A fast prediction method of target acoustic scattering

Abstract

This invention discloses a rapid prediction method for acoustic scattering of targets, relating to the field of active sonar detection technology for underwater targets. This invention significantly reduces the computation time for occlusion and ray tracing in the algorithm through the "medicine box algorithm"; it can predict multiple sets of boundary conditions for a target in a single calculation; and through a self-developed adaptive parallel technology, it can perfectly utilize all CPU cores when calculating multiple field points. Ultimately, this invention can shorten the calculation time for a single field point to within 4 seconds and increase the calculation speed for multiple field points by approximately four orders of magnitude, significantly improving the calculation speed for acoustic scattering prediction of underwater targets and meeting the requirements of real-time detection and analysis in engineering.

Description

Technical Field

[0001] This invention relates to the field of active sonar detection technology for underwater targets, and in particular to a method for rapid prediction of target acoustic scattering. Background Technology

[0002] Currently, the calculation of mid-to-high frequency target acoustic scattering all adopts simulation methods. It is necessary to divide the target surface into surface elements according to a certain linear dimension, resulting in tens of thousands or even hundreds of thousands of surface elements. Then, the final result is obtained by integrating these surface elements. In the calculation process, it is necessary to remove surface elements that have been blocked multiple times through occlusion technology. At the same time, it is necessary to use ray tracing algorithms to determine the propagation paths of incident and outgoing sound rays, multiple transmissions, and multiple scatterings. The calculation speed is extremely slow and cannot meet the requirements of real-time detection and analysis in engineering. Summary of the Invention

[0003] To overcome the aforementioned problems in the existing technology, this invention proposes a rapid prediction method for target acoustic scattering.

[0004] The technical solution adopted by this invention to solve its technical problem is: a method for rapid prediction of target acoustic scattering, specifically including the following steps:

[0005] Step 1: Given the calculation range and calculation step size, determine the number of field points to be calculated based on the calculation range and step size;

[0006] Step 2: Automatically obtain the current number of computer CPU cores. Based on the number of field points to be calculated and the number of CPU cores involved in the calculation obtained in Step 1, automatically assign a calculation task to each CPU core.

[0007] Step 3: Automatically create a corresponding parallel computing folder for each CPU core participating in the computation, automatically generate a parallel computing task book, and copy the main executable program into the parallel folder;

[0008] Step 4: Sequentially activate the executable programs in each parallel computing folder. After each executable program is activated, check if there is a parallel computing task book. If there is, perform the calculation directly according to the requirements of the computing task book and store the data calculation results one by one. If there is no task book, do not perform parallel computing and enter the main program interface.

[0009] Step 5: Perform occlusion calculations for each single field point that needs to be calculated, traverse all face elements, and remove occluded face elements.

[0010] Step 6: Using a ray tracing algorithm, calculate the acoustic scattering values ​​of all inner shell surface elements and all outer shell surface elements after removing the occluding surface elements. Then, vector-wise sum the acoustic scattering values ​​of all inner shell surface elements and all outer shell surface elements after removing the occluding surface elements to obtain the acoustic scattering values ​​of the complex double-shell target.

[0011] In the above-mentioned method for rapid prediction of target acoustic scattering, when performing occlusion calculation in step 5, the outer shell surface element and the inner shell surface element start cycling simultaneously. If the surface element to be judged is an inner shell surface element, all inner shell surface elements are extracted and judged against the inner shell surface element; if the surface element to be judged is an outer shell surface element, all outer shell and inner shell surface elements are extracted and judged.

[0012] The above-mentioned method for rapid prediction of target acoustic scattering, specifically includes the following process in step 5:

[0013] Step a: Based on the calculated far field or near field, sort the inner shell elements by depth or distance, and convert the three-dimensional coordinates of the target element nodes into two-dimensional coordinates on the projection plane.

[0014] Step b: In the projection plane, determine whether a node of a certain inner shell element falls into other elements.

[0015] Step c, repeat step b, and determine whether all nodes of the current face element fall into other face elements. If yes, the current face element is occluded; otherwise, the current face element is not occluded.

