A Particle-Tracking-Based Ocean Simulation System and Method

By performing particle culling, splitting, and tracking on the GPU, combined with a hierarchical bounding box acceleration structure, the problems of data volume growth and low computational efficiency in ocean simulation are solved, achieving highly efficient ocean simulation results.

CN117710556BActive Publication Date: 2026-11-17SICHUAN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410025229.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-08
Publication Date
2026-11-17
Estimated Expiration
2044-01-08

AI Technical Summary

Technical Problem

In existing ocean simulation methods, the amount of data increases exponentially after particles split on the CPU, leading to performance bottlenecks, low computational efficiency, uneven height fields at particle splitting positions, and wasted computation on invisible particles.

Method used

Particle culling, splitting, and tracking are performed on the GPU. A hierarchical bounding box acceleration structure is used for data acceleration. Particle splitting computation has high parallelism, and collision and reflection events are processed on the GPU to reduce CPU load.

Benefits of technology

It improves the efficiency of ocean simulation by more than 90%, enhances the detail and accuracy of boundary collisions and reflections, avoids performance waste, and generates a smoother and more natural height field.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117710556B_ABST
    Figure CN117710556B_ABST
Patent Text Reader

Abstract

This invention discloses a particle tracking-based ocean simulation system and method, including a data preprocessing module, a wave generation module, and a visualization module. The method includes: Step 1, generating an accelerated structure; Step 2, dividing the ocean into blocks; Step 3, interactively generating particles; and Step 4, generating an ocean height map. This invention removes invisible particles on the GPU side, avoiding performance waste. All calculations of particle splitting, collision, and reflection events are performed on the GPU side, resulting in high parallelism and computational efficiency, avoiding performance bottlenecks caused by CPU processing. By increasing the number of particle splits, the height field becomes smoother and more natural. The use of a hierarchical bounding box acceleration structure to accelerate particle tracking further improves processing speed. Compared to wave-particle algorithms, this invention significantly improves the efficiency of wave simulation, and provides higher accuracy in particle-boundary collision and reflection, resulting in better visual effects and richer wave details.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of marine simulation technology, specifically to a marine simulation system and method based on particle tracking. Background Technology

[0002] Ocean simulation uses computer technology to model ocean scenes. Because the ocean is a complex fluid system, it's difficult to achieve good real-time performance and realism simply by solving the Navier-Stokes equations (NS). Existing ocean simulation methods generally employ wave-particle duality algorithms, specifically: dividing the water surface into multiple small grids and discretizing the flowing water surface into flowing particles. Each particle carries physical information such as amplitude, velocity, and spread angle, managed on the CPU. During propagation, particles gradually split as they travel distance, splitting from one particle into three at a time, with each particle carrying one-third of its original amplitude information. The split particle information is transmitted to the GPU to calculate the grid where the particle resides. Several particles interact with their corresponding grids, forming a local height field. A filtering algorithm is used to fuse these local height fields to obtain a complete surface height field, thus providing a complete ocean surface simulation.

[0003] The above method has the following drawbacks when applied: First, after the particles split several times on the CPU, the amount of data that needs to be transferred to the GPU increases exponentially, forming a performance bottleneck; second, the calculation of invisible particles together causes unnecessary performance waste; third, the calculation of particle splitting, collision, reflection and other events is all performed on the CPU with fewer cores, resulting in low computational efficiency; fourth, the amplitude information carried by the particles during splitting becomes one-third of the original particle, causing abrupt changes in the height field at the corresponding splitting position, which is not smooth and natural. Summary of the Invention

[0004] The purpose of this invention is to provide a particle-tracking-based ocean simulation system and method to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a particle tracking-based ocean simulation system, comprising a data preprocessing module, a wave generation module, and a visualization module. The wave generation module establishes data connections with the data preprocessing module and the visualization module, respectively. The data preprocessing module includes a scene import unit and an accelerated structure generation unit. The wave generation module includes a particle tracking unit and a texture generation unit. The visualization module includes a scene rendering unit, an effect generation unit, and an interaction support unit.

[0006] Preferably, the scene import unit is used to obtain vertex data from the imported scene boundary information; the acceleration structure generation unit analyzes the imported scene data to generate the bottom-level acceleration structure and top-level acceleration structure required for tracking particles; the particle tracking unit is used to track the interactively generated particles on the top-level acceleration structure to obtain the final action position of the particles; the texture generation unit is used to generate the height map and normal map required for drawing the ocean based on the local height field generated by the particle action after tracking; the scene rendering unit is used to perform scene rendering based on the generated texture and other scene elements; the effect generation unit is used to generate the visualization effect of the open ocean for the scene; and the interaction support unit is used to support users to interact with the 3D visualization scene.

