10 results about "Grid computation" patented technology

This application provides a method and related equipment for preprocessing game trajectory data. The method includes: converting game trajectory data into a first grid sequence; merging high-frequency adjacent grids in the first grid sequence to calculate a second grid sequence; extracting embedding vectors from the second grid sequence to calculate an embedding vector sequence; and using the embedding vector sequence as a training set input to a deep learning model. This application's embodiment, by converting game trajectory data into a first grid sequence, merging high-frequency adjacent grids in the first grid sequence, and then extracting embedding vectors from the second grid sequence, can capture the prior influence of environmental factors on trajectory distribution and the potential contextual semantic information in the trajectory, achieving effective compression of long trajectory data.

Embodiments of the present application disclose a reservoir over-standard flood inundation risk zoning method, device and equipment, by determining the reservoir siltation influence probability factor and the reservoir dam break influence probability factor under the siltation condition, the reservoir siltation influence probability factor and the reservoir dam break influence probability factor are used to calibrate the reservoir inundation comprehensive risk degree of siltation-dam break double influence probability under the over-standard flood condition, and the K-means clustering algorithm is used to cluster and analyze the reservoir inundation comprehensive risk degree of each grid calculation unit, and finally the flood risk grade is determined. The technical problem of low accuracy of prediction and zoning of flood inundation risk in the prior art for the reservoir area with high sediment concentration and weak dredging capacity is solved, and a scientific basis and certain technical support are provided for reservoir flood inundation risk assessment and disaster prevention and mitigation planning in high sediment concentration basin areas.

The application belongs to the field of computational mechanics, and discloses a scale order reduction near-field finite element calculation acceleration method and implementation method for structure fracture analysis.The acceleration method comprises the following steps: firstly, geometric modeling is performed on a target structure; secondly, the split finite element is recorded as a fine scale element and constitutes a fine scale grid; thirdly, based on a spatial neighborhood relationship, the fine scale element is collected into a coarse scale element and constitutes a coarse scale grid; fourthly, an equivalent shape function of the coarse scale element is obtained; fifthly, the overall stiffness matrix and the load vector of the coarse scale grid are calculated; sixthly, the overall node displacement of the coarse scale and fine scale grid is solved; seventhly, it is judged whether the current loading step or time step reaches the maximum loading step or time step; if yes, the displacement nephogram and equivalent damage nephogram of the target structure are output; if not, the key state of all fine scale near-field elements is updated; and finally, the implementation method is that the acceleration method is taken as a core calculation module to be implemented in different scenes.The application significantly improves the calculation efficiency of near-field dynamics in large-scale simulation analysis.

The present application relates to the technical field of printed circuit control, and in particular to a multi-layer PCB winding pressing alignment method for high-frequency planar transformers. The method obtains a PCB layout and divides a grid, calculates the copper foil area proportion of each grid; determines the radial unit vector of each positioning pin based on the vector relationship between the positioning pin area and the layout geometric center; in the pressing process, the resin extrusion trend scalar is generated by combining the real-time obtained deformation height of each grid under pressure and the copper foil area proportion, and the two-dimensional flow vector is obtained through local space partial derivative analysis; then the vector and the radial unit vector are combined to calculate the radial extrusion component of each positioning pin area, and finally the pressing process is dynamically controlled according to the numerical distribution of all components to realize high-precision alignment control.

This invention relates to the field of geological exploration and 3D geological modeling technology, specifically a method and system for 3D modeling of ore bodies based on multi-source exploration geological data. The method includes: acquiring multi-source spatial discrete data and spatial continuous gradient data within a preset 3D spatial range, and constructing a spatial voxel mesh based on the multi-source spatial discrete data and spatial continuous gradient data; calculating the spatial metric tensor matrix of each voxel node in the spatial voxel mesh to generate a spatial topological barrier field; using the source point of the target object as the initial zero-order jump set to generate an initial symbolic distance field; extracting the zero isosurface of the initial symbolic distance field to obtain the Euclidean deviation residual; if the Euclidean deviation residual exceeds a preset residual threshold, continuing until the Euclidean deviation residual does not exceed the residual threshold; when the Euclidean deviation residual does not exceed the residual threshold, generating a 3D triangular mesh of the target object. This invention achieves simultaneous continuous generation of the ore body surface and fault blocking.

