Twin power station collision visual simulation method and system based on physical simulation

Abstract

The application provides a kind of twin power station collision visualization simulation method and system based on physical simulation, it is related to data processing technical field, including: obtaining the three-dimensional geometric model of twin power station, space division is carried out, geometric parameter is extracted, material data is obtained to construct material correlation model, mechanics model is fused, and physical simulation model is obtained;Collision initial condition parameters are obtained and used as boundary conditions, local grid encryption is carried out, nonlinear dynamics equation of collision process is constructed, time domain solution is carried out, equivalent stress distribution and equivalent plastic strain distribution are determined, and collision result is generated;Whether plastic deformation or fracture failure occurs is judged, full collision process physical field data is generated, collision animation frame sequence is generated, stress distribution nephogram and plastic strain nephogram are generated, fracture special effect is constructed, and collision visualization scene is comprehensively generated.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a physical simulation-based visual simulation method and system for twin power plant collisions. Background Technology

[0002] As an important infrastructure for national economic and social development, the safe and stable operation of power plants is of paramount importance. During the construction and operation of power plants, component collision accidents may occur due to factors such as natural disasters and human errors, which seriously threaten the safety and reliability of power plants. Traditional power plant component collision analysis mainly relies on on-site investigation and experience judgment, which has problems such as low analysis accuracy and poor evaluation efficiency, making it difficult to meet the ever-increasing safety management needs of power plants. In existing technologies, physical simulation and visualization are disconnected, making it impossible to dynamically display the entire collision process. At the same time, there is a lack of high-fidelity digital twin models based on actual power plants, making it difficult to conduct accurate simulation analysis. Therefore, there is an urgent need for a solution to address the problems existing in the current technologies. Summary of the Invention

[0003] This invention provides a physical simulation-based visual simulation method and system for twin power plant collisions, which can at least solve some of the problems existing in the prior art.

[0004] A first aspect of this invention provides a physical simulation-based visual simulation method for twin power plant collisions, comprising: A three-dimensional geometric model of a twin power station is obtained. The twin power station is spatially partitioned using an octree spatial partitioning algorithm to generate a voxel model. Geometric parameters of the power station components in the twin power station are extracted based on the topological relationship of the voxel units. Material data of the power station components are obtained based on the entity design specifications and bill of materials of the twin power station. The geometric parameters and the material data are associated to construct a material association model. The mechanical model of the internal frame structure of the entity power station is obtained according to the structural design drawings of the entity power station. The force distribution is calculated by combining mechanical principles to obtain a mechanical simulation model. This mechanical simulation model is then fused with the material association model to obtain a physical simulation model. The physical simulation model is added to a pre-set physical simulation engine to generate a basic collision simulation model. The initial collision condition parameters of the two power station components that collide are obtained and added to the basic collision simulation model as boundary conditions. The local mesh of the power station components is refined using an adaptive mesh subdivision algorithm. Combined with the material data corresponding to the power station components, a material constitutive model is defined using a pre-set finite element model. The nonlinear dynamic equation of the collision process is constructed based on the nonlinear properties of the power station components. The nonlinear dynamic equation is solved in the time domain using an explicit central difference algorithm. Physical field information is generated, and the strain field and stress field of the power station components are calculated using the conserved nodal mean stress method. The equivalent stress distribution and equivalent plastic strain distribution are determined. The collision results are generated based on the yield criterion and fracture failure criterion. Based on the collision results, it is determined whether the power station component has undergone plastic deformation or fracture failure. If plastic deformation occurs, the finite element model is updated using a node displacement correction algorithm. If fracture failure occurs, the failed element is removed from the finite element model using an element deletion method, and a free surface is generated. The simulation is repeated until the collision ends, generating physical field data for the entire collision process. The displacement field data in the physical field data of the entire collision process is mapped to the three-dimensional geometric model. The position and shape of the power station component are updated using a vertex displacement difference algorithm, generating a collision animation frame sequence. Based on the equivalent stress distribution and equivalent plastic strain distribution, a stress distribution cloud map and a plastic strain cloud map are generated using a volume rendering algorithm. For power station components that have fractured, fracture textures and fracture morphologies are generated using a mesh breaking algorithm based on the position and normal vector of the fracture element. Fracture effects are constructed using flying fragments. The collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects are combined to generate a collision visualization scene.

[0005] In one alternative implementation, A three-dimensional geometric model of a twin power station is obtained. The twin power station is spatially partitioned using an octree space partitioning algorithm to generate a voxel model. Geometric parameters of the power station components are extracted based on the topological relationships of the voxel units. Material data of the power station components are obtained based on the twin power station's solid design specifications and bill of materials. The geometric parameters and material data are correlated to construct a material correlation model. A mechanical model of the internal frame structure of the solid power station is obtained based on its structural design drawings. The stress distribution is calculated using mechanical principles to obtain a mechanical simulation model, which is then fused with the material correlation model to obtain a physical simulation model, including: Point cloud data corresponding to the twin power station is obtained through 3D laser scanning technology. The 3D geometric model of the twin power station is obtained through point cloud data preprocessing and registration stitching. Based on the octree space partitioning algorithm, the root node of the octree is determined according to the bounding box of the twin power station. At each node, the complexity metric is determined according to the number of point clouds. The complexity threshold is combined with a pre-set complexity threshold. If the complexity of the current node is greater than the complexity threshold, the current node is recursively divided into eight equal parts. The judgment is repeated until the complexity of all nodes is less than the complexity threshold, and leaf nodes are obtained. For each leaf node, the corresponding voxel unit is generated and combined to obtain the voxelized model. Based on the topological relationship corresponding to the voxel unit, different voxel sets corresponding to the power station components in the twin power station are identified. The outer envelope box of the component is calculated based on the voxel set belonging to the same component to obtain the size information of the current component. The shape features of the current component are extracted by fitting the surface mesh of the voxel set. The geometric center of the outer envelope box is used as the spatial position coordinate of the current component. The geometric parameters of the current component are obtained by combining the size information, shape features and spatial position coordinates. Each component in the twin power station is traversed and the corresponding geometric parameters are obtained. Based on the physical design specifications and bill of materials of the twin power stations, the material composition of the power station components in each power station is obtained. For metallic materials, a material property database is constructed by combining physical and mechanical property parameters. The property parameters of the material composition in each component are retrieved from the material property database. For non-metallic materials and new materials, mechanical performance parameters are determined by combining simulation experiments. The material data is obtained by combining the property parameters and the mechanical performance parameters. The bill of materials is traversed and a first relationship table and a second relationship table corresponding to the power station components and the geometric parameters and the material data are generated respectively. The first relationship table and the second relationship table are connected, and the material property parameters or mechanical performance parameters are retrieved and combined to obtain the material association model. Based on the structural design drawings of the actual power station, a mechanical model of the internal frame structure of the actual power station is obtained. Key load modes are determined by combining mechanical principles and load condition analysis techniques. These key load modes are divided into static loads and dynamic loads. For static loads, the static force distribution is calculated using the finite element statics analysis method. For dynamic loads, the vibration modes and frequency response characteristics of the internal frame structure are obtained using the finite element dynamics analysis method. Based on these vibration modes and frequency response characteristics, and combined with transient mechanical analysis methods, the dynamic force distribution is obtained by solving the dynamic equations using the direct integration method. The static and dynamic force distributions are then combined to obtain a mechanical simulation model. Finally, the static and dynamic force distributions in the mechanical simulation model are mapped to the material correlation model to obtain the physical simulation model.

[0006] In one alternative implementation, Based on the aforementioned vibration modes and frequency response characteristics, and combined with transient mechanics analysis methods, the dynamic force distribution is obtained by solving the dynamic equations using the direct integration method, as shown in the following formula: ; in, F ( t n+1 ) indicates at time step n +1 external force load, [M] represents the mass matrix, { u n+1} represents a time step n The displacement vector is +1, and [C] represents the damping matrix. Indicates time step n velocity vector, Δ t Indicates the time increment. Indicates time step n acceleration vector, Indicates time step n The acceleration vector is +1, and [K] represents the stiffness matrix.

[0007] In one alternative implementation, The physical simulation model is added to a pre-set physical simulation engine to generate a basic collision simulation model. The initial collision condition parameters of the two colliding power station components are obtained and added as boundary conditions to the basic collision simulation model. The power station components are locally meshed using an adaptive mesh refinement algorithm. Combined with the material data corresponding to the power station components, a material constitutive model is defined using a pre-set finite element model. Based on the nonlinear properties of the power station components, a nonlinear dynamic equation for the collision process is constructed. The nonlinear dynamic equation is solved in the time domain using an explicit central difference algorithm, generating physical field information. The strain and stress fields of the power station components are calculated using the conserved nodal mean stress method to determine the equivalent stress distribution and equivalent plastic strain distribution. The collision results are generated based on the yield criterion and fracture failure criterion, including: A physical simulation engine is built based on the explicit finite element algorithm. The physical simulation model is added to the physical simulation engine and the basic collision simulation model is generated. Kinematic analysis was performed on the two power station components that collided. Trajectory tracking was performed based on real-time monitoring data to determine the position of the center of mass. The angular velocity and contact position at the time of collision were determined by combining the geometry and attitude of each component. The initial collision condition parameters were obtained by combining the center of mass position, angular velocity and contact position and added as boundary conditions to the basic collision simulation model. For the power plant components that have collided, an initial mesh is set using an adaptive mesh subdivision algorithm. Stress analysis is performed on the initial mesh to determine the equivalent stress of each element and mark the elements whose equivalent stress is greater than a preset mesh stress threshold. The marked areas are taken as the areas to be subdivided. The areas to be subdivided are recursively subdivided using an adaptive spatial partitioning algorithm until the preset mesh density requirement is met. The subdivided mesh is smoothed and coupled with the initial mesh to obtain the second finite element model. Based on the second finite element model and combined with the material data corresponding to the power station components, a bilinear elastoplastic constitutive model is constructed for isotropic materials based on the elastic modulus and yield stress through a pre-set finite element model. For anisotropic materials, the yield function is determined based on the anisotropic parameters corresponding to the current material. The nonlinear dynamic equation corresponding to the collision process is constructed by combining the nonlinear properties of the power station components. The nonlinear properties include large deformation, material nonlinearity and contact nonlinearity. The nonlinear dynamic equations are solved using the explicit central difference algorithm, and the stress, internal force, and contact state of each component are updated. Physical field information is generated, and the strain field and stress field of the power station component are calculated using the conserved nodal mean stress method. Based on the strain field and stress field, the equivalent stress distribution and equivalent plastic strain distribution are determined. According to the material properties and current state of the power station component, combined with the corresponding yield criterion and fracture failure criterion, the collision result is generated and output.