[0016] In the aforementioned method for rapid prediction of target acoustic scattering, step b involves determining whether a node of an inner shell element falls into other elements using the cross product method and the pillbox algorithm.

[0017] The aforementioned method for rapid prediction of target acoustic scattering, specifically includes the cross product judgment method: the judgment point is P0, and the vertices of the polygon are arranged in order as P1P2...P n V i =P i -P0,i=1,2,...,n,V n+1 =V1; Determine V i ×V i+1 (i = 1, 2, ... n) Check if the signs of the cross products are the same. If the signs are the same, then point P0 is inside the polygon. If the signs are different, then point P0 is outside the polygon.

[0018] The aforementioned method for rapid prediction of target acoustic scattering, specifically the "medicine box algorithm," includes: dividing the projection plane into a finite number of regions, placing all the nodes of the target in different small regions, and when one of the target surface elements needs to be judged, extracting the small region where this node is located and other nodes in the several small regions adjacent to this node for judgment.

[0019] In the aforementioned method for rapid prediction of target acoustic scattering, step 6 specifically includes:

[0020] Step 6.1: Calculate the coordinates of the center point of the inner shell surface element based on the coordinates of the three nodes of the inner shell surface element; connect the emission point to the center point of the inner shell surface element and calculate the angle between the connecting line and the normal direction of the inner shell surface element; calculate the reflection coefficient of the inner shell surface element through linear interpolation.

[0021] Step 6.2: Use a ray tracing algorithm to find the shell surface elements that intersect with the incident sound ray path, and select the shell surface element closest to the viewpoint;

[0022] Step 6.3, repeat step 6.1 to calculate the transmission coefficient of the incident sound rays passing through the outer shell surface element corresponding to the inner shell surface element described in step 6.1;

[0023] Step 6.4: For the emitted sound rays, repeat steps 6.2-6.3. This time, take the direction of the line connecting the receiving point and the center point of the surface element as the tracking direction, and obtain the transmission coefficient of the emitted sound rays corresponding to the inner shell surface element as described in step 6.1 passing through the outer shell surface element.

[0024] Step 6.5, the comprehensive reflection and transmission coefficient of the inner shell surface element = the reflection coefficient of the inner shell surface element obtained in step 6.1 * the transmission coefficient of the incident sound ray passing through the outer shell surface element corresponding to the inner shell surface element obtained in step 6.3 * the transmission coefficient of the outgoing sound ray passing through the outer shell surface element corresponding to the inner shell surface element obtained in step 6.4.

[0025] Step 6.6: Calculate the acoustic scattering value of the inner shell surface element based on the comprehensive reflection and transmission coefficient of the inner shell surface element obtained in Step 6.5.

[0026] Step 6.7: Traverse all inner shell elements and repeat steps 6.1-6.6 to finally obtain the acoustic scattering values ​​of all inner shell elements.

[0027] Step 6.8: Calculate the coordinates of the center point of the outer shell element based on the coordinates of the three nodes of the outer shell element; connect the emission point to the center point of the outer shell element and calculate the angle between the connecting line and the normal direction of the outer shell element; calculate the reflection coefficient of the outer shell element through linear interpolation; calculate the acoustic scattering value of the outer shell element; traverse all outer shell elements to obtain the acoustic scattering values ​​of all outer shell elements.

[0028] Step 6.9: Add the acoustic scattering values ​​of all inner shell elements obtained in Step 6.7 to the acoustic scattering values ​​of all outer shell elements obtained in Step 6.8 to obtain the acoustic scattering values ​​of the complex double-shell target.

[0029] In the aforementioned rapid target acoustic scattering prediction method, the calculation formulas for the acoustic scattering values ​​in steps 6.6 and 6.8 are as follows:

[0030]

[0031] Where j represents the unit imaginary number; k is the wave number in water, where k = 2πf / C, f is the calculated frequency, and C is the speed of sound in water; r represents the distance between the transmitting or receiving point and the center point of the surface element; ∫ S+ V(θ) represents the area integral of the target illumination. i ,f) represents the combined reflection and transmission coefficient, where, for the inner shell element, V(θ) i f) is the combined reflection and transmission coefficient of the inner shell element. For the outer shell element, V(θ) i f) represents the reflection coefficient of the outer shell element; θ i Indicates the angle of incidence.

