal algorithm to find the root node, calculate the importance weight values ​​of the root node's child nodes, obtain the nodes and node information on the route with the maximum weight value as the assembly content of a single AO, and if there are h root nodes, generate h AO assembly contents.

[0010] In order to better implement the present invention, further, the specific operation of step S4 is: using graph depth traversal and breadth traversal algorithms to search whether there are "unique points" or "breakpoints" in the full fuselage skin instance assembly content, determine whether all fuselage skin structural features have nodes and relationships in the instance graph, and form an evaluation conclusion with the results.

[0011] Based on the above-mentioned fuselage skin AO planning method based on typical features, in order to better realize the present invention, further, a fuselage skin AO planning system based on typical features is proposed, which is used to execute the above-mentioned fuselage skin AO planning method based on typical features; it includes a construction unit, an inference unit, and a planning unit; The construction unit is used to obtain the data of the fuselage skin according to the acquired three-dimensional model data, and construct a fuselage skin library; according to the three-dimensional model of the aircraft, extract and construct a typical planning feature library of the fuselage skin; The reasoning unit is used to analyze the typical planning feature library of the fuselage skin, combine the knowledge graph technology, clarify the relationship between the feature and the skin, reason the fuselage skin AO, and form the whole fuselage skin assembly content set; The planning unit is used to call the graph traversal algorithm to check the completeness of the whole fuselage skin assembly content set.

[0012] Based on the above-mentioned typical feature-based fuselage skin AO planning method, in order to better realize 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 typical feature-based fuselage skin AO planning method is implemented.

[0013] Based on the above-mentioned typical feature-based fuselage skin AO planning method, in order to better realize the present invention, further, a computer-readable storage medium is proposed, on which computer instructions are stored; when the computer instructions are executed on the above-mentioned electronic device, the above-mentioned typical feature-based fuselage skin AO planning method is implemented.

[0014] The present invention has the following beneficial effects: The present invention realizes the construction of a fuselage skin typical structural feature library, a fuselage skin object library, a fuselage skin typical planning feature library, and a fuselage skin AO assembly content set network, thereby improving the structured expression of the fuselage skin AO planning process and effectively improving the efficiency and accuracy of planning. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A schematic flowchart of the fuselage skin AO planning method based on typical features provided by the present invention.

[0016] Figure 2 A schematic block diagram of the reasoning process provided by the present invention. DETAILED DESCRIPTION

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. It should be understood that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments, and therefore should not be regarded as limiting the scope of protection. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technical personnel in this field without making creative work are within the scope of protection of the present invention.

[0018] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "disposed", "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 be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0019] Embodiment 1: This embodiment proposes a fuselage skin AO planning method based on typical features, which specifically includes the following steps: Step S1: acquiring fuselage skin data according to the acquired three-dimensional model data, and constructing a fuselage skin library; The specific operation of step S1 is: according to the appearance characteristics of the aircraft skin, fuselage skin data is obtained, and the fuselage skin data set is defined as a full-aircraft fuselage skin library D{D1, D2, D3...Dt}, where Dt represents the t-th specific fuselage skin object; the typical structural characteristics of the fuselage skin object are {T1, T2, T3...Tm}, where Tm represents the m-th typical characteristic of the fuselage skin.

[0020] Step S2: extracting and constructing a typical planning feature library of the fuselage skin according to the three-dimensional model of the aircraft; The step S2 specifically includes the following steps: Step S21: extracting the fuselage skin AO planned by the recognition process experience by using semantic recognition, obtaining the main influencing factors affecting the fuselage skin AO, and recording them as Q{Q1, Q2, Q3...QN}, where QN is the Nth main influencing factor affecting the fuselage skin AO; Step S22: Obtain the frequency of occurrence of the impact factor QJ and record it as PJ; Step S23: The frequency of occurrence in the assembly content is regarded as the important frequency of occurrence and recorded as Ej; Step S24: taking the n features with the highest frequency as typical planning features of the fuselage skin and recording them as F{F1, F2, F3...Fn}; Step S25: Call the BERT model to extract the relationship type between typical planning features of the fuselage skin and generate triples<Fi,Rk,Fj> A typical planning feature graph network of the fuselage skin is constructed in the form of.

[0021] Step S3: Analyze the typical planning feature library of the fuselage skin, combine the knowledge graph technology, clarify the relationship between the features and the skin, infer the fuselage skin AO, and form the full fuselage skin assembly content set; Furthermore, the step S3 specifically includes the following steps: Step S31: Based on the attention mechanism and weight mechanism, analyze the importance value T of each node and the importance weight of its neighboring nodes to it, recorded as W{W1,W2…W i}, where W i is the importance weight of the i-th neighboring node to it; Step S32: Taking the fuselage skin object as input, semantic matching and logic verification are performed on it in the fuselage skin typical planning feature graph network to form the final AO assembly content.

