57 results about "Grid deformation" patented technology

The invention discloses a DTG printing method and system based on line scanning camera visual deformation and rigid body positioning, and relates to the technical field of digital textile printing. Dynamic images are collected in real time through a line scanning camera, and image coordinates are mapped to a printing equipment world coordinate system through a coordinate calibration module in combination with a calibration matrix; real-time error compensation is used for dynamically eliminating coordinate offset, and a dual-mode positioning module intelligently switches a positioning strategy; rigid body positioning realizes accurate positioning of non-deformation materials through template matching; deformation positioning performs adaptive compensation based on feature point extraction and grid deformation analysis, an image mapping module inversely corrects target pattern deformation according to a positioning result to generate a high-precision TIF image, the high-precision TIF image is converted into a printing instruction through an RIP processing module, and an integrated control module performs one-key triggering and dynamic parameter configuration and performs printing on a target pattern. The collection efficiency is optimized through light source synchronization and self-adaptive triggering, the system stability is improved in combination with a template database and an exception handling module, and high-quality printing of complex deformation materials is achieved.

The invention relates to the technical field of rock lumpiness identification, in particular to a rock lumpiness laser indicating device, which comprises L groups of laser projectors which are respectively arranged in different directions of a breaking hammer mechanical arm; the acquisition module is used for being matched with the laser projector to acquire grid deformation size data; and the central processing unit is used for obtaining the grid deformation size data and analyzing and processing the grid deformation size data to obtain a rock lumpiness index. According to the method, laser grid light is synchronously projected in multiple directions, deformation size data are acquired from different visual angles, and complementation of the visual angles in different directions can be realized, so that the problems of'dead angles' and information loss caused by mutual stacking and shielding among rocks are remarkably reduced, the coverage rate and reconstruction integrity of the whole point cloud are improved, and the reconstruction accuracy is improved. Through multi-view point cloud splicing and fusion, the obtained point cloud model is higher in integrity and richer in edge information, and subsequent fragmentation identification based on a machine learning model is facilitated.

The invention discloses an offshore floating body FSI forecasting method and system based on bidirectional coupling of FVM and MBD. The method comprises the steps that a numerical pool model containing a floating body and surrounding fluid is established in a CFD solver; establishing a floating body structure finite element model in the MBD solver; constructing a bidirectional coupling mapping relation between a fluid grid and a structural grid based on a preCICE open source library to realize real-time transmission of data; calculating a floating body surface fluid load time calendar value through a CFD solver, and transmitting the floating body surface fluid load time calendar value to the structure model; the MBD solver calculates structural dynamic response and feeds back a deformation time calendar value to a fluid domain; fluid grid deformation is dynamically updated based on a Laplace solver; iteratively executing a load transfer-structure response-grid updating process until the pressure and displacement residual error meet a convergence condition; and outputting a coupling forecast result of the fluid load and the structural deformation. High-precision two-way coupling is achieved through the open source coupling library, and the accuracy of floating body structure dynamic response forecasting in the complex wave environment is remarkably improved.

The data-driven non-intrusive shape-topology collaborative optimization method and system disclosed by the present invention relate to the field of structural optimization. First, the present invention uses mesh deformation technology to perform parametric modeling on the structure, and uses shape equations to constrain the displacements of the structure control points, ensuring the regular and smooth shape of the structure after mesh deformation; secondly, the present invention uses a variety of solvers to perform topology optimization on models of different shapes to obtain multi-source topology optimization response data; finally, the present invention establishes a non-intrusive surrogate model between the shape design variables and the topology optimization response data, and obtains a shape-topology collaborative optimization scheme based on the non-intrusive surrogate model. The present invention takes into account the interaction between shape optimization and topology optimization, helps to expand the design space, has higher optimization efficiency and optimization ability compared with other shape-topology optimization methods, and does not require complex sensitivity formula derivation, facilitating its application to practical engineering.