[0008] In one alternative implementation, For anisotropic materials, the yield function is determined based on the anisotropic parameters corresponding to the current material, as shown in the following formula: ; in, F Indicates the material is in y shaft and z Differential resistance to yielding between shafts G Indicates the material is in z shaft and x Differential resistance to yielding between shafts H Indicates the material is in x shaft and y Differential resistance to yielding between shafts s 11 Indicates the material is in x Components of normal stress in the axial direction s 22 Indicates the material is in y Components of normal stress in the axial direction s 33 Indicates the material is in z Components of normal stress in the axial directionm This indicates the Horsford Index. s y This represents the yield stress.

[0009] In one alternative implementation, Based on the collision results, it is determined whether the power station component has undergone plastic deformation or fracture failure. If plastic deformation occurs, the finite element model is updated using a node displacement correction algorithm. If fracture failure occurs, the failed element is removed from the finite element model using an element deletion method, and a free surface is generated. The simulation is repeated until the collision ends, generating full collision process physical field data. The displacement field data in the full collision process physical field data is mapped to the three-dimensional geometric model. The position and shape of the power station component are updated using a vertex displacement difference algorithm, generating a collision animation frame sequence. Based on the equivalent stress distribution and equivalent plastic strain distribution, a stress distribution cloud map and a plastic strain cloud map are generated using a volume rendering algorithm. For power station components that have fractured, fracture textures and fracture morphologies are generated using a mesh breaking algorithm based on the position and normal vector of the fracture element. Fracture effects are constructed using flying fragments. The collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects are combined to generate a collision visualization scene, including: Based on the collision results, it is determined whether the power station component has undergone plastic deformation or fracture failure. If plastic deformation occurs, the displacement field data of the current component is extracted based on the node displacement correction algorithm. For each node, the node coordinates are updated according to the corresponding displacement vector. A finite element mesh is generated based on the updated node coordinates, and the material properties and plastic strain of the current component are updated. The update is repeated until the finite element model sequence corresponding to the current component is obtained. If the collision result indicates that the current power plant component has fractured, then the failure unit is identified based on the fracture failure criterion and a set of failure units is generated. For each failure unit, the corresponding node set is extracted and the node set is removed. Nodes are inserted at the boundary of the fracture failure region and connected to the finite element mesh before fracture to obtain the free surface. The finite element mesh is regenerated based on the free surface and the material properties and plastic strain of the current component are updated. The update is repeated until the finite element model sequence corresponding to the current component is obtained. The physical field data of the full collision process is extracted based on the finite element model sequence. The displacement field in the physical field data of the full collision process is mapped to the three-dimensional geometric model to generate the vertex displacement field. The vertex coordinates of the three-dimensional geometric model are updated by the vertex displacement difference algorithm and the corresponding normal vectors and texture coordinates are calculated. The position and shape of the power station component are updated by combining the physical field data of the full collision process to obtain the collision animation frame sequence. For the power plant structure, based on the collision animation frame sequence and the equivalent stress distribution and equivalent plastic strain distribution, a color mapping table for each voxel is determined by a volume rendering algorithm. The stress and strain values ​​are searched in the color mapping table. A ray casting algorithm is used to emit rays to each pixel and determine the intersection points with different voxels. The color corresponding to each pixel is determined. The operation is repeated to obtain the stress distribution cloud map and plastic strain cloud map. For power plant components that have experienced fracture failure, a mesh breaking algorithm is used to generate random crack distributions on the boundary surfaces corresponding to the fracture elements based on the fracture direction and the material data, according to the location and normal vector of the fracture elements. A polygonal mesh is generated with each crack line in the random crack distribution as the starting point. The polygonal mesh is triangulated and a normal map and height map are added to generate an initial fracture morphology, and random perturbations are added to obtain the fracture texture and fracture morphology. Fragment particles are randomly generated within the fracture elements. The kinematic parameters of each fragment particle are initialized, and the scattering trajectory of each fragment particle is determined by explicit integration. Fracture effects are constructed based on the scattering trajectories and fragment particles. A collision visualization scene is generated by combining the collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects.

[0010] In one alternative implementation, By emitting rays to each pixel using a ray casting algorithm and determining the intersection points with different voxels, the color corresponding to each pixel is determined by the following formula: ; in, CP ( u , v ) represents a pixel ( u , v ) color, CV ( x , y , z ) represents voxels ( x , y , z The color value of ) S ( x , y , z ) represents voxels ( x , y , z The scattering coefficient of ) t ( x , y , z ) represents voxels ( x , y , z The absorption coefficient of ) lThe path of light propagation represents the distance light travels within a voxel. dλ A ray element is used to calculate the path of a ray. l The points on the scale.

[0011] A second aspect of the present invention provides a physical simulation-based visual simulation system for twin power plant collisions, comprising: The first unit is used to obtain a three-dimensional geometric model of the twin power station. The twin power station is spatially partitioned using an octree spatial partitioning algorithm to generate a voxel model. Based on the topological relationship of the voxel units, the geometric parameters of the power station components in the twin power station are extracted. The material data of the power station components are obtained based on the entity design specifications and bill of materials of the twin power station. The geometric parameters and the material data are associated to construct a material association model. The mechanical model of the internal frame structure of the entity power station is obtained according to the structural design drawings of the entity power station. The force distribution is calculated by combining mechanical principles to obtain a mechanical simulation model. This mechanical simulation model is then fused with the material association model to obtain a physical simulation model. The second unit is used to add the physical simulation model to a pre-set physical simulation engine to generate a basic collision simulation model, obtain the initial collision condition parameters of the two power station components that collide and add them as boundary conditions to the basic collision simulation model, refine the local mesh of the power station components through an adaptive mesh subdivision algorithm, define the material constitutive model through a pre-set finite element model in combination with the material data corresponding to the power station components, construct the nonlinear dynamic equation of the collision process based on the nonlinear properties of the power station components, solve the nonlinear dynamic equation in the time domain using an explicit central difference algorithm, generate physical field information, calculate the strain field and stress field of the power station components through the conserved nodal average stress method, determine the equivalent stress distribution and equivalent plastic strain distribution, and generate the collision result based on the yield criterion and fracture failure criterion. The third unit is used to determine whether the power station component has undergone plastic deformation or fracture failure based on the collision results. If plastic deformation occurs, the finite element model is updated using a node displacement correction algorithm. If fracture failure occurs, the failed element is removed from the finite element model using an element deletion method, and a free surface is generated. The simulation is repeated until the collision ends, generating physical field data for the entire collision process. The displacement field data in the physical field data of the entire collision process is mapped to the three-dimensional geometric model. The position and shape of the power station component are updated using a vertex displacement difference algorithm, generating a collision animation frame sequence. Based on the equivalent stress distribution and equivalent plastic strain distribution, a stress distribution cloud map and a plastic strain cloud map are generated using a volume rendering algorithm. For power station components that have fractured, fracture textures and fracture morphologies are generated using a mesh breaking algorithm based on the position and normal vector of the fracture element. Fracture effects are constructed using flying fragments. The collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects are combined to generate a collision visualization scene.

[0012] A third aspect of the present invention, An electronic device is provided, comprising: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0013] Fourth aspect of the embodiments of the present invention, A computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0014] In this invention, a detailed physical simulation model was constructed by acquiring the three-dimensional geometric model and material data of a twin power station and combining it with the mechanical model of the physical power station. Adaptive mesh subdivision and nonlinear dynamics solving techniques were employed to accurately simulate the complex mechanical behaviors of components, such as deformation and fracture. High-precision physical field data of the entire collision process was obtained, providing a reliable basis for accurately assessing the impact of the collision. Based on the physical field data obtained from the collision simulation, various visualization expressions, such as animations, cloud maps, and fracture effects of the component collision process, were generated through visualization algorithms such as displacement mapping, volume rendering, and mesh breaking. These comprehensively, three-dimensionally, and realistically demonstrate the deformation and fracture of components under collision loads. This invention provides an immersive visualization experience of complex mechanical behaviors such as collisions and fractures, intuitively revealing the collision mechanism. Through seamless integration of geometric modeling, physical simulation, and visual rendering, it achieves a closed-loop process from real-world scenarios to virtual simulation and then to visual results, greatly improving the relevance, accuracy, and persuasiveness of collision analysis. In summary, this invention, through the deep integration of physical simulation and visualization technologies, achieves refined simulation and realistic display of the collision process, establishes a multidisciplinary collision analysis framework, greatly improves the efficiency of power plant safety assessment and accident investigation, provides strong technical support for the full life cycle management of power plants, and has good versatility and scalability. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the physical simulation-based visual simulation method for twin power plant collisions according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of the physical simulation-based twin power plant collision visualization simulation system according to an embodiment of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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.