[0007] A particle-tracking-based ocean simulation method includes the following steps: Step 1, generating an acceleration structure; Step 2, dividing the ocean into blocks; Step 3, interactively generating particles; and Step 4, generating an ocean height map. In step one above, the bottom-level acceleration structure and the top-level acceleration structure are parsed and generated from the vertex data of the sea surface boundary conditions. In step two above, the ocean to be simulated is divided into blocks, and each block of ocean uses a separate height map to simulate waves; In step three above, when the object interacts with the water surface, the CPU generates particles carrying physical information based on the physical information generated by the interaction. In step four above, the particles generated in step three are passed to the GPU. The GPU performs culling, splitting and tracking of the generated particles. After tracking is completed, the particles are determined to act on the ocean blocks divided in step two based on their location, and then act on the corresponding height map.

[0008] Preferably, in step one, generating the acceleration structure specifically involves: 1.1 Constructing Boundary Conditions: Describing the ocean boundary conditions in a 2D plane using vertex data; 1.2 Generate acceleration structure: Use vertex data to generate several low-level acceleration structures, and use the collection of low-level acceleration structures to generate a top-level acceleration structure.

[0009] Preferably, in step 1.1, the vertex data is projected onto the xy plane, while vertex data that is above sea level and will not cause wave reflection is discarded; that is, for vertices... have: .

[0010] Preferably, in step 1.2, vertex buffer data VertexBuffer and index buffer data IndexBuffer are created using the projected vertex data to generate the underlying acceleration structure.

[0011] Preferably, in step two, the ocean surface to be simulated is divided into several smaller planes due to accuracy limitations. Each plane has its own height map to record the local height field generated after being affected by particles. The generation of the height field can be represented as follows: .

[0012] in, The amplitude carried by the particle. This is a fixed wave function. Then we can obtain the result through the following formula: .

[0013] Preferably, in step three, the physical information carried by the particle includes generation time, generation position, velocity, diffusion angle, amplitude, and radius. The CPU deletes the particle when the particle amplitude is lower than a specified threshold. The CPU transmits each particle and its attributes as vertex data to the GPU every frame.

[0014] Preferably, step four specifically includes: 4.1 Particle Removal: Based on the physical information carried by the particle, the diffusion radius of the wave that the particle can generate can be calculated. By intersecting the diffusion range with the view frustum used by the visualization system, particles that generate invisible waves can be eliminated. Here, the efficient parallel computing capability of the magnification shader is used to complete the elimination. 4.2 Particle Splitting: After culling, the amplification shader splits each remaining particle into... Several particles located on the diffusion arc, these particles carry amplitude information of the original particle. And ultimately, it actually acts on the height field; 4.3 Particle Tracking: After amplifying the shader, the mesh shader is called to perform parallel particle tracking calculations. The particle information passed to the mesh shader includes the velocity vector, amplitude, and diffusion radius of the particles after splitting. Including the initial particle position information, the tracking process uses a hardware-based ray tracing architecture based on modern graphics hardware. Simultaneously, it utilizes the top-level acceleration structure generated in step one for data acceleration, enabling rapid determination of the particle's intersection with the ocean boundary. Upon detecting boundary collision information, the particle undergoes reflection processing until its trajectory reaches the specified point. That is, the final position of the particle; 4.4 The data generated by the mesh shader is passed to the pixel shader to form a local height field; 4.5 The obtained local height fields are fused using a filtering algorithm to obtain the overall height field for ocean simulation.

[0015] Preferably, in step 4.1, the calculation of the diffusion radius is specifically performed by: based on the velocity information carried by the particles. The generation time can be calculated using the following formula: Particles, in time diffusion radius at time ,Right now: .

[0016] Within the radius An arc with a central angle (diffusion angle) of α contains particles after splitting. When the arc is completely outside the view frustum, the particles are considered invisible and no further calculations are needed.

[0017] In step 4.2, based on the particles removed in step 4.1, the water wave is assumed to have a radius after splitting. The particles are tangent to each other along the arc. Based on this assumption, the angle between the center of two adjacent particles and the location where the original particle was generated can be calculated. According to the diffusion angle Angle and radius of the arc This allows us to obtain the position and velocity of the desired generated particles, after splitting into After each particle, the amplitude of each particle is equal to the original amplitude. This process uses an amplification shader and directly passes the generated result to the mesh shader, without CPU intervention. (Angle) The calculation formula is: .

[0018] From the included angle diffusion angle The number of splits can be calculated. The calculation formula is: .