[0032] In the aforementioned rapid target acoustic scattering prediction method, the specific calculation formula for linear interpolation in step 6.1 is as follows:

[0033]

[0034] in,

[0035] [cosθ] i <cosθ<[cosθ] i+1 f j <f<f j+1 ;

[0036] [cosθ] i f j The base point for interpolation is 1; the cosθ value ranges from 1 to 0.1, with an interval of 0.01; f ranges from 0 Hz to 50 kHz, with an interval of 100 Hz.

[0037] The beneficial effects of this invention are that, through the medicine box algorithm, it significantly reduces the computation time for occlusion and ray tracing in the algorithm; it can predict multiple sets of boundary conditions of the target in a single calculation; and through the self-developed adaptive parallel technology, it can perfectly utilize all CPU cores when calculating multiple field points. Ultimately, this invention can shorten the calculation time for a single field point to within 4 seconds and increase the calculation speed for multiple field points by about 4 orders of magnitude, which greatly improves the calculation speed of acoustic scattering prediction for underwater targets and can meet the requirements of real-time detection and analysis in engineering. Attached Figure Description

[0038] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0039] Figure 1 This is a schematic diagram of a point transmission projection in an embodiment of the present invention;

[0040] Figure 2 This is a flowchart of the cross product determination method in an embodiment of the present invention;

[0041] Figure 3 This is a schematic diagram of the medicine box algorithm in an embodiment of the present invention;

[0042] Figure 4 This is a transmission image of a point in the observation coordinate system in an embodiment of the present invention. Detailed Implementation

[0043] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0044] In mid-to-high frequency scenarios, the Kirchoff approximate integral method is currently the primary approach for predicting the echo characteristics of complex underwater targets. This method uses surface elements to simulate and fit the target's shape, integrates each surface element to calculate its contribution to scattering, and the total scattered sound from the target is the vector sum of the scattering from each surface element.

[0045] Equation (1) is the Kirchhoff approximate integral formula for the target echo in the far-field transmit / receive combined case:

[0046]

[0047] ∫ S+ V(θ) represents the area integral of the target illumination. i f) is the reflection coefficient, θ i The angle of incidence is denoted as .

[0048] As can be seen from formula (1), to calculate the sound scattering of a target, it is not only necessary to use occlusion techniques to remove the shadowed surface of the target and integrate only on the illuminated surface, but also to use ray tracing algorithms to determine the propagation path of the sound rays, thereby accurately obtaining the V(θ) of each surface element. i θ in f) i .

[0049] When performing occlusion or ray tracing calculations, within the projection plane, each facet's three nodes (or viewpoint) needs to be iterated through and compared with all other facet nodes. Assuming there are N nodes, the number of iterations is N*(N-1). When N is in the tens or hundreds of thousands, the computation time becomes unacceptable. The "medicine box" algorithm divides the projection plane into a finite number of regions, similar to small boxes in a traditional Chinese medicine shop, and then places all the target's nodes into different small boxes. Figure 3 As shown, when one of the nodes on the target surface (such as...) Figure 3 When small and medium-sized points need to be judged, it is only necessary to extract the small box where the node is located and other nodes in the small boxes adjacent to the node for judgment (nine small boxes in the light-colored box in the figure). It is not necessary to loop through all target nodes, and the number of loops becomes N* (several to dozens), which greatly reduces the calculation time of occlusion and ray tracing.