[0017] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0018] Figure 1 This is a flowchart illustrating the physical simulation-based visual simulation method for twin power plant collisions according to an embodiment of the present invention. Figure 1 As shown, the method includes: S1. Obtain the three-dimensional geometric model of the twin power station, divide the twin power station into spaces using an octree space partitioning algorithm, generate a voxel model, and extract the geometric parameters of the power station components in the twin power station based on the topological relationship of the voxel units. Obtain the material data of the power station components based on the entity design specifications and bill of materials of the twin power station, associate the geometric parameters and the material data to construct a material association model, obtain the mechanical model of the internal frame structure of the entity power station according to the structural design drawings of the entity power station, calculate the force distribution in combination with mechanical principles, obtain a mechanical simulation model, and fuse it with the material association model to obtain a physical simulation model. The octree is a data structure used to segment three-dimensional space, recursively dividing the space into eight sub-regions to form a tree structure. The voxel is an abbreviation for "volume pixel," the smallest unit in three-dimensional space. The voxelization model is the process of decomposing a three-dimensional model into a voxel mesh, where each voxel represents a volume portion of the model. The voxel unit is the basic unit in the voxelization model, similar to a pixel in a two-dimensional image; each voxel unit represents a small cube in space. The solid design specification refers to the standards and rules used to create and describe three-dimensional models. The material correlation model considers the physical properties of materials, such as elastic modulus and Poisson's ratio, and is used to analyze the mechanical behavior of materials under different conditions. The mechanical principles are the fundamental theories for studying the behavior and design of mechanical systems, including mechanics, dynamics, and kinematics. The mechanical simulation model is a computational model used to simulate and analyze the behavior of objects under various loads and conditions.

[0019] In one alternative implementation, A three-dimensional geometric model of a twin power station is obtained. The twin power station is spatially partitioned using an octree space partitioning algorithm to generate a voxel model. Geometric parameters of the power station components are extracted based on the topological relationships of the voxel units. Material data of the power station components are obtained based on the twin power station's solid design specifications and bill of materials. The geometric parameters and material data are correlated to construct a material correlation model. A mechanical model of the internal frame structure of the solid power station is obtained based on its structural design drawings. The stress distribution is calculated using mechanical principles to obtain a mechanical simulation model, which is then fused with the material correlation model to obtain a physical simulation model, including: Point cloud data corresponding to the twin power station is obtained through 3D laser scanning technology. The 3D geometric model of the twin power station is obtained through point cloud data preprocessing and registration stitching. Based on the octree space partitioning algorithm, the root node of the octree is determined according to the bounding box of the twin power station. At each node, the complexity metric is determined according to the number of point clouds. The complexity threshold is combined with a pre-set complexity threshold. If the complexity of the current node is greater than the complexity threshold, the current node is recursively divided into eight equal parts. The judgment is repeated until the complexity of all nodes is less than the complexity threshold, and leaf nodes are obtained. For each leaf node, the corresponding voxel unit is generated and combined to obtain the voxelized model. Based on the topological relationship corresponding to the voxel unit, different voxel sets corresponding to the power station components in the twin power station are identified. The outer envelope box of the component is calculated based on the voxel set belonging to the same component to obtain the size information of the current component. The shape features of the current component are extracted by fitting the surface mesh of the voxel set. The geometric center of the outer envelope box is used as the spatial position coordinate of the current component. The geometric parameters of the current component are obtained by combining the size information, shape features and spatial position coordinates. Each component in the twin power station is traversed and the corresponding geometric parameters are obtained. Based on the physical design specifications and bill of materials of the twin power stations, the material composition of the power station components in each power station is obtained. For metallic materials, a material property database is constructed by combining physical and mechanical property parameters. The property parameters of the material composition in each component are retrieved from the material property database. For non-metallic materials and new materials, mechanical performance parameters are determined by combining simulation experiments. The material data is obtained by combining the property parameters and the mechanical performance parameters. The bill of materials is traversed and a first relationship table and a second relationship table corresponding to the power station components and the geometric parameters and the material data are generated respectively. The first relationship table and the second relationship table are connected, and the material property parameters or mechanical performance parameters are retrieved and combined to obtain the material association model. Based on the structural design drawings of the actual power station, a mechanical model of the internal frame structure of the actual power station is obtained. Key load modes are determined by combining mechanical principles and load condition analysis techniques. These key load modes are divided into static loads and dynamic loads. For static loads, the static force distribution is calculated using the finite element statics analysis method. For dynamic loads, the vibration modes and frequency response characteristics of the internal frame structure are obtained using the finite element dynamics analysis method. Based on these vibration modes and frequency response characteristics, and combined with transient mechanical analysis methods, the dynamic force distribution is obtained by solving the dynamic equations using the direct integration method. The static and dynamic force distributions are then combined to obtain a mechanical simulation model. Finally, the static and dynamic force distributions in the mechanical simulation model are mapped to the material correlation model to obtain the physical simulation model.

[0020] The point cloud data preprocessing includes cleaning, denoising, and downsampling to improve data quality and analysis accuracy. Point cloud registration is the process of aligning multiple point cloud datasets to the same coordinate system. The bounding box is a minimal rectangle or cube used to enclose and represent the spatial extent of a 3D object. The complexity metric is used to evaluate the complexity of the model or algorithm, including time and space complexity. The outer envelope is a minimal convex polyhedron containing all the point cloud data. The surface mesh is a grid composed of polygons (such as triangles) used to represent the surface of the 3D object. The physical and mechanical property parameters describe the physical and mechanical properties of the material or structure, such as density. The load condition analysis technique is used to analyze and predict the behavior and performance of structures under different load conditions, including elastic modulus and hardness. The finite element static analysis method is used to calculate and analyze the stress, strain, and displacement of structures under static loads. The finite element dynamic analysis method is used to analyze the behavior of structures under dynamic loads, including vibration and impact. The dynamic modes describe the characteristics of different deformation modes of structures during vibration. The frequency response characteristics describe the response behavior of structures at different frequencies. The transient mechanical analysis method is used to analyze and predict the response of structures under transient loads (such as impacts). The direct integration method is a numerical method used to solve the equations of dynamic systems and directly calculate the system response.

[0021] A 3D laser scanner is used to perform a comprehensive scan of the twin power station to acquire point cloud data. The acquired point cloud data is preprocessed, including denoising, filtering, and downsampling. Point cloud registration algorithms (such as ICP algorithm) are used to stitch together point cloud data from different perspectives to obtain a complete 3D point cloud model. An octree space partitioning algorithm is used to voxelize the point cloud model to determine the bounding box of the point cloud model, which serves as the root node of the octree. The complexity metric of each node is calculated based on the number of points and compared with a preset complexity threshold. If it is greater than the threshold, the current node is divided into eight equal parts. The child nodes are recursively partitioned until the complexity of all nodes is less than the threshold. Voxel units corresponding to the leaf nodes are generated and combined to obtain a complete voxelized model. Based on the topological relationship of voxel elements in the voxelization model, the voxel sets corresponding to different power plant components are identified. For each power plant component, the outer envelope box of the component voxel set is calculated to obtain the component's size information. The surface mesh of the component voxel set is fitted to extract the component's shape features. The geometric center of the component's outer envelope box is used as the component's spatial position coordinates. The geometric parameters of the component are obtained by combining the size information, shape features, and spatial position coordinates. All components in the twin power plant are traversed to obtain their respective geometric parameters. Obtain the physical design specifications and bill of materials for the twin power station, determine the material composition of each power station component, and for metallic materials, construct a material property database by combining physical and mechanical property parameters. Retrieve the property parameters of the material composition of each component from the material property database. For non-metallic materials and new materials, determine the mechanical performance parameters of the materials through simulation experiments. Combine the property parameters and mechanical performance parameters to obtain complete material data. Traverse the bill of materials to generate a first relationship table between power station components and geometric parameters, and a second relationship table between power station components and material data. Connect the first and second relationship tables, retrieve material property parameters or mechanical performance parameters, and combine them to obtain a material association model. Based on the structural design drawings of the actual power station, a mechanical model of the internal frame structure of the actual power station is obtained. Combining mechanical principles and load condition analysis technology, key load modes are determined, which are divided into static loads and dynamic loads. For static loads, the static force distribution is calculated using the finite element static analysis method. For dynamic loads, the vibration modes and frequency response characteristics of the internal frame structure are obtained using the finite element dynamic analysis method. Based on the vibration modes and frequency response characteristics, combined with the transient mechanical analysis method, the dynamic equations are solved using the direct integration method to obtain the dynamic force distribution. Combining the static and dynamic force distributions, a mechanical simulation model is obtained. The static and dynamic force distributions in the mechanical simulation model are mapped to the material correlation model to obtain the final physical simulation model. For example, taking a simplified twin power plant as an example, assuming it consists of three main components: a main plant, a cooling tower, and transmission lines, a 3D laser scanner is used to perform a comprehensive scan of the power plant to acquire point cloud data. After preprocessing, the ICP algorithm is used to stitch the point cloud into a complete 3D model. The octree algorithm is then used to voxelize the point cloud model, resulting in a voxelized model. Based on the voxelized model, the voxel sets corresponding to the main plant, cooling tower, and transmission lines are identified. For each component, its outer envelope size is calculated, shape features are extracted by fitting a surface mesh, and the geometric center is used as the spatial coordinates to finally obtain the geometric parameters of the three components. The design specifications and bill of materials for the power plant are then obtained. Assuming the main plant is constructed of reinforced concrete, the cooling tower is constructed of steel and concrete, and the power transmission lines are constructed of steel and aluminum alloy, the property parameters of these materials are retrieved from the material property database, and their mechanical performance parameters are determined through simulation experiments. A relationship table between components, geometric parameters, and material data is generated, and a material correlation model is obtained by combining them. Based on the structural design drawings of the power plant, the mechanical model of its internal frame structure is obtained, and static loads (such as self-weight) and dynamic loads (such as wind loads) are determined. The static force distribution and dynamic force distribution are calculated using the finite element method to obtain the mechanical simulation model. The force distribution in the mechanical simulation model is mapped to the material correlation model to obtain the physical simulation model.