[0019] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention removes invisible particles on the GPU side, avoiding performance waste; it performs all calculations of particle splitting, collision, and reflection events on the GPU side, resulting in high parallelism and computational efficiency, avoiding performance bottlenecks caused by CPU processing; by increasing the number of particle splits, the height field becomes smoother and more natural; and by using a hierarchical bounding box acceleration structure to accelerate particle tracking data, the processing speed is further improved. Compared with the wave particle algorithm, this invention can improve simulation efficiency by more than 90% in complex scenes, and has better image details and effects, with higher accuracy in collision and reflection with boundaries. Attached Figure Description

[0020] Figure 1 This is a system block diagram of the present invention.

[0021] Figure 2 This is a system flowchart of the present invention.

[0022] Figure 3 This is a flowchart of the method of the present invention.

[0023] Figure 4 This is a comparison chart of the frame rate curves of the open ocean simulated by the present invention and the wave-particle algorithm.

[0024] Figure 5 This is a flowchart of the steps of the present invention.

[0025] In the diagram: 1. Data preprocessing module; 11. Scene import unit; 12. Accelerated structure generation unit; 2. Wave generation module; 21. Particle tracking unit; 22. Texture generation unit; 3. Visualization module; 31. Scene rendering unit; 32. Effect generation unit; 33. Interaction support unit. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Please see Figure 1-2This invention provides an embodiment of a particle-tracking-based ocean simulation system, comprising a data preprocessing module 1, a wave generation module 2, and a visualization module 3. The wave generation module 2 establishes data connections with both the data preprocessing module 1 and the visualization module 3. The data preprocessing module 1 includes a scene import unit 11 and an accelerated structure generation unit 12. The wave generation module 2 includes a particle tracking unit 21 and a texture generation unit 22. The visualization module 3 includes a scene rendering unit 31, an effect generation unit 32, and an interaction support unit 33. The scene import unit 11 is used to obtain vertex data from imported scene boundary information to accelerate structure generation. Unit 12 analyzes the imported scene data and generates the underlying and top-level acceleration structures required for particle tracking. Particle tracking unit 21 tracks the interactively generated particles on the top-level acceleration structure to obtain the final action position of the particles. Texture generation unit 22 generates the height map and normal map required for drawing the ocean based on the local height field generated by the particle action after tracking. Scene rendering unit 31 renders the scene based on the generated texture and other scene elements. Effect generation unit 32 generates the visualization effect of the open ocean for the scene. Interaction support unit 33 supports users to interact with the 3D visualization scene.

[0028] Please see Figure 3-5 The present invention provides an embodiment of an ocean simulation method based on particle tracking, comprising the following steps: Step 1, generating an acceleration structure; Step 2, dividing the ocean into blocks; Step 3, interactively generating particles; and Step 4, generating an ocean height map.

[0029] In step one above, the bottom-level acceleration structure and the top-level acceleration structure are parsed and generated from the vertex data of the sea surface boundary conditions. The specific process for generating the acceleration structure is as follows: 1.1 Constructing Boundary Conditions: Vertex data is used to describe the ocean boundary conditions in a 2D plane. This involves projecting the vertex data onto the xy plane and discarding vertices above sea level that would not cause wave reflection. have: ; 1.2 Generate the accelerated structure: Use the projected vertex data to create vertex buffer data VertexBuffer and index buffer data IndexBuffer to generate the underlying accelerated structure, and use the collection of the underlying accelerated structure to generate a top-level accelerated structure.

[0030] In step two above, the ocean to be simulated is divided into blocks, and each block uses a separate height map to simulate waves. Specifically, the ocean plane to be simulated is divided. Due to accuracy limitations, a large open ocean plane needs to be composed of several smaller planes. Each plane has its own height map to record the local height field generated after being affected by particles. The generation of the height field can be represented as: .

[0031] in, The amplitude carried by the particle. This is a fixed wave function. Then we can obtain the result through the following formula: .

[0032] In step three above, when an object interacts with the water surface, the CPU generates particles carrying physical information based on the physical information generated by the interaction. The physical information carried by the particles includes generation time, generation position, velocity, diffusion angle, amplitude, and radius. The CPU deletes the particle when the particle amplitude is lower than a specified threshold. Every frame, the CPU transmits each particle and its attributes as vertex data to the GPU.

