Rapid identification method for aircraft thin-wall opening frame type parts

Abstract

The invention discloses a rapid identification method for an airplane thin-wall opening frame type part, which comprises the following steps: calculating the size of a bounding box of the part, reversely measuring the thickness of the part along the normal vector of a reference plane with the maximum area, and judging whether a part mathematical model is a thin-wall type part or not according to the thickness of the part; secondly, projection in the calibration direction is further carried out on the part mathematical model, so that an inner closed-loop structure and an outer closed-loop structure of the part are obtained through recognition on a projection plane, and the part mathematical model is preliminarily screened by calculating the relation between the perimeter of the closed-loop structures and a perimeter threshold value; then, the ratio of the area of the outer closed-loop structure to the area of the inner closed-loop structure is calculated, the shape consistency of the outer closed-loop structure and the inner closed-loop structure is calculated, and then whether the part digital model is a thin-wall opening frame type part or not is efficiently and accurately judged; according to the method, typical features such as the thickness of the part model and the consistency of projected image edges are identified by adopting image identification and statistical methods, so that the part digital model is identified, and whether the part digital model is a thin-wall opening frame part can be efficiently and accurately judged.

Description

Technical Field

[0001] The invention belongs to the technical field of parts identification, and in particular relates to a method for quickly identifying aircraft thin-walled mouth frame parts. Background Art

[0002] In the field of industrial design and manufacturing, the geometric shape of parts directly affects their manufacturing and assembly processes. Thin-walled mouth frame parts have typical geometric features of thin thickness, ring shape, and hollowness. They are usually used between the aircraft skin and the mouth cover. They can be used to assist in the positioning of mouth cover components with different shapes and functions, thereby realizing the process design planning of the mouth cover components. Traditional recognition methods usually rely on manual judgment or simple keyword matching, which has problems of low efficiency and large recognition errors. As aircraft design becomes more and more complex, the scale of part three-dimensional models becomes larger and larger. There is an urgent need for an efficient and accurate automatic recognition method for thin-walled mouth frame parts.

[0003] For example, in the patent application "Aircraft Skin Part Feature Detection Method (Publication Date: 2015-03-25, Publication Number: CN104462656A)," various skin processing features were first defined. For each skin part, a ring characteristic map and holographic attribute surface-edge map were constructed, and seed faces of various features were searched. On this basis, a search was performed based on the reverse identification face, seed face, and expansion rules to find all the collective elements of each type of feature, construct the skin processing features, and obtain feature recognition results. However, this method requires predefined processing features and can only identify feature-level features such as milling processing features. It cannot effectively identify global-level geometric features such as thin-walled mouth frame parts.

[0004] Therefore, in order to solve the problems of low recognition efficiency and insufficient recognition accuracy in the zero recognition process of thin-walled mouth frames in the prior art, the present invention discloses a method for rapid recognition of aircraft thin-walled mouth frames. Summary of the Invention

[0005] The present invention discloses a method for quickly identifying aircraft thin-walled mouth frame parts. The method adopts image recognition and statistical methods to identify typical features such as the thickness of the part model and the consistency of the edge of the projection image to realize the recognition of the part digital model. The method can efficiently and accurately determine whether the part digital model is a thin-walled mouth frame part.

[0006] The present invention is achieved through the following technical solutions: A method for quickly identifying aircraft thin-walled mouth frame parts includes the following steps: Step 1: Create the OBB bounding box of the part digital model and extract the length, width and height of the OBB bounding box; Step 2: Extract the reference surface with the largest area on the part model, measure the thickness of the part along the normal direction of the reference surface, and determine whether the part is a thin-walled part. If it is a thin-walled part, proceed to step 3. Step 3: Project the part digital model according to the calibration direction, identify the closed-loop structure formed on the projection surface, and calculate the perimeter of all closed-loop structures; Step 4: Set a perimeter threshold, compare the perimeter of the closed-loop structure with the perimeter threshold, and screen the part digital model based on the comparison result; Step 5: Divide the closed-loop structure into an inner closed-loop structure and an outer closed-loop structure, calculate the ratio between the area of the inner closed-loop structure and the area of the outer closed-loop structure, and perform a secondary screening of the part digital model based on the ratio; Step 6: Calculate the shape consistency of the inner closed-loop structure and the outer closed-loop structure on the projection view, and screen the part digital model three times based on the consistency to obtain thin-walled mouth frame parts.