The present application belongs to the technical field of image processing, and specifically relates to a three-dimensional face model automatic generation method, which is characterized by the following specific steps: data loading processing, user model generation based on computer vision, user model grid re-topology, model grid local shape optimization, model grid fusion splicing, and output result model. The grid generation method ensures 2D mapping accuracy according to face key points; the grid deformation and smoothing method makes the user model and the standard face model closer in 3D space under the premise of reasonable deformation, so that the mapping is more accurate; the re-topology model grid local shape optimization utilizes original grid information to correct the grid deformity and improve the quality of the model. The model grid fusion splicing adopts the grid deformation and grid smoothing method, so that the output result meets the design requirements, guarantees the integrity and high quality of the model, and reduces the post-processing work.

The invention belongs to the field of structure optimization design, and discloses a coordinate space mapping-based geometric reconstruction method for topological optimization characteristics of a curved surface blade structure. On the basis of a topological optimization result of a three-dimensional curved surface blade structure of an engine, topological optimization geometric reconstruction is carried out on internal ribs, deformation from a three-dimensional grid to a two-dimensional grid is realized firstly, and a three-dimensional node grid is mapped to a rectangular flat plate according to a certain proportion through grid deformation according to a three-dimensional density optimization result. Secondly, processing a density result of the surface of the flat plate to obtain a plane density image, capturing key geometric features according to an image morphological processing and skeleton extraction technology, and obtaining skeleton point coordinates of corresponding parts; and mapping the extracted pixel skeleton points back to grid nodes in an actual two-dimensional model, mapping back to grid nodes of an original three-dimensional blade according to the two-dimensional grid nodes, reconstructing the model in CAD software according to the three-dimensional grid nodes, and completing geometric reconstruction of topological optimization of the three-dimensional curved surface structure in combination with an original cavity structure.

The present application relates to the field of virtual reality graphics processing, and discloses a virtual reality processing method for simulating a psychological situation, comprising: acquiring a line-of-sight focus data and calling a target grid model; calculating a deformation coefficient according to a spatial distance, completing low-pass filtering in combination with a historical frame smoothing weight to calculate a grid deformation transformation matrix; multiplying the matrix by a grid vertex coordinate to output a deformed grid vertex coordinate, and simultaneously enlarging each axis radius of an axis-aligned bounding box by 1.5 to 2.0 according to an expansion coefficient and updating a scene tree search node; extracting an absolute value of the matrix determinant as a deformation rate and passing it to a fragment shader to control a diffuse reflection field, and completing grid color rendering, wherein the present application converges grid morphological distortion in a grid transformation stage, uses in-situ transformation to eliminate topology reconstruction and bus retransmission, eliminates bus bandwidth saturation hazards, eliminates visual cone body rejection defects in cooperation with dynamic expansion of a bounding box, and guarantees a view refresh rate and spatial realism.

Embodiments of the present disclosure relate to an image processing method, device, equipment and medium, wherein the method comprises: acquiring a plurality of groups of image pairs with different grid resolutions; according to the order of the grid resolutions from small to large, sequentially calculating the grid vertex positions of the corresponding second images according to the corresponding second images and the associated reference images of each group of image pairs in the plurality of groups of image pairs, and sequentially passing the grid vertex positions of the previous grid resolution to the next layer until the target grid deformation image of the image pair corresponding to the maximum grid resolution is acquired, and calculating the grid flow according to the target grid deformation image and the first image, so as to align the original reference image and the first image according to the grid flow. Thus, the grid vertex positions of the previous pair of images are passed from coarse to fine according to the order of the grid resolutions from small to large to control the grid shape deformation, and the grid flow is determined according to the deformed grid, thereby improving the acquisition efficiency and robustness of the grid flow and improving the image alignment effect.