[0022] In this embodiment, the octree space partitioning algorithm can adaptively generate voxelized models based on model complexity, ensuring geometric details while improving data processing efficiency. Through operations such as calculating the outer envelope box and fitting surface meshes, it can automatically extract geometric parameters such as the size, shape, and position of components, greatly reducing the workload of manual modeling and improving the efficiency and accuracy of parametric modeling. It generates a relationship table between components and geometric and material data and connects and combines them, enabling seamless integration of data from different fields and formats, providing consistent data support. It maps mechanical simulation results with material correlation models, achieving deep coupling of multiple physics fields (geometry, materials, and mechanics), greatly improving the accuracy and reliability of simulation. In summary, this embodiment, through systematic data acquisition, processing, integration, and analysis, constructs a high-fidelity, multi-dimensional, full-lifecycle digital twin model of a power plant, providing powerful digital tools and decision support for various stages of power plant design optimization, construction management, operation, and maintenance. It is expected to greatly improve the efficiency, quality, and safety of power plant projects and promote the development of power plants towards intelligence and digitalization.

[0023] In one alternative implementation, Based on the aforementioned vibration modes and frequency response characteristics, and combined with transient mechanics analysis methods, the dynamic force distribution is obtained by solving the dynamic equations using the direct integration method, as shown in the following formula: ; in, F ( t n+1 ) indicates at time step n +1 external force load, [M] represents the mass matrix, { u n+1} represents a time step n The displacement vector is +1, and [C] represents the damping matrix. Indicates time step n velocity vector, Δ t Indicates the time increment. Indicates time step n acceleration vector, Indicates time step n The acceleration vector is +1, and [K] represents the stiffness matrix.

[0024] In this embodiment, by reasonably setting the mass matrix, damping matrix, and stiffness matrix, the dynamic behavior of the structure can be accurately described, improving the accuracy of dynamic response calculation. By discretizing the time domain, the continuous dynamic problem is transformed into a discrete algebraic equation system, which has the characteristics of computational stability and high efficiency, and is suitable for large-scale, nonlinear dynamic problems. By applying the corresponding external load at each time step, the time-varying characteristics of dynamic loads, such as wind loads and seismic loads, can be simulated. By solving the dynamic equations, the displacement vector, velocity vector, and acceleration vector of the structure at each time step can be obtained, which can be used to evaluate the deformation, vibration, and stress state of the structure under dynamic loads, providing a basis for the analysis of the structure's seismic and wind resistance dynamic performance. In summary, this embodiment fully considers the dynamic characteristics of the structure and the time-varying nature of external loads, and obtains the dynamic response of the structure through numerical solution, providing important data support for subsequent dynamic performance analysis and fatigue assessment. By coupling with static analysis, the comprehensive consideration of static and dynamic loads is realized, improving the accuracy of simulation results, which is of great significance for analyzing and optimizing the seismic and wind resistance dynamic performance of power station structures.

[0025] S2. The physical simulation model is added to the pre-set physical simulation engine to generate a basic collision simulation model. The initial collision condition parameters of the two power station components that collide are obtained and added to the basic collision simulation model as boundary conditions. The local mesh of the power station components is refined by an adaptive mesh subdivision algorithm. Combined with the material data corresponding to the power station components, a material constitutive model is defined by a pre-set finite element model. The nonlinear dynamic equation of the collision process is constructed based on the nonlinear properties of the power station components. The nonlinear dynamic equation is solved in the time domain by combining the explicit central difference algorithm. Physical field information is generated and the strain field and stress field of the power station components are calculated by the conserved nodal average stress method. The equivalent stress distribution and equivalent plastic strain distribution are determined. The collision result is generated based on the yield criterion and fracture failure criterion. The physics simulation engine is used to simulate real-world physical phenomena, such as mechanics, collisions, and fluid dynamics. The basic collision simulation model is used to simulate collisions between objects, including collision detection and response. The initial collision condition parameters define the initial state of the collision simulation, including the object's position, velocity, direction, and collision angle. The adaptive mesh refinement algorithm dynamically adjusts the mesh fineness according to computational needs to improve simulation accuracy. Local mesh refinement refers to increasing the mesh density in specific regions to more accurately simulate complex physical phenomena. The finite element model approximates the solution of physical problems by dividing complex structures into small discrete elements. The material constitutive model describes the material's response under various loads, including stress-strain relationships. The nonlinear properties refer to the relationship between stress and strain in materials or structures. The system is not linear and may include effects such as plasticity and creep. The nonlinear dynamic equations are used to describe the dynamic behavior of nonlinear mechanical effects in the system. The explicit central difference algorithm is a numerical method for solving time-stepping dynamic problems based on the central difference formula. The time-domain solution is a method for analyzing the system behavior that changes over time, including time-stepping and dynamic response analysis. The conserved nodal mean stress method is used to calculate and average nodal stresses in finite element analysis to ensure stress conservation. The equivalent stress distribution and equivalent plastic strain distribution are used to describe the internal state of the structure under stress and deformation. The yield criterion is a standard for predicting the initiation of plastic deformation in materials under stress. The fracture failure criterion is used to predict the fracture behavior of materials under stress and crack influence, such as the stress intensity factor criterion.

[0026] In one alternative implementation, The physical simulation model is added to a pre-set physical simulation engine to generate a basic collision simulation model. The initial collision condition parameters of the two colliding power station components are obtained and added as boundary conditions to the basic collision simulation model. The power station components are locally meshed using an adaptive mesh refinement algorithm. Combined with the material data corresponding to the power station components, a material constitutive model is defined using a pre-set finite element model. Based on the nonlinear properties of the power station components, a nonlinear dynamic equation for the collision process is constructed. The nonlinear dynamic equation is solved in the time domain using an explicit central difference algorithm, generating physical field information. The strain and stress fields of the power station components are calculated using the conserved nodal mean stress method to determine the equivalent stress distribution and equivalent plastic strain distribution. The collision results are generated based on the yield criterion and fracture failure criterion, including: A physical simulation engine is built based on the explicit finite element algorithm. The physical simulation model is added to the physical simulation engine and the basic collision simulation model is generated. Kinematic analysis was performed on the two power station components that collided. Trajectory tracking was performed based on real-time monitoring data to determine the position of the center of mass. The angular velocity and contact position at the time of collision were determined by combining the geometry and attitude of each component. The initial collision condition parameters were obtained by combining the center of mass position, angular velocity and contact position and added as boundary conditions to the basic collision simulation model. For the power plant components that have collided, an initial mesh is set using an adaptive mesh subdivision algorithm. Stress analysis is performed on the initial mesh to determine the equivalent stress of each element and mark the elements whose equivalent stress is greater than a preset mesh stress threshold. The marked areas are taken as the areas to be subdivided. The areas to be subdivided are recursively subdivided using an adaptive spatial partitioning algorithm until the preset mesh density requirement is met. The subdivided mesh is smoothed and coupled with the initial mesh to obtain the second finite element model. Based on the second finite element model and combined with the material data corresponding to the power station components, a bilinear elastoplastic constitutive model is constructed for isotropic materials based on the elastic modulus and yield stress through a pre-set finite element model. For anisotropic materials, the yield function is determined based on the anisotropic parameters corresponding to the current material. The nonlinear dynamic equation corresponding to the collision process is constructed by combining the nonlinear properties of the power station components. The nonlinear properties include large deformation, material nonlinearity and contact nonlinearity. The nonlinear dynamic equations are solved using the explicit central difference algorithm, and the stress, internal force, and contact state of each component are updated. Physical field information is generated, and the strain field and stress field of the power station component are calculated using the conserved nodal mean stress method. Based on the strain field and stress field, the equivalent stress distribution and equivalent plastic strain distribution are determined. According to the material properties and current state of the power station component, combined with the corresponding yield criterion and fracture failure criterion, the collision result is generated and output.

[0027] The mesh density requirement refers to the fineness of the model mesh in finite element analysis or other numerical simulations. The smoothing process refers to improving the surface quality of the mesh model by smoothing it out, removing sharp corners and irregularities, and making the surface look smoother and more natural. The isotropic material refers to a material that has the same physical properties in all directions. The anisotropic material has different physical properties in different directions. The elastic modulus is a measure of a material's ability to resist elastic deformation. The yield stress is the stress value at which a material begins to undergo plastic deformation.