[0007] In order to better implement the present invention, further, in step 5, the calculation formula of the ratio is: ; in: It represents the ratio between the area of the inner closed loop structure and the area of the outer closed loop structure; represents the area of the i-th inner closed loop structure; represents the area of the outer closed-loop structure; k represents the number of inner closed-loop structures, 1≤i≤k.

[0008] In order to better implement the present invention, further, the step 6 specifically includes: Step 6.1: For the outer closed-loop structure, randomly select several outer closed-loop sampling points on the outer closed-loop structure, and calculate the minimum distance between each outer closed-loop sampling point and all inner closed-loop structures to obtain the minimum distance sampling set; Step 6.2: Calculate the normalized standard deviation of the minimum distance sampling set as the consistency; Step 6.3: Compare the normalized standard deviation with the consistency threshold, and screen the part digital model three times based on the comparison results to obtain thin-walled mouth frame parts.

[0009] In order to better implement the present invention, further, the formula for calculating the normalized standard deviation in step 6.2 is as follows: ; in: represents the normalized standard deviation; m represents the number of outer closed-loop sampling points; represents the minimum distance between the i-th outer closed-loop sampling point and all inner closed-loop structures; It represents the average value of the minimum distance between m outer closed-loop sampling points and all inner closed-loop structures.

[0010] In order to better implement the present invention, further, the step 2 specifically includes: Step 2.1, traverse all geometric faces on the part digital model and calculate the area of each geometric face; Step 2.2, extracting the geometric surface with the largest area as the reference surface, and discretizing the reference surface to obtain a number of reference geometric points on the reference surface; Step 2.3, extract the normal vector of the reference geometric point on the reference surface, and measure the thickness of the part at each reference geometric point along the normal vector direction; Step 2.4: Determine whether the thickness of the part at all reference geometric points is equal. If so, proceed to step 2.5. Step 2.5: Set a thickness threshold based on the size of the OBB bounding box, compare the part thickness with the thickness threshold, and determine whether the part is a thin-walled part.

[0011] In order to better implement the present invention, further, in step 2.5, the thickness of the part is compared with the thickness threshold as follows: like , then the part is judged to be a thin-walled part; if , then the part is judged to be a non-thin-wall part; in: represents the thickness of the part at the kth reference geometric point; Indicates the thickness threshold; max(R) indicates the maximum value of the length, width, and height of the OBB bounding box.

[0012] In order to better implement the present invention, further, in step 4, the perimeter of each closed-loop structure is calculated. If the perimeter of the closed-loop structure is less than a perimeter threshold, the current closed-loop structure is filtered out.

[0013] In order to better implement the present invention, further, the filtering and excluding steps of the closed-loop structure are repeated. If the number of closed-loop structures remaining is equal to 1, the current part digital model is directly judged to be a non-thin-walled mouth frame part; if the number of closed-loop structures remaining is greater than 1, then go to step 5.

[0014] In order to better implement the present invention, further, in step 1, an OBB bounding box algorithm is used to establish an OBB bounding box of the part digital model.

[0015] In order to better implement the present invention, further, in step 3, an edge detection algorithm is used to identify the closed loop structure formed on the projection surface.

[0016] Compared with the prior art, the present invention has the following advantages and beneficial effects: The present invention uses image recognition and statistical methods to identify typical features such as the thickness of the part model and the consistency of the edge of the projected image, and can efficiently and accurately identify the part digital model. It can effectively identify whether the part is a thin-walled frame part. Compared with existing recognition technologies, the present invention can improve recognition efficiency and reduce manual intervention. It is also suitable for rapid recognition and classification of large-scale parts libraries, which helps to improve aircraft knowledge management and process design levels. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of the process steps of the present invention; Figure 2 This is a schematic diagram of the first type of thin-walled mouth frame parts; Figure 3 This is a schematic diagram of the second type of thin-walled mouth frame parts; Figure 4 This is a schematic diagram of non-thin-walled mouth frame parts; Figure 5 Schematic diagram of the closed-loop structure. DETAILED DESCRIPTION