[0022] In order to better implement the present invention, further, the step S32 specifically includes the following steps: Step S321: Obtain the fuselage skin object Di and its typical structural features in the fuselage skin library; Step S322: semantically matching a local network whose structural features are satisfied in the typical planning feature network; Step S323: forming the final AO assembly content based on the screened local network; The generation logic of the final AO assembly content is specifically as follows: using a deep traversal algorithm to find the root node, calculating the importance weight values ​​of the root node's child nodes, obtaining the nodes and node information on the route with the maximum weight value as the assembly content of a single AO, and if there are h root nodes, generating h AO assembly contents.

[0023] Step S324: loop through the fuselage skin library objects.

[0024] Step S4: Call the graph traversal algorithm to check the completeness of the whole fuselage skin assembly content set.

[0025] The specific operation of step S4 is: using the graph depth traversal and breadth traversal algorithms to search whether there are "unique points" or "breakpoints" in the full fuselage skin instance assembly content, determine whether all the fuselage skin structural features have nodes and relationships in the instance graph, and form an evaluation conclusion based on the results.

[0026] Working principle: This embodiment first obtains the data of the fuselage skin according to the acquired three-dimensional model data and constructs a fuselage skin library; secondly, according to the aircraft three-dimensional model, extracts and constructs a typical planning feature library of the fuselage skin; then analyzes the typical planning feature library of the fuselage skin, combines the knowledge graph technology, clarifies the relationship between the features and the skin, infers the fuselage skin AO, and forms the whole fuselage skin assembly content set; finally, calls the graph traversal algorithm to check the completeness of the whole fuselage skin assembly content set; improves the structured expression of the fuselage skin AO planning process, and effectively improves the efficiency and accuracy of planning.

[0027] Embodiment 2: This embodiment is based on the above embodiment 1. Figure 1 , Figure 2 As shown, a specific embodiment is described in detail, which specifically includes the following steps.

[0028] Step S1: Build the fuselage skin library.

[0029] The fuselage skin of the whole aircraft has different shapes and distribution positions. Based on the data of the three-dimensional model, the appearance characteristics of the skin are defined, which is the outermost layer of the fuselage. Based on this typical feature, the data of the fuselage skin is obtained, mainly obtaining the appearance feature information, drawing number, name and other information, and the fuselage skin data set is defined as the fuselage skin library D{D1, D2, D3...Dt} of the whole aircraft, where Dt represents the tth specific fuselage skin object; the typical structural characteristics of the fuselage skin object are {T1, T2, T3...Tm}, where Tm represents the mth typical feature of the fuselage skin.

[0030] Taking a certain aircraft model as an example, there are more than 500 fuselage skin objects in total. The typical structural features extracted mainly include T{structural parts, fuel tank area, rib type, material type, name, drawing number, and belonging components} and other information.

[0031] Step S2: Based on the aircraft 3D model, extract and construct a typical planning feature library of the fuselage skin.

[0032] Acquiring and defining the fuselage skin library and its three-dimensional model features is only to extract the three-dimensional digital model into two-dimensional vector feature data to support the subsequent data feature processing. Data feature processing is strongly related to the area where the fuselage skin is located, the structural characteristics of the fuselage skin, the structural characteristics of the aircraft, etc. The traditional method is based on the experience of process personnel. By adopting methods such as semantic recognition, the fuselage skin AO previously planned based on process experience is extracted and identified, and the main influencing factors affecting the fuselage skin AO planning are obtained and recorded as Q{Q1,Q2,Q3...QN}, where QN represents the Nth influencing factor. And for each influencing factor, its frequency of occurrence is counted and recorded as P J The frequency of occurrence in the assembly content is determined, and the important frequency is recorded as E j The 10 most frequent features are obtained as typical design features of the fuselage skin, and their features are defined as F{F1, F2, F3…Fn}.

[0033] The optimized BERT algorithm is used to extract the relationship types between the typical planning features of the fuselage skin and the relationship between them in triples.<Fi,Rk,Fj> Among them, Fi represents the i-th feature of the fuselage skin planning feature, Fj represents the j-th feature of the fuselage skin planning feature, and Rk represents the relationship type between the two. Based on the relationship of the triples, a typical planning feature graph network of the fuselage skin is constructed.