[0028] Choose a suitable explicit finite element algorithm, such as the explicit central difference algorithm, to build the numerical solver of the physical simulation engine. Input the geometric model, material properties, boundary conditions and other information of the power station components into the physical simulation engine to generate the corresponding finite element model. Set the collision detection parameters in the physical simulation engine, such as the contact surface search algorithm and contact stiffness, and build the basic collision simulation model. The displacement, velocity, and other motion state data of two colliding power station components are collected in real time by sensors. Based on the kinematic model and numerical integration method, the collected data is tracked to obtain the position of the center of mass of the two components at the moment of collision. According to the geometric shape and attitude data of the components, the angular velocity and contact position at the time of collision are calculated by analytical or numerical methods. The parameters such as the center of mass position, angular velocity, and contact position are added to the basic collision simulation model as the initial conditions of the collision. According to the geometric characteristics and material properties of the components, an initial finite element mesh is generated by an adaptive mesh generation algorithm. Stress analysis is performed on the initial mesh to calculate the equivalent stress of each element. Elements with equivalent stress greater than a preset threshold are marked as regions that need to be subdivided. Adaptive spatial partitioning algorithms such as octree and quadtree are used to recursively subdivide the regions that need to be subdivided until the preset mesh density requirement is met. The subdivided mesh is smoothed to eliminate mesh distortion and improve mesh quality. The smoothed subdivided mesh is coupled with the initial mesh to form a new finite element model (second finite element model). Based on the material type (isotropic / anisotropic) of the power plant components, the corresponding constitutive model is selected. For isotropic materials, a bilinear elastoplastic constitutive relation is constructed using the material's elastic modulus and yield stress. For anisotropic materials, an anisotropic yield function is established based on anisotropic parameters (such as the six parameters of the Hill yield criterion). Considering factors such as large deformation of the power plant components, material nonlinearity, and contact nonlinearity, a nonlinear dynamic equation is established based on the second finite element model. The nonlinear dynamic equation is discretized and solved in the time domain using an explicit central difference algorithm. In each time increment step, the node coordinates and node velocities are updated, and the element stress and nodal internal forces are calculated. The contact point is found using a contact search algorithm, and the contact force is calculated. The strain field and stress field of the component are obtained by extrapolation based on the nodal stress using the conserved nodal mean stress method. The distribution of equivalent stress and equivalent plastic strain is calculated. Combining the material's yield criterion and fracture failure criterion (such as the maximum tensile stress criterion), it is determined whether the component has yielded or fractured, and the collision result data is output.

[0029] For example, suppose a steam generator collides with in-reactor components in a nuclear power plant. An explicit dynamic finite element method (FEM) solver is built using software, establishing a finite element model containing the steam generator and in-reactor components. Collision detection parameters are set to form a basic collision simulation model. High-speed cameras and inertial measurement units collect motion state data of the two components, calculating the centroid position, angular velocity, and collision point position at the time of collision. These are applied to the finite element model as initial conditions for the collision. The mesh of the collision region is refined, and an octree algorithm is used for triple mesh subdivision of stress concentration areas. The mesh quality is improved using the Laplace smoothing method to form a second finite element model. For the steam generator, its elastoplastic behavior is considered, and a bilinear elastoplastic constitutive model is adopted. The in-reactor components are made of alloys, and their corresponding anisotropy is considered, selecting the Hill-48 yield criterion. Simultaneously, the large deformation, material nonlinearity, and contact nonlinearity of the two components during the collision process are considered, establishing nonlinear dynamic equations. An explicit central difference algorithm is used to solve the dynamic equations with a time step of 1×10⁻⁶ s to obtain the stress and strain fields of the entire collision process. Von... The Mises criterion and the maximum tensile stress criterion are used to assess the yielding and fracture conditions of the component and generate an analysis report.

[0030] In this embodiment, nonlinear factors during the collision process, such as large deformation, material nonlinearity, and contact nonlinearity, are fully considered, making the simulation results closer to reality. Mesh refinement can better capture the gradient changes of stress and strain, avoid the occurrence of stress singularities, and improve the convergence and stability of the calculation. By adopting different material constitutive models (such as bilinear elastoplastic models, anisotropic yield criteria, etc.), the mechanical behavior of different materials during the collision process can be accurately described, improving the universality of the simulation results. By explicitly solving the nonlinear dynamic equations, various physical field information such as stress field, strain field, and displacement field of power plant components during the collision process can be obtained. This not only intuitively shows the impact range and damage degree of the collision, but also provides important data support for subsequent damage assessment, life prediction, and other analyses. By introducing fracture failure criteria, the crack initiation and propagation of components can be predicted, and the integrity of components can be quantitatively evaluated. In summary, this embodiment generates rich physical field information to evaluate the impact of collisions on components, providing important technical support for accident emergency decision-making and structural optimization design, and realizing high-fidelity dynamic simulation of the collision process of power plant components.

[0031] In one alternative implementation, For anisotropic materials, the yield function is determined based on the anisotropic parameters corresponding to the current material, as shown in the following formula: ; in, F Indicates the material is in y shaft andz Differential resistance to yielding between shafts G Indicates the material is in z shaft and x Differential resistance to yielding between shafts H Indicates the material is in x shaft and y Differential resistance to yielding between shafts s 11 Indicates the material is in x Components of normal stress in the axial direction s 22 Indicates the material is in y Components of normal stress in the axial direction s 33 Indicates the material is in z Components of normal stress in the axial direction m This indicates the Horsford Index. s y This represents the yield stress.

[0032] In this embodiment, the Horsford criterion considers the anisotropy of the material, making the yield function closer to the actual mechanical behavior of the material and improving the accuracy of yield judgment. By comparing the differences of the three normal stress components, the yielding state of the material under different stress states can be determined, improving the comprehensiveness of yield judgment. By using the yield stress as the judgment standard, the material's yielding state can be quantitatively assessed. By adjusting the Horsford exponent, the yield function can better fit the actual yielding behavior of the material, improving the applicability of yield judgment. By embedding the Horsford yield function into the finite element algorithm, accurate simulation of the material's nonlinear behavior can be achieved. The explicit yield function also facilitates structural optimization design. By optimizing the coupling between the objective function and the yield function, structural optimization based on the yield criterion can be achieved. In summary, this embodiment fully considers the anisotropic characteristics of the material and the influence of stress state, improving the accuracy and comprehensiveness of yield judgment, facilitating numerical solution and optimization design, and is widely applicable to various engineering materials. It has important application value in material nonlinear analysis, structural strength assessment, and impact resistance design.

[0033] S3. Based on the collision results, determine whether the power station component has undergone plastic deformation or fracture failure. If plastic deformation occurs, update the finite element model using a node displacement correction algorithm. If fracture failure occurs, remove the failed element from the finite element model using an element deletion method and generate a free surface. Repeat the simulation until the collision ends, generating full collision process physical field data. Map the displacement field data in the full collision process physical field data to the three-dimensional geometric model. Update the position and shape of the power station component using a vertex displacement difference algorithm, generating a collision animation frame sequence. Based on the equivalent stress distribution and equivalent plastic strain distribution, combine with a volume rendering algorithm to generate a stress distribution cloud map and a plastic strain cloud map. For power station components that have fractured, generate fracture textures and fracture morphologies based on the position and normal vector of the fracture element using a mesh breaking algorithm. Combine with splashed fragments to construct fracture effects. Integrate the collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects to generate a collision visualization scene.

[0034] The nodal displacement correction algorithm is used to adjust the nodal displacements in the finite element model. It is typically a correction step applied during the calculation process to improve the accuracy of the results or address numerical instability. The element deletion method is a technique for handling fracture or failure in the finite element model. It simulates the fracture process of materials or structures by removing failed elements from the model. The free surface refers to a surface unaffected by constraints or external forces, typically used to describe the surfaces of fluids, particles, or debris. The full collision process physics data includes all physical data collected during the collision event, such as forces, stresses, deformations, and temperature changes. The vertex displacement interpolation algorithm is used to calculate and interpolate vertex displacements in the three-dimensional mesh model. The values ​​are shifted to smooth deformation and improve accuracy. The collision animation frame sequence refers to the animation sequence recorded and displayed frame by frame in the collision simulation to show the dynamic process of the collision event. The volume rendering algorithm is used to convert voxel data (such as medical scan data) into a visualized image. The mesh breaking algorithm is used to simulate and generate the breaking effect in the mesh model. It usually decomposes the mesh into fragments or broken parts through the algorithm. The fracture texture refers to the pattern or texture used to describe and render the fracture surface. The fracture morphology describes the surface features and shape of the material or structure after fracture, including the shape, direction and roughness of the crack. The splashed fragments refer to the fragments that fly out from the surface of the object during the collision or explosion.

[0035] In one alternative implementation, Based on the collision results, it is determined whether the power station component has undergone plastic deformation or fracture failure. If plastic deformation occurs, the finite element model is updated using a node displacement correction algorithm. If fracture failure occurs, the failed element is removed from the finite element model using an element deletion method, and a free surface is generated. The simulation is repeated until the collision ends, generating full collision process physical field data. The displacement field data in the full collision process physical field data is mapped to the three-dimensional geometric model. The position and shape of the power station component are updated using a vertex displacement difference algorithm, generating a collision animation frame sequence. Based on the equivalent stress distribution and equivalent plastic strain distribution, a stress distribution cloud map and a plastic strain cloud map are generated using a volume rendering algorithm. For power station components that have fractured, fracture textures and fracture morphologies are generated using a mesh breaking algorithm based on the position and normal vector of the fracture element. Fracture effects are constructed using flying fragments. The collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects are combined to generate a collision visualization scene, including: Based on the collision results, it is determined whether the power station component has undergone plastic deformation or fracture failure. If plastic deformation occurs, the displacement field data of the current component is extracted based on the node displacement correction algorithm. For each node, the node coordinates are updated according to the corresponding displacement vector. A finite element mesh is generated based on the updated node coordinates, and the material properties and plastic strain of the current component are updated. The update is repeated until the finite element model sequence corresponding to the current component is obtained. If the collision result indicates that the current power plant component has fractured, then the failure unit is identified based on the fracture failure criterion and a set of failure units is generated. For each failure unit, the corresponding node set is extracted and the node set is removed. Nodes are inserted at the boundary of the fracture failure region and connected to the finite element mesh before fracture to obtain the free surface. The finite element mesh is regenerated based on the free surface and the material properties and plastic strain of the current component are updated. The update is repeated until the finite element model sequence corresponding to the current component is obtained. The physical field data of the full collision process is extracted based on the finite element model sequence. The displacement field in the physical field data of the full collision process is mapped to the three-dimensional geometric model to generate the vertex displacement field. The vertex coordinates of the three-dimensional geometric model are updated by the vertex displacement difference algorithm and the corresponding normal vectors and texture coordinates are calculated. The position and shape of the power station component are updated by combining the physical field data of the full collision process to obtain the collision animation frame sequence. For the power plant structure, based on the collision animation frame sequence and the equivalent stress distribution and equivalent plastic strain distribution, a color mapping table for each voxel is determined by a volume rendering algorithm. The stress and strain values ​​are searched in the color mapping table. A ray casting algorithm is used to emit rays to each pixel and determine the intersection points with different voxels. The color corresponding to each pixel is determined. The operation is repeated to obtain the stress distribution cloud map and plastic strain cloud map. For power plant components that have experienced fracture failure, a mesh breaking algorithm is used to generate random crack distributions on the boundary surfaces corresponding to the fracture elements based on the fracture direction and the material data, according to the location and normal vector of the fracture elements. A polygonal mesh is generated with each crack line in the random crack distribution as the starting point. The polygonal mesh is triangulated and a normal map and height map are added to generate an initial fracture morphology, and random perturbations are added to obtain the fracture texture and fracture morphology. Fragment particles are randomly generated within the fracture elements. The kinematic parameters of each fragment particle are initialized, and the scattering trajectory of each fragment particle is determined by explicit integration. Fracture effects are constructed based on the scattering trajectories and fragment particles. A collision visualization scene is generated by combining the collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects.

