A frequency-selective radome modeling method based on Catia point cloud technology

Abstract

This invention relates to a frequency-selective radome modeling method based on Catia point cloud technology, comprising the following steps: S1, importing the radome's 3D model and dividing it into multiple surfaces; S2, converting the multiple surfaces of the radome into an initial point cloud dataset; S3, downsampling and optimizing the initial point cloud dataset according to the periodic size and array distribution of the frequency-selective elements to obtain a filtered point cloud dataset; S4, obtaining the normal vectors of the filtered point cloud dataset; S5, drawing the frequency-selective element structure on the tangent plane of each normal vector in each filtered point cloud normal vector dataset; S6, projecting or overlaying the drawn frequency-selective element structure onto the radome surface to complete the conformal modeling design of the frequency-selective radome. This invention can be used for rapid modeling of rotationally symmetric and non-rotationally symmetric frequency-selective radomes, greatly improving the accuracy and speed of frequency-selective radome modeling.

Description

Technical Field

[0001] This invention relates to a frequency-selective radome modeling method based on Catia point cloud technology, belonging to the field of radome technology. Background Technology

[0002] With the increasing demands for stealth warfare, aircraft require not only frequency-selective radomes for their radar antennas at the front, but also for radomes at the leading edge of wings, wingtips, and vertical tail to meet different functional requirements. These radomes protect the communication modules within the radome from harsh environments such as wind, rain, high temperatures, humidity, and sandstorms, without affecting normal communication functions. Therefore, frequency-selective radomes are key components determining the combat performance of stealth aircraft. Due to the special nature of their surface features, the local shape is often complex, typically exhibiting significant curvature variations and being non-linear, non-deployable irregular surfaces. This poses significant challenges to modeling the radomes in these locations. Traditional frequency-selective radome modeling methods are mostly applicable to deployable curved surface structures, resulting in slow modeling speeds, low efficiency, and a tendency to produce errors such as overlapping frequency-selective structures in the top and edge areas of the radome. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings and defects of the existing technology and provide a frequency-selective radome modeling method based on Catia point cloud technology, which specifically includes the following steps:

[0004] S1. Use the computer software Catia to create a three-dimensional model of the radome in STL format. Divide the three-dimensional model of the radome into M surfaces, where the i-th surface is denoted as FI, and the range of i is from 1 to M.

[0005] S2. Use the computer software Cloud Compare to read the M surfaces of the radome in STL format and convert them into M initial point cloud datasets respectively. The initial point cloud dataset of the i-th surface FI is DI, which contains dense points.

[0006] S3. Based on the period size and array distribution of the frequency selection unit, downsample and optimize the M initial point cloud datasets respectively, delete unnecessary point clouds and point clouds with irregular smooth density, and obtain M filtered point cloud datasets. The i-th initial point cloud dataset DI is optimized to obtain the filtered point cloud dataset DI', which contains several sparse points.

[0007] S4. Calculate the normal vector of each point in the M filtered point cloud datasets respectively, and form the corresponding normal vector dataset. For the i-th filtered point cloud dataset DI', obtain the normal vector of each point in DI', and denote it as the normal vector dataset VI of DI'.

[0008] S5. Perform the following operations on all M normal vector datasets: draw the tangent plane of each normal vector in the i-th normal vector dataset VI, and draw the frequency-selective unit topology on each tangent plane; S6. Import the radome 3D model into Catia, and project the drawn frequency-selective units onto the surface of the radome 3D model to complete the conformal modeling of the frequency-selective radome.

[0009] Preferably, the 3D model of the radome established by the Catia software in S1 is a rotationally symmetric structure or a non-rotationally symmetric structure; it is a developable surface or a non-developable surface; the 3D model is divided into multiple surfaces, and the radius of curvature of each surface is smaller than the spatial straight-line distance Q between the geometric centers of adjacent frequency-selective units.

[0010] Preferably, the frequency-selective array in S3 can be distributed in a triangular, square, or rhomboid pattern.

[0011] Specifically, the S3 operation of filtering the DI of the i-th initial point cloud dataset includes:

[0012] S3.1: On surface F i A point PI0(x0,y0,z0) is randomly selected from the initial point cloud dataset DI as the initial vertex; PI0 is included in the filtered point cloud dataset DI'; all points in the initial point cloud dataset DI whose distance from PI0 is less than Q-Δd are deleted; where Δd is the spacing error.