[0050] From formula (1), we can see that V(θ) i V(θ) is a two-dimensional array that depends on the incident angle and frequency. However, there are many possible boundary conditions for the target, i.e., V(θ) i There are several sets of boundary conditions (f), but only one set can be calculated at a time, which greatly reduces the computational efficiency. Now, by using a self-developed technique to calculate multiple sets of boundary conditions at once, the algorithm can predict multiple sets of boundary conditions for the target in a single calculation. The method is to use V(θ) i f) becomes a three-dimensional array. After the occlusion and ray tracing calculations are completed, multiple sets of boundary conditions are substituted into the final integration. The final calculation result is also changed to a three-dimensional array, and all results will be generated at the same time.

[0051] Both of the above methods are for improving the calculation speed when there is a single field point (i.e., one transmitter and one receiver). If there are multiple transmitters and multiple receivers that need to be calculated, a self-developed adaptive parallel algorithm can be used to improve the calculation speed for multiple field points.

[0052] The adaptive parallel algorithm first determines the number of field points that need to be computed. Then, it automatically obtains the number of logical cores of the current computer CPU. Based on the total number of field points and the number of CPU logical cores, it automatically assigns a computation task book to each CPU core. It automatically creates a corresponding parallel computing folder for each participating CPU core and automatically copies the main executable program to the corresponding parallel computing folder. Finally, it sequentially activates the executable programs in each parallel computing folder. After each parallel computing executable program is activated, it first checks whether a parallel computing task book exists. If it exists, it directly performs the computation according to the requirements of the computation task book and stores the data computation results one by one. If no task book exists, parallel computing is not performed.

[0053] This method can perfectly utilize all the cores of a computer without human intervention. It adapts to the number of CPU cores, automatically allocates the computational load of each CPU core (evenly distributed), automatically generates a computational task sheet, and automatically stores the computational results.

[0054] The above three methods can ultimately reduce the calculation time for a single field point to within 4 seconds, and increase the calculation speed for multiple fields by about 4 orders of magnitude (taking 48 CPU cores as an example).

[0055] The specific steps in this embodiment are as follows:

[0056] 1. Given the calculation range (including distance, pitch angle, horizontal angle, and split angle) and the calculation step size;

[0057] 2. Determine the total number of field points to be calculated by using the range and step size of the four variables;

[0058] 3. Automatically obtain the current number of computer CPU cores, and automatically assign computing tasks to each CPU core based on the total computing load (total number of points) and the number of CPU cores participating in the computing.

[0059] 4. Automatically create a corresponding parallel computing folder for each CPU core participating in the computation, automatically generate the corresponding parallel computing task book, and automatically copy the main executable program to the corresponding parallel computing folder.

[0060] 5. Activate the executable programs in each parallel computing folder sequentially. After each parallel computing executable program is activated, first check if there is a parallel computing task book. If there is, perform the calculation directly according to the requirements of the computing task book and store the data calculation results one by one. If there is no task book, do not perform parallel computing.

[0061] This is mainly because the executable program for parallel computing is exactly the same as the main executable program. In order to avoid getting stuck in an infinite loop, a flag variable must be given. This flag variable is the computing task book. If there is a task book, parallel computing will start; if there is no task book, the main program interface will be entered.

[0062] 6. For each single field point that needs to be calculated, occlusion calculation should be performed first. Based on the calculation results (far field or near field), the target surface elements should be sorted by depth or distance. Different projection methods should be selected to convert the three-dimensional coordinates of the target surface element nodes into two-dimensional coordinates on the projection plane.

[0063] In the far field, the incident wave is a plane wave, and the projection plane can be simply set as the coordinate plane. For example, when the projection plane is the xoy plane, the x and y coordinates of each node of each surface element are the coordinates of each node of that surface element on the projection plane.

[0064] In the near field, the incident wave is a spherical wave. Solving for the coordinates of each node of the surface element on the projection plane involves the following two cases:

[0065] (1) The viewpoint is located on the x-axis, y-axis or z-axis.