[0036] The displacement vector represents the amount of displacement of an object in a certain direction, typically including three components: displacement in the X, Y, and Z directions. The finite element mesh is a grid that divides the computational domain into small discrete elements (such as triangles or tetrahedrons) used to approximate problems in finite element analysis. The set of failed elements refers to the set of mesh elements identified as failed or fractured in finite element analysis. The fracture failure region refers to the area in the material or structure where fracture or failure occurs, typically manifested as cracks or fractures. The vertex displacement field describes the spatial distribution of displacements at each vertex in the mesh model, including the actual displacement of each vertex. The color mapping table is used to map numerical data (such as temperature and stress) onto a color range for visual representation in graphical displays. The ray casting algorithm is used to sample the scene from the viewpoint along the ray direction to generate an image or perform volume rendering. The random crack distribution refers to generating cracks in the material surface or structure according to a random pattern to simulate the distribution of natural cracks. The normal map is a texture mapping technique used to simulate details and bump effects on the surface, enhancing surface details by storing normal directions. The height map is an image used to represent changes in surface height, defining terrain or surface bumps through grayscale values. The fragment particles refer to small fragments or particles generated during breakage or collision, used to simulate the scattering effect of materials. The scattering trajectory describes the path and trajectory of fragments or splashes during collision or explosion.

[0037] Extract the equivalent stress and equivalent plastic strain data from the collision results, compare the equivalent stress with the yield stress of the material to determine whether plastic deformation has occurred. If the equivalent stress is greater than the yield stress, plastic deformation has occurred. Extract the displacement field data of the component, update the node coordinates according to the displacement vector for each node, generate a new finite element mesh based on the updated node coordinates, update the material properties and plastic strain of the component, and repeat the update until the finite element model sequence of the current component is obtained. If the equivalent stress is less than the yield stress, no plastic deformation has occurred. Equivalent stress and equivalent plastic strain are substituted into the fracture failure criterion to determine whether fracture failure has occurred. If the fracture failure criterion is met, fracture failure has occurred. Based on the fracture failure criterion, failure elements are identified and a set of failure elements is generated. For each failure element, the corresponding node set is extracted and these nodes are removed. New nodes are inserted at the boundary of the fracture failure region and connected to the mesh before fracture to form a free surface. The finite element mesh is regenerated based on the free surface, and the material properties and plastic strain of the component are updated. This update is repeated until the finite element model sequence of the component is obtained. If the fracture failure criterion is not met, fracture failure has not occurred. The physical field data of the entire collision process is extracted based on the finite element model sequence. The displacement field data is mapped to the three-dimensional geometric model to generate the vertex displacement field. The vertex coordinates of the three-dimensional model are updated by the vertex displacement difference algorithm. The updated vertex normal vector and texture coordinates are calculated. The position and shape of the components are updated in combination with the physical field data. The update is repeated until the collision animation frame sequence is obtained. Based on the collision animation frame sequence and the equivalent stress and plastic strain distribution, the color map of each voxel is determined by the volume rendering algorithm. For each pixel, the color is looked up in the color map based on the stress value and strain value. The light projection algorithm emits light to each pixel to determine the intersection point with different voxels. The color of each pixel is determined based on the intersection point information. The determination is repeated until the stress distribution cloud map and plastic strain cloud map are obtained. Random crack distribution is generated on the boundary surface of fracture units using a mesh breaking algorithm. A polygonal mesh is generated with each crack line as the starting point. The polygonal mesh is triangulated, and a normal map and height map are added to generate the initial fracture morphology. Random perturbation is added to obtain the final fracture texture and fracture morphology. Fragment particles are randomly generated within the fracture unit. The kinematic parameters of each fragment particle are initialized. The scattering trajectory of each fragment particle is determined by explicit integration. Fracture effects are constructed based on the scattering trajectory and fragment particles. The final collision visualization scene is generated by combining the collision animation frame sequence, stress distribution cloud map, plastic strain cloud map and fracture effects. For example, a pump body of a power station was hit by a foreign object during operation. It is necessary to perform a visual analysis of the collision process and determine whether the pump body has undergone plastic deformation or fracture failure based on the collision simulation results. Comparing the equivalent stress and yield stress, it was found that the maximum equivalent stress was 400 MPa, which is greater than the yield strength of the pump body material of 350 MPa. Therefore, the pump body has undergone plastic deformation. Substituting the equivalent stress and plastic strain into the fracture failure criterion for judgment, it was found that some units met the failure conditions. Therefore, the pump body also experienced local fracture failure. For regions undergoing plastic deformation, the displacement field is extracted and node coordinates are updated to generate a new finite element mesh. Material properties and plastic strain are updated, and this process is repeated multiple times to obtain a sequence of finite element models of the pump body after deformation. For regions experiencing fracture failure, nodes of the failed elements are extracted and removed from the mesh. New nodes are inserted at the fracture boundary and the mesh is regenerated. Material properties and plastic strain are updated, and this process is repeated multiple times to obtain a sequence of finite element models of the pump body after fracture. Based on the finite element model sequence, physical field data of the entire collision process is extracted. The displacement field is mapped to the 3D model of the pump body to generate vertex displacement fields. Vertex coordinates, normal vectors, and texture coordinates are updated. The position and shape of the pump body are updated in combination with the physical field data, and a collision animation frame sequence is generated. Based on the collision animation frame sequence and equivalent stress / strain distribution, stress cloud maps and plastic strain cloud maps of the pump body are generated through volume rendering and ray casting algorithms, which intuitively show the stress and deformation of the pump body. For fracture failure regions, random crack distribution is generated through a mesh breaking algorithm. A polygonal mesh is constructed and triangulated to generate fracture morphology and fracture texture. Fragment particles are generated within the fracture unit, their scattering trajectories are calculated, and fracture effects are constructed. Finally, the collision animation, stress cloud map, strain cloud map, and fracture effects are integrated to generate a visualized scene of the entire pump body collision process, intuitively showing the complete process of the pump body's deformation, fracture, and breakage under the impact of the collision.

[0038] In this embodiment, by extracting the equivalent stress and equivalent plastic strain data from the collision results and comparing them with the material yield strength and fracture failure criteria, the system accurately determines whether the component has undergone plastic deformation and fracture failure, providing a basis for subsequent visualization processing. The node coordinates are updated using a node displacement correction algorithm, and a finite element mesh is generated based on the new coordinates. Simultaneously, material properties and plastic strain are updated. Multiple iterations yield a sequence of deformed component models, realistically reproducing the progressive deformation process of the component under collision load. By identifying failed elements and removing corresponding nodes, new nodes are inserted at the fracture boundary, and the mesh is regenerated, while simultaneously updating material properties and plastic strain. Multiple iterations yield a sequence of fractured component models, realistically reproducing the component fracture. By extracting the physical field data corresponding to the finite element model sequence, the complex failure process was analyzed, and rich data information such as the displacement field, stress field, and strain field of the component during the entire collision process was obtained. This provided a data foundation for visualization rendering. By mapping the displacement field to the 3D model of the component and updating the vertex coordinates, normal vectors, and texture coordinates, and combining the physical field data to update the position and shape of the component in real time, a continuous, smooth, and fluid collision animation frame sequence was generated. This intuitively demonstrated the motion changes of the component during the collision process. In summary, this embodiment not only intuitively demonstrated the deformation and failure process of the component, but also revealed the stress and strain distribution law inside the component, providing intuitive and reliable technical support for collision consequence analysis, accident investigation, safety assessment, and other work.

[0039] In one alternative implementation, By emitting rays to each pixel using a ray casting algorithm and determining the intersection points with different voxels, the color corresponding to each pixel is determined by the following formula: ; in, CP ( u , v ) represents a pixel ( u , v ) color, CV ( x , y , z ) represents voxels ( x , y , z The color value of ) S ( x , y , z ) represents voxels ( x , y , z The scattering coefficient of ) t ( x , y , z ) represents voxels ( x ,y , z The absorption coefficient of ) l The path of light propagation represents the distance light travels within a voxel. dλ A ray element is used to calculate the path of a ray. l The points on the scale.