[0018] Example 1: A method for quickly identifying thin-walled frame parts of an aircraft according to this embodiment is as follows: Figure 1 As shown, the following steps are included: Step 1: Create the OBB bounding box of the part digital model and extract the length, width and height of the OBB bounding box; Step 2: Extract the reference surface with the largest area on the part model, measure the thickness of the part along the normal direction of the reference surface, and determine whether the part is a thin-walled part. If it is a thin-walled part, proceed to step 3. Step 3: Project the part digital model according to the calibration direction, identify the closed-loop structure formed on the projection surface, and calculate the perimeter of all closed-loop structures; Step 4: Set a perimeter threshold, compare the perimeter of the closed-loop structure with the perimeter threshold, and screen the part digital model based on the comparison result; Step 5: Divide the closed-loop structure into an inner closed-loop structure and an outer closed-loop structure, calculate the ratio between the area of the inner closed-loop structure and the area of the outer closed-loop structure, and perform a secondary screening of the part digital model based on the ratio; Step 6: Calculate the shape consistency of the inner closed-loop structure and the outer closed-loop structure on the projection view, and screen the part digital model three times based on the consistency to obtain thin-walled mouth frame parts.

[0019] In the step 1, an OBB bounding box algorithm is used to establish an OBB bounding box of the part digital model. In the step 3, an edge detection algorithm is used to identify a closed-loop structure formed on the projection surface.

[0020] The step 2 specifically includes: Step 2.1, traverse all geometric faces on the part digital model and calculate the area of each geometric face; Step 2.2, extracting the geometric surface with the largest area as the reference surface, and discretizing the reference surface to obtain a number of reference geometric points on the reference surface; Step 2.3, extract the normal vector of the reference geometric point on the reference surface, and measure the thickness of the part at each reference geometric point along the normal vector direction; Step 2.4: Determine whether the thickness of the part at all reference geometric points is equal. If so, proceed to step 2.5. Step 2.5: Set a thickness threshold based on the size of the OBB bounding box, compare the part thickness with the thickness threshold, and determine whether the part is a thin-walled part.

[0021] In step 2.5, the part thickness is compared with the thickness threshold as follows: like , then the part is judged to be a thin-walled part; if , then the part is judged to be a non-thin-wall part; in: represents the thickness of the part at the kth reference geometric point; Indicates the thickness threshold; max(R) indicates the maximum value of the length, width, and height of the OBB bounding box.

[0022] In step 4, the perimeter of each closed-loop structure is calculated. If the perimeter of the closed-loop structure is less than the perimeter threshold, the current closed-loop structure is filtered out. The closed-loop structure filtering and elimination steps are repeated. If the number of closed-loop structures remaining is equal to 1, the current part model is directly determined to be a non-thin-walled mouth frame part. If the number of closed-loop structures remaining is greater than 1, the process proceeds to step 5.

[0023] In step 5, the calculation formula of the ratio is: ; in: It represents the ratio between the area of the inner closed loop structure and the area of the outer closed loop structure; represents the area of the i-th inner closed loop structure; represents the area of the outer closed-loop structure; k represents the number of inner closed-loop structures, 1≤i≤k.

[0024] The step 6 specifically includes: Step 6.1: For the outer closed-loop structure, randomly select several outer closed-loop sampling points on the outer closed-loop structure, and calculate the minimum distance between each outer closed-loop sampling point and all inner closed-loop structures to obtain the minimum distance sampling set; Step 6.2: Calculate the normalized standard deviation of the minimum distance sampling set as the consistency; Step 6.3: Compare the normalized standard deviation with the consistency threshold, and screen the part digital model three times based on the comparison results to obtain thin-walled mouth frame parts.

[0025] The formula for calculating the normalized standard deviation in step 6.2 is as follows: ; in: represents the normalized standard deviation; m represents the number of outer closed-loop sampling points; represents the minimum distance between the i-th outer closed-loop sampling point and all inner closed-loop structures; It represents the average value of the minimum distance between m outer closed-loop sampling points and all inner closed-loop structures.

[0026] Example 2: This embodiment discloses a method for quickly identifying thin-walled aircraft frame parts, which is further optimized based on the first embodiment, specifically: Given a part model to be identified, calculate the OBB bounding box of the part model ,in 、 、 are the length, width and height of the bounding box respectively, in mm; Figure 2 As an example, the OBB bounding box size is ; Traverse all geometric faces of the part model , j represents the geometric face number. Loop and calculate each geometric face Area , traverse the areas of all the calculated geometric surfaces and find the geometric surface with the largest area as the reference surface , for the reference surface Discretize according to the set accuracy to obtain several reference geometric points on the reference surface , k represents the serial number of the reference geometry point. Get the reference geometry point On the reference surface The normal vector on the part is specified to point to the interior of the part model 3D entity, and the part thickness t at each reference geometric point in the direction of the normal vector is calculated. k , t k Represents the part thickness along the normal direction at the kth reference geometry point.