This application discloses a hybrid acceleration method and related apparatus for computational fluid dynamics simulation based on data dimensionality reduction and multigrid interpolation, relating to the field of computational fluid dynamics technology. The method includes: calling a reduced-order prediction model based on the fine-grid physical field data of the current time step to generate a coarse-grid predicted physical field for the current time step; spatially interpolating the coarse-grid predicted physical field of the current time step and mapping it onto a fine-grid to generate initial fine-grid field data for the current time step; using the initial fine-grid field data of the current time step as initial conditions, continuing to perform computational fluid dynamics solutions for the next number of time steps on the fine-grid to generate updated fine-grid physical field data; using the updated fine-grid physical field data as the benchmark for the next iteration cycle until a preset total simulation time step is reached, and outputting the final simulation result of the target physical field. This application fundamentally reduces the dependence of high-fidelity simulation on fine-grid computational resources.

This application relates to an adaptive cross-tree mesh computation method and apparatus for complex microscale structures. The method is applicable to efficient numerical solutions in automatically generated cross-tree meshes for complex micro-surfaces. Addressing the problems of large local geometric scale spans, complex microstructure surface morphology, difficulty in generating high-quality meshes near boundaries, and insufficient adaptability of traditional solvers to special meshes in complex microscale structure scenarios of aircraft, this invention constructs an adaptive cross-tree mesh computation method for complex microscale structures. By combining the adaptive cross-tree mesh with a flux-solving algorithm, stable and efficient numerical solutions for flow fields of complex shapes can be achieved, thereby improving the flexibility of mesh generation, flow field resolution, and overall computational efficiency under complex configuration conditions.

This invention discloses a control method for an adaptive deformable wing based on Hessian vector product, belonging to the field of aircraft design technology. The method includes: S1, defining the wing's design variables; S2, updating the design variables using an optimizer based on gradient information; S3, updating the wing's geometry using the FFD geometric parameterization method and adjusting the fluid computation grid; S4, calculating the aerodynamic performance data and first-order gradient of sample points using the fluid computation grid after perturbation of the design variables; S5, calculating the robust objective function and perturbation vector, and calculating the total gradient using the Hessian vector product module; S6, determining whether the wing configuration determined by the design variables has converged based on the robust objective function and the total gradient. If converged, the wing configuration is taken as the final robust optimal wing design; otherwise, the total gradient is used as gradient information, and the process returns to S2. This achieves a balance between efficiency and accuracy, ensuring the reliability and economy of the aircraft in actual operation.

The invention discloses a polygon filling method, system and device based on global subdivision grid multi-level combination and a medium, and belongs to the field of geographic information processing and computer graphics. According to the polygon filling method, the advantage that different levels of grids of global subdivision grids can be rapidly indexed based on coding rules is utilized, the number of grids to be calculated is reduced to hundreds of levels according to the size of a polygon bounding rectangle through the thought of first coarse and then fine, further study and judgment are carried out through the distance and position of the center points of the grids, and the calculation accuracy is improved. And obtaining a geometrical relationship between all grids and polygons, and completing filling value assignment. The method is a filling technology for subdivision grids, and is suitable for filling of current mainstream subdivision grids such as GeoSOT and the like; meanwhile, processing of complex conditions such as polygon boundary area grid filling judgment is optimized, and the seamlessness and geometric correctness of the filling result are ensured; in addition, the grid-by-grid calculation mode is especially suitable for the GPU parallel acceleration technology, and the calculation efficiency is improved.