[0013] S3.2: Randomly select a point PI1(x1,y1,z1) in the initial point cloud dataset DI that is more than Q-Δd and less than Q+Δd from the initial point PI0, and use it as the second vertex. Include PI1 in the filtered point cloud dataset DI' and delete all points in the initial point cloud dataset DI that are less than Q-Δd from PI1.

[0014] S3.3: Using PI1 as the vertex, construct spheres with radii of Q-Δd and Q+Δd respectively. Based on the frequency-selective array distribution, divide the initial point cloud dataset DI into N equal parts, where N can be any value from 4, 6, or 8. Randomly select a point in the region where the dividing line intersects with the annulus with radii from Q-Δd to Q+Δd as the third vertex PI2(x2,y2,z2). PI2 is included in the filtered point cloud dataset DI'. Delete all points in the initial point cloud dataset DI whose distance from PI2 is less than Q-Δd.

[0015] S3.4: Using PI2(x2,y2,z2) as the vertex, execute S3.3 again to obtain the fourth vertex PI3(x3,y3,z3). Add PI3(x3,y3,z3) to the filtered point cloud dataset DI', and delete all points in the initial point cloud dataset DI whose distance from PI3 is less than Q-Δd. Continue in this manner to obtain subsequent points in the filtered point cloud dataset DI' until only one point remains in the initial point cloud dataset DI.

[0016] Furthermore, in S3.3 and S3.4, the (j+2)th vertex PI j+2 (x j+2 ,y j+2 ,z j+2 ), where j is an integer greater than or equal to 0, and must satisfy: ①PI j+1 PI j+2 The distance is greater than Q-Δd and less than Q+Δd; ②PI j+1 PI j Two-point line and PI j+1 PI +2 The included angle between the lines connecting the two points satisfies a condition greater than 360° / N-ΔΦ and less than 360° / N+ΔΦ; where ΔΦ is the angular error.

[0017] Preferably, in step S3.3, if the frequency selection array is distributed in an equilateral triangle, the initial point cloud dataset is divided into 6 equal parts; if the frequency selection array is distributed in a square, the initial point cloud dataset is divided into 4 equal parts; and if the frequency selection array is distributed in a rhombus, the initial point cloud dataset is divided into 8 equal parts.

[0018] Specifically, the operation of S4 on the i-th filtered point cloud dataset DI' includes:

[0019] Import the filtered point cloud dataset DI' into the Catia software to obtain the PI of each point. j The tangent plane SI at the location j and the corresponding outer normal LI j SI j and LI j They are perpendicular to each other, with their midpoint PI j (x j ,y j ,z j The corresponding normal vector VI j It can be calculated using the following formula:

[0020]

[0021] Where F(x) j ,y j ,z j ) is point PI j (xj ,y j ,z j Local surface equations, filtering point cloud dataset DI' corresponding normal vector dataset VI from point PI j Normal vector VI at point j Composition; Point PI j Corresponding outer normal LI j Any point (x, y, z) on (x, y, z) satisfies the following equation:

[0022]

[0023] Where point PI j Corresponding tangent plane SI j Any point (x, y, z) on (x, y, z) satisfies the following equation:

[0024]

[0025] Specifically, the operation of S5 on the j-th point in the i-th filtered point cloud dataset DI' is as follows: using point PI in the filtered point cloud dataset DI' j Let LI be the origin of the coordinate system. j The normal is in the tangent plane SI. j A local coordinate system is established; the frequency selection unit is transformed using the Catia software to copy the frequency selection unit from the initial coordinate system to all local coordinate systems, and to ensure that the spatial orientation of the frequency selection unit is consistent in all local coordinate systems.

[0026] Specifically, the operation of S6 on the j-th point in the i-th filtered point cloud dataset DI' is as follows: in the Catia software, the topology of the j-th frequency selection unit is aligned with the corresponding normal vector VI. j The parallel projection onto the surface of the radome 3D model yields the frequency-selective conformal radome 3D model.