[0034] By extracting the previous fuselage cover skin AO data, the main influencing factors are obtained, including: connected objects (structural parts or inside the skin), double bolts, support plate nuts, inside and outside of the fuel tank, rib type and other planning features. The relationship includes whether it is inside or outside the fuel tank, whether there are ribs, and whether it is connected to the skeleton. A total of 7 root nodes are defined, and a single skin can generate up to 7 AOs.

[0035] Step S3: Analyze the typical feature library of the fuselage skin, and combine it with the knowledge graph technology to clarify the relationship between the features and the skin, infer the fuselage skin AO, and form the full fuselage skin assembly content set.

[0036] Taking the fuselage skin object as input, it is semantically matched and logically verified in the fuselage skin typical planning feature graph network to form the final AO assembly content. The main steps are as follows: Firstly, obtain the fuselage skin object Di and its typical structural features in the fuselage skin library; Secondly, in the typical planning feature network, the semantics matches the local network whose structural features are satisfied; Then, based on the selected local network, the final AO assembly content is formed. The generation logic is to use the depth traversal algorithm to find the root node, the nodes on the route and their information as the assembly content of a single AO. If there are h (up to 7) root nodes, h AO assembly contents are generated.

[0037] Finally loop through the fuselage skin library objects.

[0038] Through the above steps, the AO of the fuselage skin cover can be generated.

[0039] Step S4: Use a graph traversal algorithm to check the completeness of the entire fuselage skin assembly content set.

[0040] Using graph depth traversal and breadth traversal algorithms, we search for "unique points" and "breakpoints" in the full fuselage skin instance assembly content. We determine whether all fuselage skin structural features have nodes and relationships in the instance graph. And we form an evaluation conclusion based on the results.

[0041] Through analysis, taking a certain aircraft as an example, it is possible to accurately determine whether it is missing equipment, thereby improving the efficiency and accuracy of AO planning.

[0042] Working principle: In order to solve the problems of low efficiency of fuselage skin, difficulty in ensuring completeness, and inability to accumulate process technology knowledge, this embodiment proposes a fuselage skin AO planning method based on typical features in combination with a three-dimensional model to improve the efficiency of fuselage skin planning and carry process planning knowledge; it realizes the construction of a fuselage skin typical structural feature library, a fuselage skin object library, a fuselage skin typical planning feature library, and a fuselage skin AO assembly content set network, and builds a fuselage skin planning graph network based on various parameter libraries to perform structured graph expression on its planning process and results, thereby improving the structured expression of the fuselage skin AO planning process and effectively improving the planning efficiency.

[0043] The other parts of this embodiment are the same as those of the above-mentioned embodiment 1, and thus will not be described in detail.

[0044] Embodiment 3: This embodiment, based on any one of the above-mentioned embodiments 1-2, proposes a fuselage skin AO planning system based on typical features, which is used to execute the above-mentioned fuselage skin AO planning method based on typical features; it includes a construction unit, a reasoning unit, and a planning unit; The construction unit is used to obtain the data of the fuselage skin according to the acquired three-dimensional model data, and construct a fuselage skin library; according to the three-dimensional model of the aircraft, extract and construct a typical planning feature library of the fuselage skin; The reasoning unit is used to analyze the typical planning feature library of the fuselage skin, combine the knowledge graph technology, clarify the relationship between the feature and the skin, reason the fuselage skin AO, and form the whole fuselage skin assembly content set; The planning unit is used to call the graph traversal algorithm to check the completeness of the whole fuselage skin assembly content set.

[0045] This embodiment also proposes an electronic device, including a memory and a processor; a computer program is stored in the memory; when the computer program is executed on the processor, the above-mentioned fuselage skin AO planning method based on typical features is implemented.

[0046] This embodiment also proposes a computer-readable storage medium, on which computer instructions are stored; when the computer instructions are executed on the above-mentioned electronic device, the above-mentioned fuselage skin AO planning method based on typical features is implemented.

[0047] The other parts of this embodiment are the same as any one of the above-mentioned embodiments 1-2, so they will not be repeated here.

[0048] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any simple modification or equivalent change made to the above embodiment based on the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A fuselage skin AO planning method based on typical features, characterized in that: The specific steps include: Step S1: acquiring fuselage skin data according to the acquired three-dimensional model data, and constructing a fuselage skin library; Step S2: extracting and constructing a typical planning feature library of the fuselage skin according to the three-dimensional model of the aircraft; Step S3: Analyze the typical planning feature library of the fuselage skin, combine the knowledge graph technology, clarify the relationship between the features and the skin, infer the fuselage skin AO, and form the full fuselage skin assembly content set; Step S4: Call the graph traversal algorithm to check the completeness of the whole fuselage skin assembly content set.