The invention discloses a composite material lightning stroke dynamic ablation simulation method based on ABAQUS. The method comprises the specific steps that S1, an anisotropic composite material three-dimensional model is established; s2, applying lightning stroke current and heat flow load, setting a variable initial value at normal temperature, and calculating the pyrolysis degree, potential and temperature field of each node; s3, giving a grid deformation displacement load in a static ablation depth direction, and driving the nodes to move to simulate deformation; and S4, constructing an information transfer matrix, mapping a temperature update grid, and taking the update grid as a next electrothermal analysis initial material domain. According to the ABAQUS-based composite material lightning stroke dynamic ablation simulation method provided by the invention, flowing of a dynamic grid between different layers of an anisotropic composite material laminated plate is realized, and parallel coupling calculation of composite material temperature rise and surface ablation depression is realized.

The application discloses a kind of based on dynamic mesh technology's hydraulic machinery system component progressive wear quantity quantification prediction method, comprising: constructing fluid machinery three-dimensional geometric model;CFD numerical calculation is carried out under specified boundary condition based on fluid machinery three-dimensional geometric model, obtains fluid machinery internal steady-state initial field;Based on fluid machinery internal steady-state initial field, inject silt particles, introduce wear model, obtain transient sand steady flow field;Based on transient sand steady flow field, according to wall surface damage rate, calculate wear depth;The difference of wear depth of current time and last time is calculated, and the grid deformation of current time is obtained;According to the grid deformation of current time, it is decomposed in particle impact velocity direction, reach the effect of wall surface topography change, obtain wear depth under total time, wall surface topography change, transient flow field change.The application improves the prediction precision of fluid machinery wear depth.

The invention discloses a three-dimensional net capture visualization method based on simulation data. The three-dimensional net capture visualization method is used for efficiently presenting the dynamic wrapping effect of a rope net capture target at a Web end. Comprising the following steps: S1, calculating AABB and OBB of a target to be captured, uniformly sampling on the surfaces of bounding boxes, emitting rays along an outer normal, colliding with the surfaces of the target to solve an offset distance, comparing average fitting distances of the two types of bounding boxes, and selecting an optimal bounding box; s2, the sampling density is determined according to the mesh size, the opposite face of the incoming direction is removed, sampling points are generated on the other five faces, the sampling points and the mesh body nodes are bound one to one according to the minimum space distance, redistribution is carried out if conflicts occur, and a mapping dictionary M is output; and S3, by taking the simulation position at the contact moment as a starting point, performing linear interpolation on the nodes in the M to a target position according to a fixed time window delta T, reading simulation data by other nodes, and submitting WebGL to drive grid deformation, thereby realizing vivid transition from free motion to target fitting, avoiding real-time recalculation at a browser end and improving rendering performance.

The invention discloses a hyperelastic material large strain visual measurement method based on surface grid deformation identification, which comprises the following steps: processing a hyperelastic material to be measured into a dumbbell-shaped sample, and constructing uniformly distributed regular grid lines on the surface of a scale distance section of the sample; stretching the sample at a preset stretching rate, and continuously collecting surface grid deformation images of the sample while stretching; identifying the position of each grid line in each frame of image; and according to the position change condition of the corresponding grid lines in the two adjacent frames of images, determining the strain condition of the hyperelastic material at each moment under the preset stretching rate. The invention aims to solve the problems that a traditional measurement method is inaccurate in measurement under the condition of large strain of the hyperelastic material, for example, speckles chapped and deformed in a random speckle sample laying method are difficult to recover, an extensometer cannot obtain local strain and results have uncontrollable factors, and a traditional DIC method is large in measurement error under large strain.

The invention provides a grid deformation and particle deposition layer and wall surface transient thermal coupling temperature prediction method. In the modern industrial field, high-efficiency operation and precise maintenance of high-temperature heat flow equipment are crucial. The traditional simulation method neglects the existence of a sedimentary layer or simplifies the sedimentary layer, so that the prediction of the heat transfer characteristics of the thermal boundary has deviation. Through the user-defined function UDF, the deformation of the deposition layer is dynamically predicted, the heat conductivity coefficient of the deposition layer is synchronously updated and assigned, the deposition layer and the solid wall surface are coupled, and then the influence of the particulate matter deposition layer on the heat exchange characteristic of the high-temperature heat flow equipment is predicted. The prediction accuracy of thermal boundary change caused by particle deposition in high-temperature heat flow equipment is improved, and the problem of prediction deviation in traditional simulation is solved. According to the method, the influence of the particulate matter deposition layer on the heat exchange characteristics of the high-temperature heat flow equipment can be accurately predicted, and the method has certain guiding significance on design, operation and maintenance of the high-temperature heat flow equipment through which the particulate matter flows.