[0027] In summary, compared with the prior art, the present invention provides a fast, accurate, and efficient modeling method for frequency-selective radomes with deployable / non-deployable curved surfaces. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the modeling method for frequency-selective radomes based on Catia point cloud technology according to the present invention.

[0029] Figure 2 This is a schematic diagram of a three-dimensional model of the frequency-selective radome based on Catia point cloud technology of the present invention.

[0030] Figure 3 This is a schematic diagram of the initial point cloud dataset for the frequency-selective radome based on Catia point cloud technology according to the present invention.

[0031] Figure 4 This is a schematic diagram of the filtered point cloud dataset of the frequency-selective radome based on Catia point cloud technology according to the present invention.

[0032] Figure 5 This is a schematic diagram of the frequency selection unit of the frequency-selective radome based on Catia point cloud technology of the present invention;

[0033] Figure 6 This is a schematic diagram of the frequency-selective radome array distribution based on Catia point cloud technology according to the present invention;

[0034] Figure 7 This is a schematic diagram of the frequency-selective radome point cloud screening method based on Catia point cloud technology of the present invention;

[0035] Figure 8 This is a schematic diagram of the modeling results of the frequency-selective radome based on Catia point cloud technology according to the present invention. Detailed Implementation

[0036] The following will be combined with the appendix in the embodiments of the present invention. Figure 1 ~Attached Figure 8 The specific operation process of the frequency-selective radome modeling method based on Catia point cloud technology of the present invention will be described in detail in the embodiments.

[0037] It should be noted that the accompanying drawings are in a very simplified form and use non-precise proportions. They are only used to facilitate and clarify the purpose of illustrating the embodiments of the present invention, and are not intended to limit the implementation conditions of the present invention. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportional relationship, or adjustments to the size should still fall within the scope of the technical content disclosed in the present invention, provided that they do not affect the effects and objectives that the present invention can produce.

[0038] It should be noted that, in this invention, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only the expressly listed elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0039] Figure 1 This is a flowchart illustrating the steps of a frequency-selective radome modeling method based on Catia point cloud technology provided by this invention. Figure 1 As shown, the present invention includes the following steps:

[0040] S1: Use the computer software Catia to create a three-dimensional model of the radome in STL format. Divide the three-dimensional model of the radome into M surfaces, where the i-th surface is denoted as FI, and the range of i is from 1 to M.

[0041] S2: Use the computer software Cloud Compare to read the M surfaces of the radome in STL format and convert them into M initial point cloud datasets respectively. The initial point cloud dataset of the i-th surface FI is DI, which contains dense points.

[0042] S3: Based on the period size and array distribution of the frequency selection unit, downsample and optimize the M initial point cloud datasets respectively, delete unnecessary point clouds and point clouds with irregular smooth density, and obtain M filtered point cloud datasets. The i-th initial point cloud dataset DI is optimized to obtain the filtered point cloud dataset DI', which contains several sparse points.

[0043] S4: Calculate the normal vector of each point in the M filtered point cloud datasets respectively, and form the corresponding normal vector dataset. For the i-th filtered point cloud dataset DI', obtain the normal vector of each point in DI', and denote it as the normal vector dataset VI of DI'.

[0044] S5: Draw a tangent plane for each normal vector in each of the M normal vector datasets, and draw the frequency-selective unit topology on each normal vector tangent plane.

[0045] S6: Import the 3D model of the radome into Catia, and project the drawn frequency-selective unit topology onto the surface of the 3D model of the radome to complete the conformal modeling of the frequency-selective radome.

[0046] like Figure 2 As shown, in step S1 of this embodiment, a three-dimensional model of a rotationally symmetric radome is established using Catia software; and the three-dimensional model of the radome is divided into M surfaces F1, F2, F3..., FM.

[0047] Furthermore, the 3D model of the radome created by the Catia software can be a rotationally symmetric structure or a non-rotationally symmetric structure; it can be a developable surface or a non-developable surface.

[0048] Specifically, the 3D model of the radome is divided into multiple curved surfaces, and the radius of curvature of each surface is smaller than the spatial straight-line distance Q between the geometric centers of adjacent frequency selective units. Q is also the periodic dimension of the frequency selective unit, which is 10mm in this embodiment.