[0066] The example illustrates the case where the viewpoint is on the z-axis; the same logic applies to the case where the viewpoint is on other axes. Let P be the viewpoint (emission point) of the perspective projection. c The coordinates are (x c ,y c ,z c The projection plane is the xoy plane, and the coordinates of a point P(x,y,z) on the target on the xoy plane are (x...y...z ... s ,y s ),like Figure 1 As shown. By P c The projection equation for P can be obtained as follows:

[0067] x s =x c +(xx c )t

[0068] y s =y c +(yy c )t

[0069] z s =z c +(zz c )t

[0070] It intersects the XOY plane at (x s ,y s ,z s According to z s =0 can be used to find:

[0071] x s =(x c z-xz c ) / (zz c )

[0072] y s =(y c z-yz c ) / (zz c ).

[0073] (2) The viewpoint is located at any point in space other than the coordinate axes.

[0074] When the viewpoint is located at any point in space other than the coordinate axes, a transition coordinate system, called the observation coordinate system, needs to be introduced, such as... Figure 4As shown. In the observation coordinate system, let the viewpoint be (a, b, c), and the distance of the viewing plane from the viewpoint in the observation direction be z. s ,set up The vertex coordinates of the shape are (x w ,y w ,z w The coordinates transformed to the observation coordinate system are (x... e ,y e ,z e The coordinates of the projection onto the view plane after perspective projection are (x...). s ,y s If the perspective projection transformation formula is:

[0075]

[0076] Then we can obtain:

[0077] x e =-b / v·x w +a / v·y w

[0078] y e =-ac / uv·x w -bc / uv·y w +v / u·z w

[0079] z e =-c / u·x w -b / u·y w -c / u·z w +u

[0080] x s =x e ·z s / z e

[0081] y s =y e ·z s / z e .

[0082] 7. Following the rule that the depth (or distance) of polygons increases from far to near, determine whether a node of a given polygon falls within other polygons in the projection plane. Here, the cross product method is still used. Assume the judgment point is P0, and the polygon vertices are arranged in order as P1P2...P n , making V i =P i -P0, i = 1, 2, ..., n, V n+1 =V1. Therefore, the necessary and sufficient condition for P0 to lie within the polygon is that the cross product V... i ×Vi+1 (i = 1, 2, ..., n) have the same sign. See the flowchart for the cross product test. Figure 2 .

[0083] In this judgment process, the medicine box algorithm is used to divide the projection plane into a finite number of regions and put all the nodes of the target into different small regions. When one of the nodes of the target surface element needs to be judged, it is only necessary to extract the small region where this node is located and other nodes in the small regions adjacent to this node for judgment.

[0084] 8. Repeat step 7 to determine whether all nodes of the current face element fall within other face elements. If so, the face element is occluded; otherwise, the face element is not occluded. Traverse all face elements to completely remove the occluded part of the target.

[0085] 9. Using a ray tracing algorithm, calculate the acoustic scattering values ​​of all inner shell elements and all outer shell elements after removing the occluded elements. Then, add the acoustic scattering values ​​of all inner shell elements and all outer shell elements after removing the occluded elements to obtain the acoustic scattering values ​​of the complex double-shell target.

[0086] Step 9 specifically includes:

[0087] 9.1 First, determine the coordinates of the center point of the inner shell element based on the coordinates of its three nodes.

[0088] O x =(R0[P0[i]]+R0[P1[i]]+R0[P2[i]]) / 3.0;

[0089] O y =(R1[P0[i]]+R1[P1[i]]+R1[P2[i]]) / 3.0;

[0090] O z =(R2[P0[i]]+R2[P1[i]]+R2[P2[i]]) / 3.0;

[0091] Then, connect the emission point to the center point of the surface element, and use the law of cosines to find the cosine of the angle between the line connecting the two points and the normal direction of the surface element:

[0092] cosk ni =(InX-O) x )*N0[i]+(InY-O y )*N1[i]+(InZ-O z )*N2[i];

[0093] Finally, the length of the connecting line is obtained, and the cosine value of the included angle is normalized.

[0094] R i O = sqrt((InX-O) x )*(InX-O x )+(InY-O y )*(InY-O y )+(InZ-O z )*(InZ-O z ));

[0095] cosk ni / =R i O.