The present invention discloses a radial basis function mesh deformation method based on virtual nodes and an airfoil design method. Currently, generating a high-quality deformed mesh by a mesh deformation method is still an open problem. The present invention first obtains the distribution of internal virtual nodes based on the internal normal of the moving boundary point; then reduces the number of internal virtual nodes to obtain a small number of uniformly distributed internal virtual nodes; then obtains the displacement of each internal virtual point by interpolation of K adjacent boundary nodes; finally, uses the internal virtual nodes and the original boundary nodes to construct the linear equilibrium equation of the object surface, solves the coefficient matrix, and thus interpolates and calculates the displacement of the internal nodes to obtain the deformed mesh. The present invention effectively improves the quality of the deformed mesh and the degree of mesh deformation, and realizes the rapid generation of high-quality mesh units for the deformation of the aircraft airfoil section. This greatly reduces the time and computing power required for aircraft airfoil optimization.

Aspects and features of the present disclosure relate to providing injective three-dimensional (3D) deformations based on two-dimensional (2D) mesh deformations. For example, a method involves defining at least one 2D mesh deformation based on a designated position of an object represented by an input neural radiance field (NeRF). The method also involves applying the 2D mesh deformation(s) to a 3D piecewise-linear map that operates over a plane and preserves a normal direction to produce prismatic maps. The method further involves composing a 3D deformation for the object from layers defined by the prismatic maps, and parameterizing the 3D piecewise-linear map. The method additionally involves storing or rendering, using the 3D piecewise-linear map, a deformed NeRF injectively representing the object in the designated position. Aspects also include computer systems, apparatus, and computer programs configured to perform the method.

This invention belongs to the field of automotive parts manufacturing, specifically involving a simulation system, method, equipment, and medium for evaluating the impact of materials on wheels. Based on a target wheel test bench, a corresponding simulation model is established; several sets of material properties, yield strength and tensile strength, are input to generate SN curves; corresponding stress criteria are generated based on the SN curves; the corresponding weight is obtained, optimization results are obtained, and the optimization results are evaluated. This invention converts the wheel bending fatigue life into equivalent fatigue strength through SN curves, and uses high-order mesh deformation technology to quickly and efficiently evaluate the impact of yield strength and tensile strength on wheel weight. It provides a scientific and accurate simulation prediction method for evaluating the impact of material properties on wheel weight, with broad application prospects and market value. This addresses the problem of difficulty in quickly and accurately evaluating the quantitative impact of new materials on wheel weight under bending fatigue conditions.

The present invention relates to a spherical grid hoisting construction control method and system based on overall positioning, which relates to the field of hoisting control. The method pre-collects a welding parameter array to perform welding deformation analysis, utilizes a hoisting equipment array to hoist according to preset parameters and performs overall positioning monitoring at a specific time, analyzes grid deformation in combination with welding deformation parameters, and then optimizes hoisting parameters to achieve precise hoisting of the spherical grid. The method solves the technical problems of difficult installation, difficult precision assurance, high safety risks of high-altitude operations, and long construction period in the hoisting construction of the spherical grid, thereby achieving the technical effects of improving hoisting precision, shortening construction period, reducing safety risks, and ensuring structural stability.