[0049] In the detailed description of each step below, only surface F2 will be used as an example. Those skilled in the art will understand that the same method will be used to process other surfaces.

[0050] like Figure 3 As shown, in this embodiment S2, the Cloud Compare software is used to read the surface F2 in STL format, and the corresponding initial point cloud dataset D2 is obtained by conversion.

[0051] In step S3 of this embodiment, as follows Figure 5 The diagram shows a schematic of the frequency-selective unit. The material used is copper foil with a thickness of 0.017 mm. The Jerusalem cross structure is equivalent to an inductor, and the metal square ring is equivalent to a capacitor, forming an LC resonant circuit. The topology of the frequency-selective unit is optimized according to the required transmission frequency band of the radome. Figure 6 As shown, the frequency-selective unit array is a square array distribution. The initial point cloud dataset D2 is downsampled and optimized according to the following steps to obtain the following result. Figure 4 The filtered point cloud dataset D2' is shown.

[0052] S3.1: Randomly select a point P20(x0,y0,z0) in the initial point cloud dataset D2 of surface F2 as the initial vertex, and include it in the filtered point cloud dataset D2'. Delete all points in the initial point cloud dataset D2 that are less than Q-Δd from P20, where Δd is an acceptable spacing error. In this embodiment, Δd is taken as 0.5mm.

[0053] S3.2: In the initial point cloud dataset D2, select a point P21(x1,y1,z1) whose distance from the initial point P20 is greater than Q-Δd and less than Q+Δd as the second vertex, and include it in the filtered point cloud dataset D2'. Delete all points in the initial point cloud dataset D2 whose distance from P21 is less than Q-Δd.

[0054] S3.3: Using P21 as the vertex, construct spheres with radii of Q-Δd and Q+Δd respectively. Based on the frequency-selective unit structure and its square array distribution, divide the initial point cloud dataset D2 into four equal parts at 90°. Randomly select a point in the region where the dividing line intersects with the annulus with radii from Q-Δd to Q+Δd as the third vertex P22(x2,y2,z2). Include P22 in the filtered point cloud dataset D2'. Delete all points in the initial point cloud dataset D2 whose distance from P22 is less than Q-Δd. The third vertex P22(x2,y2,z2) must satisfy: ① The distance between P21 and P22 is greater than Q-Δd and less than Q+Δd; ② The angle between the line connecting P21 and P20 and the line connecting P21 and P22 is greater than 90°-ΔΦ and less than 90°+ΔΦ. Here, ΔΦ is an acceptable angle error, which is 5° in this embodiment.

[0055] S3.4: Using P22 as the vertex, execute S3.3 again to obtain the fourth vertex P23(x3,y3,z3). Add P23(x3,y3,z3) to the filtered point cloud dataset D2', and delete all points in the initial point cloud dataset D2 whose distance from P23 is less than Q-Δd. The fourth vertex P23(x2,y2,z2) must satisfy the following: ① The distance between P22 and P23 is greater than Q-Δd and less than Q+Δd; ② The angle between the line connecting P22 and P21 and the line connecting P22 and P23 is greater than 90°-ΔΦ and less than 90°+ΔΦ. Continue in this manner to obtain subsequent points in the filtered point cloud dataset D2' until only one point remains in the initial point cloud dataset D2.

[0056] Furthermore, in S3.3 and S3.4, the (j+2)th vertex P2 j+2 (x j+2 ,y j+2 ,z j+2 ), where j is an integer greater than or equal to 0, ①P2 j+1 P2 j+2 The distance is greater than Q-Δd and less than Q+Δd; ②P2 j+1 P2 j The line connecting the two points and P2 j+1 P2 j+2 The included angle between the lines connecting the two points satisfies a value greater than 360° / N-ΔΦ and less than 360° / N+ΔΦ; where ΔΦ is the angle error, which is taken as 5° in this embodiment.

[0057] Furthermore, such as Figure 6 and Figure 7As shown, the array distribution of the frequency selection unit can be a square array, a triangular array, or a rhombus array. If it is an equilateral triangular distribution, the initial point cloud dataset is divided into 6 equal parts at 60°; if it is a rhombus distribution, the initial point cloud dataset is divided into 8 equal parts at 45°.