[0040] In this embodiment, a ray casting algorithm is used to emit light to each pixel and calculate the intersection point of the light with the voxel. The pixel's color is then calculated based on the voxel attributes (color, scattering coefficient, absorption coefficient) at the intersection point, achieving realistic rendering of 3D volumetric data. This generates realistic visualization results with effects such as translucency, lighting, and shadows. The algorithm comprehensively considers the absorption, scattering, and transmission of light by the voxel, making the rendering results more realistic and accurately representing the optical properties of different materials. By integrating the absorption coefficient along the light propagation path, the energy attenuation process of light propagating through the voxel is simulated, resulting in voxels farther from the light source being darker, and voxels farther from the light source being darker. Voxels closer to the light source are brighter, producing a realistic light attenuation effect. By introducing light element-wise operations into the integral calculation, the light propagation path is discretized, improving the efficiency of rendering calculations. By setting appropriate sampling step size and termination conditions, unnecessary calculations are reduced while ensuring rendering quality, thus improving the algorithm's execution efficiency. In summary, this embodiment achieves high-quality, highly realistic volumetric data rendering, generating visualization results with effects such as semi-transparency, lighting, and shadows. It restores the physical mechanism of light and object interaction in the real world, making the rendering results more realistic and natural, and providing strong technical support for the visualization analysis of complex volumetric data.

[0041] Figure 2 This is a schematic diagram of the structure of the physical simulation-based twin power plant collision visualization simulation system according to an embodiment of the present invention, as shown below. Figure 2 As shown, the system includes: The second unit is used to add the physical simulation model to a pre-set physical simulation engine to generate a basic collision simulation model, obtain the initial collision condition parameters of the two power station components that collide and add them as boundary conditions to the basic collision simulation model, refine the local mesh of the power station components through an adaptive mesh subdivision algorithm, define the material constitutive model through a pre-set finite element model in combination with the material data corresponding to the power station components, construct the nonlinear dynamic equation of the collision process based on the nonlinear properties of the power station components, solve the nonlinear dynamic equation in the time domain using an explicit central difference algorithm, generate physical field information, calculate the strain field and stress field of the power station components through the conserved nodal average stress method, determine the equivalent stress distribution and equivalent plastic strain distribution, and generate the collision result based on the yield criterion and fracture failure criterion. The third unit is used to determine whether the power station component has undergone plastic deformation or fracture failure based on the collision results. If plastic deformation occurs, the finite element model is updated using a node displacement correction algorithm. If fracture failure occurs, the failed element is removed from the finite element model using an element deletion method, and a free surface is generated. The simulation is repeated until the collision ends, generating physical field data for the entire collision process. The displacement field data in the physical field data of the entire collision process is mapped to the three-dimensional geometric model. The position and shape of the power station component are updated using a vertex displacement difference algorithm, generating a collision animation frame sequence. Based on the equivalent stress distribution and equivalent plastic strain distribution, a stress distribution cloud map and a plastic strain cloud map are generated using a volume rendering algorithm. For power station components that have fractured, fracture textures and fracture morphologies are generated using a mesh breaking algorithm based on the position and normal vector of the fracture element. Fracture effects are constructed using flying fragments. The collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects are combined to generate a collision visualization scene.

[0042] A third aspect of the present invention, An electronic device is provided, comprising: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0043] Fourth aspect of the embodiments of the present invention, A computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0044] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0045] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A physical simulation-based visual simulation method for collisions in twin power plants, characterized in that, include: A three-dimensional geometric model of a twin power station is obtained. The twin power station is spatially partitioned using an octree spatial partitioning algorithm to generate a voxel model. Geometric parameters of the power station components in the twin power station are extracted based on the topological relationship of the voxel units. Material data of the power station components are obtained based on the entity design specifications and bill of materials of the twin power station. The geometric parameters and the material data are associated to construct a material association model. The mechanical model of the internal frame structure of the entity power station is obtained according to the structural design drawings of the entity power station. The force distribution is calculated by combining mechanical principles to obtain a mechanical simulation model. This mechanical simulation model is then fused with the material association model to obtain a physical simulation model. The physical simulation model is added to a pre-set physical simulation engine to generate a basic collision simulation model. The initial collision condition parameters of the two power station components that collide are obtained and added to the basic collision simulation model as boundary conditions. The local mesh of the power station components is refined using an adaptive mesh subdivision algorithm. Combined with the material data corresponding to the power station components, a material constitutive model is defined using a pre-set finite element model. The nonlinear dynamic equation of the collision process is constructed based on the nonlinear properties of the power station components. The nonlinear dynamic equation is solved in the time domain using an explicit central difference algorithm. Physical field information is generated, and the strain field and stress field of the power station components are calculated using the conserved nodal mean stress method. The equivalent stress distribution and equivalent plastic strain distribution are determined. The collision results are generated based on the yield criterion and fracture failure criterion. Based on the collision results, it is determined whether the power station component has undergone plastic deformation or fracture failure. If plastic deformation occurs, the finite element model is updated using a node displacement correction algorithm. If fracture failure occurs, the failed element is removed from the finite element model using an element deletion method, and a free surface is generated. The simulation is repeated until the collision ends, generating physical field data for the entire collision process. The displacement field data in the physical field data of the entire collision process is mapped to the three-dimensional geometric model. The position and shape of the power station component are updated using a vertex displacement difference algorithm, generating a collision animation frame sequence. Based on the equivalent stress distribution and equivalent plastic strain distribution, a stress distribution cloud map and a plastic strain cloud map are generated using a volume rendering algorithm. For power station components that have fractured, fracture textures and fracture morphologies are generated using a mesh breaking algorithm based on the position and normal vector of the fracture element. Fracture effects are constructed using flying fragments. The collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects are combined to generate a collision visualization scene.

2. The method according to claim 1, characterized in that, A three-dimensional geometric model of a twin power station is obtained. The twin power station is spatially partitioned using an octree space partitioning algorithm to generate a voxel model. Geometric parameters of the power station components are extracted based on the topological relationships of the voxel units. Material data of the power station components is obtained based on the twin power station's solid design specifications and bill of materials. The geometric parameters and material data are correlated to construct a material correlation model. A mechanical model of the internal frame structure of the solid power station is obtained based on its structural design drawings. The stress distribution is calculated using mechanical principles to obtain a mechanical simulation model, which is then fused with the material correlation model to obtain a physical simulation model, including: Point cloud data corresponding to the twin power station is obtained through 3D laser scanning technology. The 3D geometric model of the twin power station is obtained through point cloud data preprocessing and registration stitching. Based on the octree space partitioning algorithm, the root node of the octree is determined according to the bounding box of the twin power station. At each node, the complexity metric is determined according to the number of point clouds. The complexity threshold is combined with a pre-set complexity threshold. If the complexity of the current node is greater than the complexity threshold, the current node is recursively divided into eight equal parts. The judgment is repeated until the complexity of all nodes is less than the complexity threshold, and leaf nodes are obtained. For each leaf node, the corresponding voxel unit is generated and combined to obtain the voxelized model. Based on the topological relationship corresponding to the voxel unit, different voxel sets corresponding to the power station components in the twin power station are identified. The outer envelope box of the component is calculated based on the voxel set belonging to the same component to obtain the size information of the current component. The shape features of the current component are extracted by fitting the surface mesh of the voxel set. The geometric center of the outer envelope box is used as the spatial position coordinate of the current component. The geometric parameters of the current component are obtained by combining the size information, shape features and spatial position coordinates. Each component in the twin power station is traversed and the corresponding geometric parameters are obtained. Based on the physical design specifications and bill of materials of the twin power stations, the material composition of the power station components in each power station is obtained. For metallic materials, a material property database is constructed by combining physical and mechanical property parameters. The property parameters of the material composition in each component are retrieved from the material property database. For non-metallic materials and new materials, mechanical performance parameters are determined by combining simulation experiments. The material data is obtained by combining the property parameters and the mechanical performance parameters. The bill of materials is traversed and a first relationship table and a second relationship table corresponding to the power station components and the geometric parameters and the material data are generated respectively. The first relationship table and the second relationship table are connected, and the material property parameters or mechanical performance parameters are retrieved and combined to obtain the material association model. Based on the structural design drawings of the actual power station, a mechanical model of the internal frame structure of the actual power station is obtained. Key load modes are determined by combining mechanical principles and load condition analysis techniques. These key load modes are divided into static loads and dynamic loads. For static loads, the static force distribution is calculated using the finite element statics analysis method. For dynamic loads, the vibration modes and frequency response characteristics of the internal frame structure are obtained using the finite element dynamics analysis method. Based on these vibration modes and frequency response characteristics, and combined with transient mechanical analysis methods, the dynamic force distribution is obtained by solving the dynamic equations using the direct integration method. The static and dynamic force distributions are then combined to obtain a mechanical simulation model. Finally, the static and dynamic force distributions in the mechanical simulation model are mapped to the material correlation model to obtain the physical simulation model.

3. The method according to claim 2, characterized in that, Based on the aforementioned vibration modes and frequency response characteristics, and combined with transient mechanics analysis methods, the dynamic force distribution is obtained by solving the dynamic equations using the direct integration method, as shown in the following formula: ； in, F ( t n+1 ) indicates at time step n +1 external force load, [M] represents the mass matrix, { u n+1 } represents a time step n The displacement vector is +1, and [C] represents the damping matrix. Indicates time step n velocity vector, Δ t Indicates the time increment. Indicates time step n acceleration vector, Indicates time step n The acceleration vector is +1, and [K] represents the stiffness matrix.