[0027] Setting the thickness threshold , first determine several reference geometric points The corresponding part thickness t kAre they equal? If the thickness of the parts at all reference geometric points is equal, then: like , then the current part model is judged to be a non-thin-wall part; if , then the current part model is judged to be a thin-walled part. Figure 2 For example, the thickness of the part at each reference geometric point is 1.2 mm, and , then judge Figure 2 The parts shown in the figure are thin-walled parts.

[0028] Project the part model along the main direction of the OBB bounding box and identify all closed loop structures formed by the edge of the part model on the projection surface. The outermost ring of the closed loop structure is regarded as the outer closed loop structure, and the ring contained within the outer closed loop structure is regarded as the inner closed loop structure. Usually, there is only one outer closed loop structure, and there is at least one inner closed loop structure inside the outer closed loop structure. Calculate the perimeter of the outer closed loop structure and the inner closed loop structure. Figure 3 Taking the part shown in as an example, all closed-loop structures identified in the projection image are as follows Figure 5 As shown, it includes 1 outer closed loop structure and 3 inner closed loop structures.

[0029] Set the perimeter threshold c = 10mm and calculate the perimeter of each closed-loop structure. If the perimeter of a closed-loop structure is less than the perimeter threshold, filter the closed-loop structure. If the number of closed-loop structures after filtering is equal to 1, directly determine that the part digital model is a non-thin-walled mouth frame part. Otherwise, proceed to the next step.

[0030] Calculate the ratio of the sum of the areas of all inner closed-loop structures of the part model to the area of the outer closed-loop structure , The smaller it is, the larger the hollow area of the part. The calculation method is as follows: ; in, It represents the ratio between the area of the inner closed loop structure and the area of the outer closed loop structure; represents the area of the i-th inner closed loop structure; represents the area of the outer closed-loop structure; k represents the number of inner closed-loop structures, 1≤i≤k.

[0031] The shape consistency of the outer closed-loop structure and the inner closed-loop structure on the part projection view is calculated using a statistical method, which specifically includes the following steps: For the outer closed-loop structure, 100 outer closed-loop sampling points are randomly selected from the outer closed-loop structure, and the minimum distance between each outer closed-loop sampling point and all inner closed-loop structures is calculated to obtain the minimum distance sampling set. , It represents the minimum distance between the i-th outer closed-loop sampling point and all inner closed-loop structures, 1≤i≤100.

[0032] Calculate the normalized standard deviation of the minimum distance sampling set As consistency, The smaller it is, the more consistent the shapes of the outer closed-loop structure and the inner closed-loop structure of the part are, and vice versa. The calculation method is as follows: , where m=100; in: represents the normalized standard deviation; m represents the number of outer closed-loop sampling points; represents the minimum distance between the i-th outer closed-loop sampling point and all inner closed-loop structures; It represents the average value of the minimum distance between m outer closed-loop sampling points and all inner closed-loop structures.

[0033] Set ratio threshold and consistency threshold , if satisfied ,and , then the part is considered to be a thin-walled frame part. Figure 2-Figure 4 The parts shown in They are 0.76, 0.81, and 0.15 respectively. They are 0.09, 0.01, and 0.42 respectively, so we can judge Figure 2 and Figure 3 For thin-walled frame parts, Figure 4 These are non-thin-walled mouth frame parts.

[0034] This embodiment shows that the automatic identification method for thin-walled frame parts proposed in the present invention can be used to quickly identify flat plate parts in a three-dimensional digital model of an aircraft and can achieve good results.

[0035] The rest of this embodiment is the same as that of embodiment 1, so it will not be described again.

[0036] 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 scope of protection of the present invention.