[0058] In step S4 of this embodiment, the filtered point cloud dataset D2' is imported into the Catia software to obtain the value of each point P2. j Tangent plane S2 at the location j and the corresponding outer normal L2 j S2 j and L2 j They are perpendicular to each other, with point P2 as the midpoint. j (x j ,y j ,z j The corresponding normal vector V2 j It can be calculated using the following formula:

[0059]

[0060] Where F(x) j ,y j ,z j Point P2 is a point on the P2 line. j (x j ,y j ,z j The local surface equation, the filtered point cloud dataset D2' corresponding to the normal vector dataset V2 is derived from the normal vector V2. j Composition; Point P2 j Corresponding outward normal L2 j Any point (x, y, z) on (x, y, z) satisfies the following equation:

[0061]

[0062] Point P2 j The corresponding tangent plane S2 j Any point (x, y, z) on (x, y, z) satisfies the following equation:

[0063]

[0064] In step S5 of this embodiment, each point P2 in the filtered point cloud dataset D2' is used as an example. j Using L2 as the origin of the coordinate system j As the normal, in the tangent plane S2 jEstablish local coordinate systems; use Catia software to perform coordinate transformation on the frequency selection unit, so that the frequency selection unit is copied from the initial coordinate system to all local coordinate systems, and ensure that the spatial orientation of the frequency selection unit in all local coordinate systems is consistent.

[0065] like Figure 8 As shown, in step S6 of this embodiment, the topology of the j-th frequency selection unit is aligned with the corresponding normal vector V2 in the Catia software. j The parallel projection onto the surface of the radome 3D model yields the frequency-selective conformal radome 3D model.

[0066] In summary, compared with existing frequency-selective radome modeling methods, the frequency-selective radome modeling method based on Catia point cloud technology provided by this invention can quickly, accurately, and efficiently model frequency-selective radomes with deployable / non-deployable curved surfaces.