[0033] In step four above, the particles generated in step three are fed into the graphics processing unit (GPU). The GPU performs culling, splitting, and tracking of the generated particles. After tracking, it determines their location and applies them to the ocean blocks divided in step two, and then to the corresponding height map. Specifically: 4.1 Particle Removal: Based on the physical information carried by the particle, the diffusion radius of the wave that the particle can generate can be calculated. By intersecting the diffusion range with the view frustum used in the visualization system, particles that generate invisible waves can be removed. This removal is accomplished using the efficient parallel computing capabilities of the Amplification Shader. The diffusion radius is calculated specifically based on the velocity information carried by the particles. The generation time can be calculated using the following formula: Particles, in time diffusion radius at time ,Right now: ; Within the radius A circular arc with a central diffusion angle of α contains particles after splitting. When the arc is completely outside the visual cone, the particles are considered invisible and no further calculations are needed. 4.2 Particle Splitting: After culling, the amplification shader splits each remaining particle into... Several particles located on the diffusion arc, these particles carry amplitude information of the original particle. And ultimately, it actually acts on the height field, that is, based on the particles removed in step 4.1, the water wave is assumed to have a radius of [missing information]. The particles are tangent to each other along the arc. Based on this assumption, the angle between the center of two adjacent particles and the location where the original particle was generated can be calculated. According to the diffusion angle Angle and radius of the arc This allows us to obtain the position and velocity of the desired generated particles, after splitting into After each particle, the amplitude of each particle is equal to the original amplitude. This process uses an amplification shader and directly passes the generated result to the mesh shader, without CPU intervention. (Angle) The calculation formula is: ; From the included angle diffusion angle The number of splits can be calculated. The calculation formula is: ; 4.3 Particle Tracking: After amplifying the shader, the mesh shader is called to perform parallel particle tracking calculations. The particle information passed to the mesh shader includes the velocity vector, amplitude, and diffusion radius of the particles after splitting. Including the initial particle position information, the tracking process uses a hardware-based ray tracing architecture based on modern graphics hardware. Simultaneously, it utilizes the top-level acceleration structure generated in step one for data acceleration, enabling rapid determination of the particle's intersection with the ocean boundary. Upon detecting boundary collision information, the particle undergoes reflection processing until its trajectory reaches the specified point. That is, the final position of the particle; 4.4 The data generated by the mesh shader is passed to the pixel shader to form a local height field; 4.5 The obtained local height fields are fused using a filtering algorithm to obtain the overall height field for ocean simulation.

[0034] Based on the above, the advantages of this invention are as follows: When using this invention, the data preprocessing module 1 first preprocesses the ocean boundary data, the scene import unit 11 obtains the vertex data from the boundary information in the scene, the acceleration structure generation unit 12 analyzes the imported scene data, and generates the bottom-level acceleration structure and top-level acceleration structure required for tracking particles; the preprocessed data is sent to the wave generation module 2, which generates an open ocean simulation, the particle tracking unit 21 tracks the interactively generated particles on the top-level acceleration structure to obtain the final action position of the particles, the texture generation unit 22 generates the height map and normal map required for drawing the ocean based on the local height field generated by the particle action after tracking, the visualization module 3 generates the visualization effect of the open ocean simulation generated by the wave generation module 2, the scene rendering unit 31 renders the scene based on the texture and other scene elements generated by the wave generation module 2, the effect generation unit 32 generates the visualization effect of the open ocean for the scene, and the interaction support unit 33 provides interaction support for the user and the 3D visualization scene.

[0035] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A particle-tracking-based ocean simulation method, comprising: Step 1, generating an acceleration structure; Step 2, dividing the ocean into fixed-size blocks; Step 3, interactively generating particles; Step 4, generating an ocean height map; characterized in that: In step one above, the bottom-level acceleration structure and the top-level acceleration structure are parsed and generated from the vertex data of the sea surface boundary conditions. In step two above, the ocean to be simulated is divided into several ocean blocks of the same size, and each ocean block uses a separate height map to simulate waves. In step three above, when an object interacts with the water surface, the CPU generates particles carrying physical information based on the physical information generated by the interaction. The physical information includes generation time, generation position, velocity, diffusion angle, amplitude, and radius. The CPU deletes the particle when the particle amplitude is lower than a specified threshold. The CPU transmits each particle and its attributes as vertex data to the GPU every frame. In step four above, the particles generated in step three are fed into the GPU. The GPU performs culling, splitting, and tracking of the generated particles. After tracking is completed, the ocean block where the particles are located is determined, and the corresponding height map of the ocean block is applied. Specifically: 4.1 Particle culling: According to the physical information carried by the particle, the diffusion radius of the wave generated by the particle can be calculated The diffusion radius is intersected with the viewing frustum used by the visualization system, and particles that generate invisible waves are culled. The high-efficiency parallel computing capability of the magnification shader is used to complete the culling. 4.2 Particle Splitting: After culling, the amplification shader splits each remaining particle into... Several particles located on the diffusion arc, these particles carry amplitude information of the original particle. And ultimately, it actually acts on the height field; 4.3 Particle Tracking: After amplifying the shader, the mesh shader is called to perform parallel particle tracking calculations. The particle information passed to the mesh shader includes the velocity vector, amplitude, and diffusion radius of the particles after splitting. Including the initial particle position information, the tracking process uses a hardware-based ray tracing architecture based on modern graphics hardware. Simultaneously, it utilizes the top-level acceleration structure generated in step one for data acceleration, enabling rapid determination of the particle's intersection with the ocean boundary. Upon detecting boundary collision information, the particle undergoes reflection processing until its trajectory reaches the specified point. That is, the final position of the particle; 4.4 The data generated by the mesh shader is passed to the pixel shader to form a local height field; 4.5 The obtained local height fields are fused using a filtering algorithm to obtain the overall height field for ocean simulation.