[0096] 9.2. Calculate the reflection coefficient on the inner shell using linear interpolation, using cosθ values ​​from 1 to 0.1 in intervals of 0.01 and frequencies from 0 Hz to 50 kHz in intervals of 100 Hz, for the reflection coefficient V([cosθ]). i ,f j The value of [cosθ], linearly interpolated reflection coefficients for other angles and frequencies. i f j This is the interpolation base point. The calculation formula is:

[0097]

[0098] in

[0099] [cosθ] i <cosθ<[cosθ] i+1 f j <f<f j+1 .

[0100] 9.3. To determine which (or several) of the outer shell elements intersect the incident sound ray path using ray tracing, the first step is to project the coordinates of all outer shell element nodes onto the corresponding view plane using a one-point transmission projection method, such as... Figure 1 As shown.

[0101] Let P be the viewpoint (emission point) of the perspective projection. c The coordinates are (x c ,y c ,z c The projection plane is the XOY plane (taking the Z-axis as an example). The XOY coordinates of a point P(x,y,z) on the target are (x,y,z) and (y,z) respectively. s ,y s ). By P c The projection equation for P can be obtained as follows:

[0102] x s =xc +(xx c )t

[0103] y s =y c +(yy c )t

[0104] z s =z c +(zz c )t

[0105] It intersects the XOY plane at (x s ,y s ,z s According to z s =0 can be used to find:

[0106] x s =(x c z-xz c ) / (zz c )

[0107] y s =(y c z-yz c ) / (zz c ).

[0108] When the viewpoint is located at any point in space other than the coordinate axes, a transition coordinate system is introduced, called the observation coordinate system. In the observation coordinate system, let the viewpoint be (a, b, c), and the distance of the viewing plane from the viewpoint along the observation direction be z. s ,set up The vertex coordinates of the shape are (x w ,y w ,z w The coordinates transformed to the observation coordinate system are (x... e ,y e ,z e The coordinates of the projection onto the view plane after perspective projection are (x...). s ,y s If the perspective projection transformation formula is:

[0109]

[0110] Then we can obtain:

[0111] x e =-b / v·x w +a / v·y w

[0112] y e =-ac / uv·x w -bc / uv·yw +v / u·z w

[0113] z e =-c / u·x w -b / u·y w -c / u·z w +u

[0114] x s =x e ·z s / z e

[0115] y s =y e ·z s / z e .

[0116] 9.4. To determine whether a viewpoint intersects with a certain shell element, the cross product method is used. Assume the judgment point is P0, and the polygon vertices are arranged in order as P1P2...P n , making V i =P i -P0, i = 1, 2, ..., n, V n+1 =V1. Therefore, the necessary and sufficient condition for P0 to lie within the polygon is that the cross product V... i ×V i+1 (i = 1, 2, ..., n) have the same sign. See the flowchart for the cross product test. Figure 2 In this judgment process, the medicine box algorithm is used to divide the projection plane into a finite number of regions and put all the nodes of the target into different small regions. When the viewpoint needs to be judged with the surface node, it is only necessary to extract the surface node in the small region where the viewpoint is located and other surface nodes in the small regions adjacent to the viewpoint for judgment.

[0117] 9.5 During the judgment process, there may be a situation where several shell elements overlap on the projection plane, causing the incident sound ray to pass through several shell elements at the same time. Only the shell element closest to the viewpoint needs to be selected, and the number of the shell element should be recorded.

[0118] 9.6. Obtain the coordinates of the three nodes of the outer shell element by using its number. Repeat steps 9.1 and 9.2 to obtain the transmission coefficient of the incident sound ray penetrating the outer shell element corresponding to the inner shell element of the incident sound ray.

[0119] 9.7 Similarly, for the emitted sound rays, repeat steps 9.3 to 9.6. Note that this time, the direction of the line connecting the receiving point and the center point of the surface element is used as the tracking direction to obtain the transmission coefficient of the emitted sound ray corresponding to the inner shell surface element that penetrates the outer shell surface element.