[0067] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. After reading the above, those skilled in the art will readily recognize various modifications and substitutions to the present invention. For example, the present invention, being a rotationally symmetric radome, can also be applied to non-rotationally symmetric radomes; the present invention, being a single-layer frequency-selective structure, can also be applied to multi-layer frequency-selective structures. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A method for modeling frequency-selective radomes based on Catia point cloud technology, characterized in that, Includes the following steps: S1. Use the computer software Catia to create a 3D model of the radome. Divide the 3D model of the radome into M surfaces, where the i-th surface is represented as... FI The range of i is from 1 to M, and the radius of curvature of each surface is smaller than the spatial straight-line distance Q between the geometric centers of adjacent frequency-selective units; S2. Read the M surfaces of the radome and convert them into M initial point cloud datasets, where the i-th surface... FI The initial point cloud dataset is DI ; S3. Based on the period size and array distribution of the frequency selection unit, downsample and optimize the M initial point cloud datasets respectively to obtain M filtered point cloud datasets, where the i-th initial point cloud dataset... DI The optimized filtered point cloud dataset is obtained. DI' ; S4. Calculate the normal vector for each point in each of the M filtered point cloud datasets, forming M corresponding normal vector datasets. For the i-th filtered point cloud dataset... DI' , obtain DI' The normal vector of each point in the vector is denoted as . DI' Normal vector dataset VI ; S5. Perform the following operation on all M normal vector datasets: plot the i-th normal vector dataset. VI Find the tangent plane for each normal vector and draw the frequency-selective unit topology on each tangent plane; S6. Import the radome 3D model into Catia, and project the drawn frequency-selective unit topology onto the surface of the radome 3D model to complete the conformal modeling of the frequency-selective radome. Specifically, S3 includes: S3.1: On the surface FI Initial point cloud dataset DI Randomly select a point PI 0 (x 0 ,y 0 ,z 0 ) As the initial vertex; PI 0 Point cloud dataset after filtering DI' ; Delete the initial point cloud dataset DI and PI 0 Distance less than Q-Δ d All points; where Δ d This refers to the spacing error; S3.2: In the initial point cloud dataset DI Randomly select one of the initial vertices. PI 0 Spacing greater than Q-Δ d And less than Q+Δ d point PI 1 (x 1 ,y 1 ,z 1 ) As the second vertex, PI 1 Point cloud dataset after filtering DI' Delete the initial point cloud dataset DI Zhongyu PI 1 Distance less than Q-Δ d All points; S3.3: with PI 1 Let Q be the vertex, and let Q-Δ be the radius. d and Q+Δ d The sphere, based on the frequency-selective array distribution method, is used for the initial point cloud dataset. DI Divide the material into N equal parts, where N can be any value from 4, 6, or 8; randomly select the dividing lines and the radius Q-Δ. d To Q+Δ d A point in the intersecting region of the annulus is taken as the third vertex. PI 2 (x 2 ,y 2 ,z 2 ) , PI 2 Point cloud dataset after filtering DI' ; Delete the initial point cloud dataset DI Zhongyu PI 2 Distance less than Q-Δ d All points; S3.4: with PI 2 For the first vertex, execute S3.3 again to obtain the fourth vertex. PI 3 (x 3 ,y 3 ,z 3 ) ,Will PI 3 (x 3 ,y 3 ,z 3 ) Add to filtered point cloud dataset DI' Delete the initial point cloud dataset DI Zhongyu PI 3 Distance less than Q-Δ d All points; and so on to obtain the filtered point cloud dataset. DI' Subsequent points, up to the initial point cloud dataset. DI Only a little bit remains in the middle; In S3.3 and S3.4, the (j+2)th vertex PI j+2 (x j+2 ,y j+2 ,z j+2 ) j is an integer greater than or equal to 0, and the following conditions must be met: ① PI j+1 , PI j+2 The distance is greater than Q-Δ d And less than Q+Δ d ;② PI j+1 , PI j Two points connected by a line and PI j+1 , PI j+2 The angle between the lines connecting the two points must be greater than 360° / N-Δ Φ And less than 360° / N +Δ Φ ; where Δ Φ This is for angular error; The S4 method filters the i-th point cloud dataset. DI' The specific operation is as follows: The filtered point cloud dataset DI' Import the data into Catia software and obtain the values ​​for each point. PI j Tangent plane at the location SI j and the corresponding outer normal LI j , midpoint corresponding normal vector VI j It can be calculated using the following formula: ； in It is a point Local surface equations, filtering point cloud datasets DI' Corresponding normal vector dataset VI From point Normal vector at point VI j Composition; Point Corresponding outer normal any point on Satisfy the following formula: ； midpoint Corresponding tangent plane any point on Satisfy the following formula: 。 2. The frequency-selective radome modeling method based on Catia point cloud technology as described in claim 1, characterized in that, The three-dimensional model of the radome established in S1 is either a rotationally symmetric structure or a non-rotationally symmetric structure; it is either a developable surface or a non-developable surface.

3. The frequency-selective radome modeling method based on Catia point cloud technology as described in claim 1, characterized in that, The frequency-selective unit array in S3 is distributed in a triangular, square, or rhomboid pattern.

4. The frequency-selective radome modeling method based on Catia point cloud technology as described in claim 3, characterized in that, In S3.3, if the frequency selection array is distributed in an equilateral triangle, the initial point cloud dataset is divided into 6 equal parts; if the frequency selection array is distributed in a square, the initial point cloud dataset is divided into 4 equal parts; and if the frequency selection array is distributed in a rhombus, the initial point cloud dataset is divided into 8 equal parts.

5. The frequency-selective radome modeling method based on Catia point cloud technology as described in claim 1, characterized in that, The S5 method filters the i-th point cloud dataset. DI' The specific operation on the j-th point is: using the filtered point cloud dataset... DI' Points in With the origin of the coordinate system, As the normal, in the tangent plane A local coordinate system is established; the frequency selection unit is transformed using the Catia software to copy the frequency selection unit from the initial coordinate system to all local coordinate systems, and to ensure that the spatial orientation of the frequency selection unit is consistent in all local coordinate systems.

6. The frequency-selective radome modeling method based on Catia point cloud technology as described in claim 5, characterized in that, The S6 method filters the i-th point cloud dataset. DI' The specific operation at the j-th point is as follows: In the Catia software, the topology of the j-th frequency selection unit is aligned along the corresponding normal vector. VI j The parallel projection onto the surface of the radome 3D model yields the frequency-selective conformal radome 3D model.

Images