Claims

1. A method for quickly identifying aircraft thin-walled frame parts, characterized in that: The following steps are involved: Step 1: Create the OBB bounding box of the part digital model and extract the length, width and height of the OBB bounding box; Step 2: Extract the reference surface with the largest area on the part model, measure the thickness of the part along the normal direction of the reference surface, and determine whether the part is a thin-walled part. If it is a thin-walled part, proceed to step 3. Step 3: Project the part digital model according to the calibration direction, identify the closed-loop structure formed on the projection surface, and calculate the perimeter of all closed-loop structures; Step 4: Set a perimeter threshold, compare the perimeter of the closed-loop structure with the perimeter threshold, and screen the part digital model based on the comparison result; Step 5: Divide the closed-loop structure into an inner closed-loop structure and an outer closed-loop structure, calculate the ratio between the area of the inner closed-loop structure and the area of the outer closed-loop structure, and perform a secondary screening of the part digital model based on the ratio; Step 6: Calculate the shape consistency of the inner closed-loop structure and the outer closed-loop structure on the projection view, and screen the part digital model three times based on the consistency to obtain thin-walled mouth frame parts.

2. A method for quickly identifying aircraft thin-walled mouth frame parts according to claim 1, characterized in that: In step 5, the calculation formula of the ratio is: ； in: It represents the ratio between the area of the inner closed loop structure and the area of the outer closed loop structure; represents the area of the i-th inner closed loop structure; represents the area of the outer closed-loop structure; k represents the number of inner closed-loop structures, 1≤i≤k.

3. A method for quickly identifying aircraft thin-walled mouth frame parts according to claim 2, characterized in that: The step 6 specifically includes: Step 6.1: For the outer closed-loop structure, randomly select several outer closed-loop sampling points on the outer closed-loop structure, and calculate the minimum distance between each outer closed-loop sampling point and all inner closed-loop structures to obtain the minimum distance sampling set; Step 6.2: Calculate the normalized standard deviation of the minimum distance sampling set as the consistency; Step 6.3: Compare the normalized standard deviation with the consistency threshold, and screen the part digital model three times based on the comparison results to obtain thin-walled mouth frame parts.

4. A method for quickly identifying aircraft thin-walled mouth frame parts according to claim 3, characterized in that: The formula for calculating the normalized standard deviation in step 6.2 is as follows: ； in: represents the normalized standard deviation; m represents the number of outer closed-loop sampling points; represents the minimum distance between the i-th outer closed-loop sampling point and all inner closed-loop structures; It represents the average value of the minimum distance between m outer closed-loop sampling points and all inner closed-loop structures.

5. A method for quickly identifying aircraft thin-walled mouth frame parts according to any one of claims 1 to 4, characterized in that: The step 2 specifically includes: Step 2.1, traverse all geometric faces on the part digital model and calculate the area of each geometric face; Step 2.2, extracting the geometric surface with the largest area as the reference surface, and discretizing the reference surface to obtain a number of reference geometric points on the reference surface; Step 2.3, extract the normal vector of the reference geometric point on the reference surface, and measure the thickness of the part at each reference geometric point along the normal vector direction; Step 2.4: Determine whether the thickness of the part at all reference geometric points is equal. If so, proceed to step 2.

5. Step 2.5: Set a thickness threshold based on the size of the OBB bounding box, compare the part thickness with the thickness threshold, and determine whether the part is a thin-walled part.

6. A method for quickly identifying aircraft thin-walled mouth frame parts according to claim 5, characterized in that: In step 2.5, the part thickness is compared with the thickness threshold as follows: like , then the part is judged to be a thin-walled part; if , then the part is judged to be a non-thin-wall part; in: represents the thickness of the part at the kth reference geometric point; Indicates the thickness threshold; max(R) indicates the maximum value of the length, width, and height of the OBB bounding box.

7. A method for quickly identifying aircraft thin-walled mouth frame parts according to any one of claims 1 to 4, characterized in that: In step 4, the perimeter of each closed-loop structure is calculated. If the perimeter of the closed-loop structure is less than a perimeter threshold, the current closed-loop structure is filtered out.

8. A method for quickly identifying aircraft thin-walled mouth frame parts according to claim 7, characterized in that: Repeat the filtering and exclusion steps for closed-loop structures. If the number of closed-loop structures remaining is equal to 1, directly determine that the current part model is a non-thin-walled mouth frame part; if the number of closed-loop structures remaining is greater than 1, proceed to step 5.

9. A method for quickly identifying aircraft thin-walled mouth frame parts according to any one of claims 1 to 4, characterized in that: In step 1, an OBB bounding box algorithm is used to establish an OBB bounding box of the part digital model.

10. A method for quickly identifying aircraft thin-walled mouth frame parts according to any one of claims 1 to 4, characterized in that: In step 3, an edge detection algorithm is used to identify the closed-loop structure formed on the projection surface.

Images