[0120] 9.8 At this point, the reflection coefficient of the inner shell surface element in step 9.2, the transmission coefficient of the incident sound ray corresponding to the inner shell surface element through the outer shell surface element in step 9.6, and the transmission coefficient of the outgoing sound ray corresponding to the inner shell surface element through the outer shell surface element in step 9.7 have all been obtained. Therefore, the comprehensive reflection and transmission coefficient corresponding to the inner shell surface element = the reflection coefficient of the inner shell surface element in step 9.2 * the transmission coefficient of the incident sound ray corresponding to the inner shell surface element through the outer shell surface element in step 9.6 * the transmission coefficient of the outgoing sound ray corresponding to the inner shell surface element through the outer shell surface element in step 9.7.

[0121] 9.9 Calculate the acoustic scattering value of the inner shell element using formula (1);

[0122] 9.10. Traverse all inner shell elements and repeat steps 9.1 to 9.9 to finally obtain the acoustic scattering values ​​of all inner shell elements;

[0123] 9.11 For the outer shell element, since it does not involve ray tracing, it is only necessary to traverse all the outer shell elements, repeat steps 9.1 and 9.2 to obtain the reflection coefficient of all the outer shell elements, and then the acoustic scattering value of all the outer shell elements can be obtained by formula (1).

[0124] 9.12. Add the acoustic scattering contributions of all inner shell elements to the acoustic scattering contribution vectors of all outer shell elements to obtain the final acoustic scattering values ​​of the overall complex double-shell target.

[0125] The above embodiments are merely exemplary embodiments of the present invention and are not intended to limit the present invention. The scope of protection of the present invention is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to the present invention within its spirit and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of the present invention.

Claims

1. A method of fast prediction of target acoustic scattering, characterized in that, Specifically, the steps include the following: Step 1: Given the calculation range and calculation step size, determine the number of field points to be calculated based on the calculation range and step size; Step 2: Automatically obtain the current number of computer CPU cores. Based on the number of field points to be calculated and the number of CPU cores involved in the calculation obtained in Step 1, automatically assign a calculation task to each CPU core. Step 3: Automatically create a corresponding parallel computing folder for each CPU core participating in the computation, automatically generate a parallel computing task book, and automatically copy the main executable program to the parallel folder. Step 4: Activate the executable programs in each parallel computing folder in sequence. After each executable program is activated, check if there is a parallel computing task book. If there is, perform the calculation directly according to the requirements of the computing task book and store the data calculation results one by one. If no task description exists, parallel computing will not be performed, and the main program interface will be entered. Step 5: Perform occlusion calculations for each single field point that needs to be calculated, traverse all face elements, and remove occluded face elements. Step 6: Using a ray tracing algorithm, calculate the acoustic scattering values ​​of all inner shell surface elements and all outer shell surface elements after removing the occluded surface elements. Then, vectorically add the acoustic scattering values ​​of all inner shell surface elements and all outer shell surface elements after removing the occluded surface elements to obtain the acoustic scattering values ​​of the complex double-shell target. Step 6 specifically includes: Step 6.1: Calculate the coordinates of the center point of the inner shell surface element based on the coordinates of the three nodes of the inner shell surface element; connect the emission point to the center point of the inner shell surface element and calculate the angle between the connecting line and the normal direction of the inner shell surface element; calculate the reflection coefficient of the inner shell surface element through linear interpolation. Step 6.2: Use a ray tracing algorithm to find the shell surface elements that intersect with the incident sound ray path, and select the shell surface element closest to the viewpoint; Step 6.3, repeat step 6.1 to calculate the transmission coefficient of the incident sound rays passing through the outer shell surface element corresponding to the inner shell surface element described in step 6.1; Step 6.4: For the emitted sound rays, repeat steps 6.2-6.

3. This time, take the direction of the line connecting the receiving point and the center point of the surface element as the tracking direction, and obtain the transmission coefficient of the emitted sound rays corresponding to the inner shell surface element as described in step 6.1 passing through the outer shell surface element. Step 6.5, the comprehensive reflection and transmission coefficient of the inner shell surface element = the reflection coefficient of the inner shell surface element obtained in step 6.1 * the transmission coefficient of the incident sound ray corresponding to the inner shell surface element that penetrates the outer shell surface element obtained in step 6.3 * the transmission coefficient of the outgoing sound ray corresponding to the inner shell surface element that passes through the outer shell surface element obtained in step 6.