4. The method according to claim 1, characterized in that, The physical simulation model is added to a pre-set physical simulation engine to generate a basic collision simulation model. The initial collision condition parameters of the two colliding power station components are obtained and added as boundary conditions to the basic collision simulation model. The power station components are locally meshed using an adaptive mesh refinement algorithm. Combined with the material data corresponding to the power station components, a material constitutive model is defined using a pre-set finite element model. Based on the nonlinear properties of the power station components, a nonlinear dynamic equation for the collision process is constructed. The nonlinear dynamic equation is solved in the time domain using an explicit central difference algorithm, generating physical field information. The strain and stress fields of the power station components are calculated using the conserved nodal mean stress method to determine the equivalent stress distribution and equivalent plastic strain distribution. The collision results are generated based on the yield criterion and fracture failure criterion, including: A physical simulation engine is built based on the explicit finite element algorithm. The physical simulation model is added to the physical simulation engine and the basic collision simulation model is generated. Kinematic analysis was performed on the two power station components that collided. Trajectory tracking was performed based on real-time monitoring data to determine the position of the center of mass. The angular velocity and contact position at the time of collision were determined by combining the geometry and attitude of each component. The initial collision condition parameters were obtained by combining the center of mass position, angular velocity and contact position and added as boundary conditions to the basic collision simulation model. For the power plant components that have collided, an initial mesh is set using an adaptive mesh subdivision algorithm. Stress analysis is performed on the initial mesh to determine the equivalent stress of each element and mark the elements whose equivalent stress is greater than a preset mesh stress threshold. The marked areas are taken as the areas to be subdivided. The areas to be subdivided are recursively subdivided using an adaptive spatial partitioning algorithm until the preset mesh density requirement is met. The subdivided mesh is smoothed and coupled with the initial mesh to obtain the second finite element model. Based on the second finite element model and combined with the material data corresponding to the power station components, a bilinear elastoplastic constitutive model is constructed for isotropic materials based on the elastic modulus and yield stress through a pre-set finite element model. For anisotropic materials, the yield function is determined based on the anisotropic parameters corresponding to the current material. The nonlinear dynamic equation corresponding to the collision process is constructed by combining the nonlinear properties of the power station components. The nonlinear properties include large deformation, material nonlinearity and contact nonlinearity. The nonlinear dynamic equations are solved using the explicit central difference algorithm, and the stress, internal force, and contact state of each component are updated. Physical field information is generated, and the strain field and stress field of the power station component are calculated using the conserved nodal mean stress method. Based on the strain field and stress field, the equivalent stress distribution and equivalent plastic strain distribution are determined. According to the material properties and current state of the power station component, combined with the corresponding yield criterion and fracture failure criterion, the collision result is generated and output.

5. The method according to claim 4, characterized in that, For anisotropic materials, the yield function is determined based on the anisotropic parameters corresponding to the current material, as shown in the following formula: ； in, F Indicates the material is in y shaft and z Differential resistance to yielding between shafts G Indicates the material is in z shaft and x Differential resistance to yielding between shafts H Indicates the material is in x shaft and y Differential resistance to yielding between shafts σ 11 Indicates the material is in x Components of normal stress in the axial direction σ 22 Indicates the material is in y Components of normal stress in the axial direction σ 33 Indicates the material is in z Components of normal stress in the axial direction m This indicates the Horsford Index. σ y This represents the yield stress.

6. The method according to claim 1, characterized in that, Based on the collision results, it is determined whether the power station component has undergone plastic deformation or fracture failure. If plastic deformation occurs, the finite element model is updated using a node displacement correction algorithm. If fracture failure occurs, the failed element is removed from the finite element model using an element deletion method, and a free surface is generated. The simulation is repeated until the collision ends, generating full collision process physical field data. The displacement field data in the full collision process physical field data is mapped to the three-dimensional geometric model. The position and shape of the power station component are updated using a vertex displacement difference algorithm, generating a collision animation frame sequence. Based on the equivalent stress distribution and equivalent plastic strain distribution, a stress distribution cloud map and a plastic strain cloud map are generated using a volume rendering algorithm. For power station components that have fractured, fracture textures and fracture morphologies are generated using a mesh breaking algorithm based on the position and normal vector of the fracture element. Fracture effects are constructed using flying fragments. The collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects are combined to generate a collision visualization scene, including: Based on the collision results, it is determined whether the power station component has undergone plastic deformation or fracture failure. If plastic deformation occurs, the displacement field data of the current component is extracted based on the node displacement correction algorithm. For each node, the node coordinates are updated according to the corresponding displacement vector. A finite element mesh is generated based on the updated node coordinates, and the material properties and plastic strain of the current component are updated. The update is repeated until the finite element model sequence corresponding to the current component is obtained. If the collision result indicates that the current power plant component has fractured, then the failure unit is identified based on the fracture failure criterion and a set of failure units is generated. For each failure unit, the corresponding node set is extracted and the node set is removed. Nodes are inserted at the boundary of the fracture failure region and connected to the finite element mesh before fracture to obtain the free surface. The finite element mesh is regenerated based on the free surface and the material properties and plastic strain of the current component are updated. The update is repeated until the finite element model sequence corresponding to the current component is obtained. The physical field data of the full collision process is extracted based on the finite element model sequence. The displacement field in the physical field data of the full collision process is mapped to the three-dimensional geometric model to generate the vertex displacement field. The vertex coordinates of the three-dimensional geometric model are updated by the vertex displacement difference algorithm and the corresponding normal vectors and texture coordinates are calculated. The position and shape of the power station component are updated by combining the physical field data of the full collision process to obtain the collision animation frame sequence. For the power plant structure, based on the collision animation frame sequence and the equivalent stress distribution and equivalent plastic strain distribution, a color mapping table for each voxel is determined by a volume rendering algorithm. The stress and strain values ​​are searched in the color mapping table. A ray casting algorithm is used to emit rays to each pixel and determine the intersection points with different voxels. The color corresponding to each pixel is determined. The operation is repeated to obtain the stress distribution cloud map and plastic strain cloud map. For power plant components that have experienced fracture failure, a mesh breaking algorithm is used to generate random crack distributions on the boundary surfaces corresponding to the fracture elements based on the fracture direction and the material data, according to the location and normal vector of the fracture elements. A polygonal mesh is generated with each crack line in the random crack distribution as the starting point. The polygonal mesh is triangulated and a normal map and height map are added to generate an initial fracture morphology, and random perturbations are added to obtain the fracture texture and fracture morphology. Fragment particles are randomly generated within the fracture elements. The kinematic parameters of each fragment particle are initialized, and the scattering trajectory of each fragment particle is determined by explicit integration. Fracture effects are constructed based on the scattering trajectories and fragment particles. A collision visualization scene is generated by combining the collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects.

7. The method according to claim 6, characterized in that, By emitting rays to each pixel using a ray casting algorithm and determining the intersection points with different voxels, the color corresponding to each pixel is determined by the following formula: ； in, CP ( u , v ) represents a pixel ( u , v ) color, CV ( x , y , z ) represents voxels ( x , y , z The color value of ) S ( x , y , z ) represents voxels ( x , y , z The scattering coefficient of ) τ ( x , y , z ) represents voxels ( x , y , z The absorption coefficient of ) λ The path of light propagation represents the distance light travels within a voxel. dλ A ray element is used to calculate the path of a ray. λ The points on the scale.

8. A physical simulation-based visual simulation system for twin power plant collisions, used to implement the method described in any one of claims 1-7, characterized in that, include: The first unit is used to obtain a three-dimensional geometric model of the twin power station. The twin power station is spatially partitioned using an octree spatial partitioning algorithm to generate a voxel model. Based on the topological relationship of the voxel units, the geometric parameters of the power station components in the twin power station are extracted. The material data of the power station components are obtained based on the entity design specifications and bill of materials of the twin power station. The geometric parameters and the material data are associated to construct a material association model. The mechanical model of the internal frame structure of the entity power station is obtained according to the structural design drawings of the entity power station. The force distribution is calculated by combining mechanical principles to obtain a mechanical simulation model. This mechanical simulation model is then fused with the material association model to obtain a physical simulation model. The second unit is used to add the physical simulation model to a pre-set physical simulation engine to generate a basic collision simulation model, obtain the initial collision condition parameters of the two power station components that collide and add them as boundary conditions to the basic collision simulation model, refine the local mesh of the power station components through an adaptive mesh subdivision algorithm, define the material constitutive model through a pre-set finite element model in combination with the material data corresponding to the power station components, construct the nonlinear dynamic equation of the collision process based on the nonlinear properties of the power station components, solve the nonlinear dynamic equation in the time domain using an explicit central difference algorithm, generate physical field information, calculate the strain field and stress field of the power station components through the conserved nodal average stress method, determine the equivalent stress distribution and equivalent plastic strain distribution, and generate the collision result based on the yield criterion and fracture failure criterion. The third unit is used to determine whether the power station component has undergone plastic deformation or fracture failure based on the collision results. If plastic deformation occurs, the finite element model is updated using a node displacement correction algorithm. If fracture failure occurs, the failed element is removed from the finite element model using an element deletion method, and a free surface is generated. The simulation is repeated until the collision ends, generating physical field data for the entire collision process. The displacement field data in the physical field data of the entire collision process is mapped to the three-dimensional geometric model. The position and shape of the power station component are updated using a vertex displacement difference algorithm, generating a collision animation frame sequence. Based on the equivalent stress distribution and equivalent plastic strain distribution, a stress distribution cloud map and a plastic strain cloud map are generated using a volume rendering algorithm. For power station components that have fractured, fracture textures and fracture morphologies are generated using a mesh breaking algorithm based on the position and normal vector of the fracture element. Fracture effects are constructed using flying fragments. The collision animation frame sequence, the stress distribution cloud map, the plastic strain cloud map, and the fracture effects are combined to generate a collision visualization scene.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Images