2. The fuselage skin AO planning method based on typical features according to claim 1, characterized in that: The specific operation of step S1 is: according to the appearance characteristics of the aircraft skin, fuselage skin data is obtained, and the fuselage skin data set is defined as a full-aircraft fuselage skin library D{D1, D2, D3...Dt}, where Dt represents the t-th specific fuselage skin object; the typical structural characteristics of the fuselage skin object are {T1, T2, T3...Tm}, where Tm represents the m-th typical characteristic of the fuselage skin.

3. The fuselage skin AO planning method based on typical features according to claim 2, characterized in that: The step S2 specifically includes the following steps: Step S21: extracting the fuselage skin AO planned by the recognition process experience by using semantic recognition, obtaining the main influencing factors affecting the fuselage skin AO, and recording them as Q{Q1, Q2, Q3...QN}, where QN is the Nth main influencing factor affecting the fuselage skin AO; Step S22: Obtain the frequency of occurrence of the impact factor QJ and record it as PJ; Step S23: The frequency of occurrence in the assembly content is regarded as the important frequency of occurrence and recorded as Ej; Step S24: taking the n features with the highest frequency as typical planning features of the fuselage skin and recording them as F{F1, F2, F3...Fn}; Step S25: Call the BERT model to extract the relationship type between typical planning features of the fuselage skin and generate triples<Fi,Rk,Fj> A typical planning feature graph network of the fuselage skin is constructed in the form of; where: Fi represents the i-th feature of the fuselage skin planning feature, Fj represents the j-th feature of the fuselage skin planning feature, and Rk represents the relationship type between the two.

4. The fuselage skin AO planning method based on typical features according to claim 3, characterized in that: The step S3 specifically comprises the following steps: Step S31: Based on the attention mechanism and weight mechanism, analyze the importance value T of each node and the importance weight of its neighboring nodes to it, recorded as W{W1,W2…W i }, where W i is the importance weight of the i-th neighboring node to it; Step S32: Taking the fuselage skin object as input, semantic matching and logic verification are performed on it in the fuselage skin typical planning feature graph network to form the final AO assembly content.

5. The fuselage skin AO planning method based on typical features according to claim 4, characterized in that: The step S32 specifically includes the following steps: Step S321: Obtain the fuselage skin object Di and its typical structural features in the fuselage skin library; Step S322: semantically matching a local network whose structural features are satisfied in the typical planning feature network; Step S323: forming the final AO assembly content based on the screened local network; Step S324: loop through the fuselage skin library objects.

6. The fuselage skin AO planning method based on typical features according to claim 5, characterized in that: The generation logic of the final AO assembly content is as follows: using the depth traversal algorithm, find the root node, calculate the importance weight value of the root node child node, obtain the nodes and node information on the route with the maximum weight as the assembly content of a single AO, and if there is h root nodes, then generate h AO assembly content.

7. The fuselage skin AO planning method based on typical features according to claim 4, characterized in that: The specific operation of step S4 is: using the graph depth traversal and breadth traversal algorithms to search whether there are "unique points" and "breakpoints" in the full fuselage skin instance assembly content, determine whether all the fuselage skin structural features have nodes and relationships in the instance graph, and form an evaluation conclusion with the results.

8. A fuselage skin AO planning system based on typical features, used to execute the fuselage skin AO planning method based on typical features as claimed in claim 1; characterized in that: It includes construction unit, reasoning unit and planning unit; The construction unit is used to obtain fuselage skin data according to the acquired three-dimensional model data, and to construct a fuselage skin library; According to the aircraft 3D model, extract and construct the typical planning feature library of fuselage skin; The reasoning unit is used to analyze the typical planning feature library of the fuselage skin, combine the knowledge graph technology, clarify the relationship between the feature and the skin, reason the fuselage skin AO, and form the whole fuselage skin assembly content set; The planning unit is used to call the graph traversal algorithm to check the completeness of the whole aircraft fuselage skin assembly content set.

9. An electronic device, characterized in that: It comprises a memory and a processor; a computer program is stored in the memory; when the computer program is executed on the processor, the fuselage skin AO planning method based on typical features as described in any one of claims 1-7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions; when the computer instructions are executed on the electronic device as described in claim 9, the fuselage skin AO planning method based on typical features as described in any one of claims 1-7 is implemented.

Images