2. The ocean simulation method based on particle tracking according to claim 1, characterized in that: In step 4.1, the diffusion radius is calculated specifically as follows: based on the velocity information carried by the particles... The generation time can be calculated using the following formula: Particles, in time diffusion radius at time ,Right now: ; Within the radius The arc with a diffusion angle of α contains split particles. When the arc is completely outside the view frustum, the particles are considered invisible and no further calculations are needed. In step 4.2, based on the particles removed in step 4.1, the water wave is assumed to have a radius after splitting. The particles are tangent to each other along the arc. Based on this assumption, the angle between the center of two adjacent particles and the original particle's generation position can be calculated. According to the diffusion angle Angle and radius of the arc This allows us to obtain the position and velocity of the desired generated particles, after splitting into After each particle, the amplitude of each particle is equal to the original amplitude. This process uses an amplification shader and directly passes the generated result to the mesh shader, without CPU intervention. (Angle) The calculation formula is: ; Number of splits The calculation formula is: 。 3. The ocean simulation method based on particle tracking according to claim 1, characterized in that: In step one, generating the acceleration structure specifically involves: 1.1 Constructing Boundary Conditions: Describing the ocean boundary conditions in a 2D plane using vertex data; 1.2 Generate acceleration structure: Use vertex data to generate several low-level acceleration structures, and use the collection of low-level acceleration structures to generate a top-level acceleration structure.

4. The ocean simulation method based on particle tracking according to claim 3, characterized in that: In step 1.1, the vertex data is projected onto the xy plane, while vertex data that is above sea level and will not cause wave reflection is discarded. have: 。 5. The ocean simulation method based on particle tracking according to claim 3, characterized in that: In step 1.2, vertex buffer data VertexBuffer and index buffer data IndexBuffer are created using the projected vertex data to generate the underlying acceleration structure.

6. The ocean simulation method based on particle tracking according to claim 1, characterized in that: In step two, specifically: the ocean surface to be simulated is divided into subdivisions. Due to precision limitations, a large open ocean surface needs to be composed of several smaller subdivisions. Each subdivision has its own height map, used to record the local height field generated after being affected by particles. The generation of the height field can be represented as: ; in, The initial height of the local height field. The amplitude carried by the particle. This is a fixed wave function. The effect of a single particle on the local height field can be obtained by the following formula: 。 7. A particle-tracking-based ocean simulation system for performing a particle-tracking-based ocean simulation method as described in claim 1, comprising a data preprocessing module (1), a wave generation module (2), and a visualization module (3), characterized in that: The wave generation module (2) establishes data connections with the data preprocessing module (1) and the visualization module (3) respectively. The data preprocessing module (1) includes a scene import unit (11) and an accelerated structure generation unit (12). The wave generation module (2) includes a particle tracking unit (21) and a texture generation unit (22). The visualization module (3) includes a scene rendering unit (31), an effect generation unit (32), and an interaction support unit (33).

8. A particle-tracking-based ocean simulation system according to claim 7, characterized in that: The scene import unit (11) is used to obtain vertex data from the imported scene boundary information. The acceleration structure generation unit (12) analyzes the imported scene data and generates the bottom-level acceleration structure and top-level acceleration structure required for tracking particles. The particle tracking unit (21) is used to track the interactively generated particles on the top-level acceleration structure and obtain the final action position of the particles. The texture generation unit (22) is used to generate the height map and normal map required for drawing the ocean based on the local height field generated by the particle action after tracking. The scene rendering unit (31) is used to render the scene based on the generated texture and other scene elements. The effect generation unit (32) is used to generate the visualization effect of the open ocean for the scene. The interaction support unit (33) is used to support the user to interact with the three-dimensional visualization scene.