4. Step 6.6: Calculate the acoustic scattering value of the inner shell surface element based on the comprehensive reflection and transmission coefficient of the inner shell surface element obtained in Step 6.

5. Step 6.7: Traverse all inner shell elements and repeat steps 6.1-6.6 to finally obtain the acoustic scattering values ​​of all inner shell elements. Step 6.8: Calculate the coordinates of the center point of the outer shell element based on the coordinates of the three nodes of the outer shell element; connect the emission point to the center point of the outer shell element and calculate the angle between the connecting line and the normal direction of the outer shell element; calculate the reflection coefficient of the outer shell element through linear interpolation; calculate the acoustic scattering value of the outer shell element; traverse all outer shell elements to obtain the acoustic scattering values ​​of all outer shell elements. Step 6.9: Add the acoustic scattering values ​​of all inner shell elements obtained in Step 6.7 to the acoustic scattering values ​​of all outer shell elements obtained in Step 6.8 to obtain the acoustic scattering values ​​of the complex double-shell target.

2. The method of claim 1, wherein, When performing the occlusion calculation in step 5, the outer shell surface element and the inner shell surface element start looping simultaneously. If the surface element to be judged is an inner shell surface element, all inner shell surface elements are extracted and judged against the inner shell surface element; if the surface element to be judged is an outer shell surface element, all outer shell and inner shell surface elements are extracted and judged.

3. The method for rapid prediction of target acoustic scattering according to claim 1, characterized in that, The specific process of the occlusion calculation in step 5 includes: Step a: Based on the calculated far field or near field, sort the inner shell elements by depth or distance, and convert the three-dimensional coordinates of the target element nodes into two-dimensional coordinates on the projection plane. Step b: In the projection plane, determine whether a node of a certain inner shell element falls into other elements. Step c, repeat step b, and determine whether all nodes of the current face element fall into other face elements. If yes, the current face element is occluded; otherwise, the current face element is not occluded.

4. The method for rapid prediction of target acoustic scattering according to claim 3, characterized in that, In step b, the method of determining whether a node of an inner shell element falls into another element is the cross product method and the medicine box algorithm.

5. The method for rapid prediction of target acoustic scattering according to claim 4, characterized in that, The cross product judgment method specifically includes: the judgment point is... The vertices of the polygon are arranged in order as follows ; ;judge Check if the signs of the cross products are the same; if they are the same, then determine the point. Within a polygon, if the signs are different, then determine the points. Outside the polygon.

6. The method for rapid prediction of target acoustic scattering according to claim 4, characterized in that, The specific steps of the medicine box algorithm include: dividing the projection plane into a finite number of regions, placing all the nodes of the target into different small regions, and when one of the nodes of the target surface element needs to be judged, extracting the small region where this node is located and other nodes in the small regions adjacent to this node for judgment.

7. The method for rapid prediction of target acoustic scattering according to claim 1, characterized in that, The formulas for calculating the acoustic scattering values ​​in steps 6.6 and 6.8 are as follows: ； Where j represents the imaginary unit; k is the wave number in water, where f is the calculated frequency, C is the speed of sound in water, and r represents the distance between the transmitting or receiving point and the center of the surface element. Represents the area integral of the target illumination; This represents the combined reflection and transmission coefficient, where, for the inner shell element, For the inner shell surface element, the comprehensive reflection and transmission coefficient is given; for the outer shell surface element, ... The reflectance coefficient of the outer shell surface element; Indicates the angle of incidence.

8. The method for rapid prediction of target acoustic scattering according to claim 1, characterized in that, The specific calculation formula for linear interpolation in step 6.1 is as follows: ； in, , ， ， ； , Used as the interpolation base point; The values ​​range from 1 to 0.1, with intervals of 0.01; f ranges from 0 Hz to 50 kHz, with intervals of 100 Hz.

Images