The invention discloses a moving femtosecond laser processing morphology prediction method considering light spot evolution, and particularly relates to the field of femtosecond laser processing. Comprising the following steps: establishing a three-dimensional geometric model of a to-be-processed material, and performing mesh generation on the three-dimensional geometric model to obtain a mesh model; establishing a dual-temperature model; determining laser intensity; determining a laser source item according to the laser intensity, and inputting the laser source item into the double-temperature model to obtain lattice temperature and electron temperature; when the lattice temperature is greater than or equal to the sublimation temperature, determining a grid deformation speed, removing a material of the grid model according to the grid deformation speed to obtain a deformed grid model, and obtaining an ablation depth; in the next pulse, the radius of the light spot is updated; and the lattice temperature, the electron temperature and the ablation depth are updated until the lattice temperature and the electron temperature are balanced, and the final ablation depth is obtained. Based on the method, the prediction accuracy can be improved.

This invention relates to the fields of remote sensing and geographic information engineering technology, and particularly to a land use change identification method driven by multi-source spatial data. The method includes: acquiring multi-source spatial data covering a target area and performing standardized preprocessing to generate a priori land use map as the data assimilation background field; constructing a land use abundance change process model based on material conservation, and obtaining a spatial discrete framework through adaptive grid deformation; establishing a multi-source differentiable observation operator, and solving the core parameters through four-dimensional variational assimilation with land use conversion constraints to complete land use change identification and rolling dynamic monitoring. In this invention, land use change is modeled as a continuous physical process, overcoming the limitations of traditional discrete-time hard classification and comparison methods. This effectively improves the temporal coherence of change identification and the consistency of multi-source data fusion, meeting the needs of high-precision long-term land use dynamic monitoring, and providing stable and reliable data support for natural resource surveys, monitoring, and land spatial management.

The present invention discloses a crane monitoring method, computer equipment, and storage medium based on digital twins, which relate to the field of crane technology. The method includes: constructing a crane arm grid model; inputting real-time dynamic data into a pre-trained RBF proxy model to output real-time coarse grid deformation data and real-time coarse grid stress data of coarse grid nodes; coloring the crane arm grid model according to the real-time coarse grid deformation data or the real-time coarse grid stress data to generate a crane arm model with a cloud effect; processing a pre-constructed crane arm proportional three-dimensional model according to real-time component relationships to generate a prototype model without a cloud effect; and constructing a stress-strain digital twin model of the crane arm based on the crane arm model and the prototype model. By adopting the present invention, the changes in the crane's geometric shape and stress-strain state during operation can be accurately understood, ensuring the safe operation of the crane and reducing safety hazards during construction.

The invention provides a finite element grid rapid reconstruction method based on linear elastic mechanics, which comprises the following steps of: after all node displacement vectors are obtained, superposing all node displacement vectors to all corresponding node coordinates before deformation to obtain all node coordinates after deformation: in the formula, the node coordinates after deformation and the node coordinates before deformation are shown in the description; keeping the corresponding relation between the units and the node numbers unchanged, and modifying node coordinates in the finite element solving file to complete finite element grid reconstruction; calculating a Jacobian determinant, a length-width ratio and a warping factor index of the reconstructed grid, and judging whether Jacobian determinants gt of all units are reconstructed or not; 0.5, the length-width ratio is lt; 5, a warping factor lt; and when 6.5, if yes, determining that the finite element mesh meets the use requirement. According to the method, a physical and geometric dual-drive grid deformation mechanism is established, global displacement distribution is restrained through a linear elastic mechanical constitutive equation, deformation of non-reasonable units is restrained, and the problem of unreasonable topology caused by a traditional geometric interpolation method is solved.

The application discloses a wing robust optimization design method based on Hessian eigenvector dimension reduction, and belongs to the technical field of aircraft design optimization, and comprises the following steps: determining initial design variables corresponding to an initial geometric shape of a wing; updating a wing geometric shape, and then adjusting a fluid calculation grid by using a grid deformation module; performing dimension reduction on an uncertainty space based on a Hessian eigenvector, and calculating aerodynamic performance data and a first-order gradient of sample points; inputting the aerodynamic performance data and the first-order gradient of the sample points into a UQ module, and calculating gradient information of the initial design variables; and judging, by an optimizer, whether a wing configuration determined by the initial design variables is convergent. The application reduces the calculation cost and improves the calculation efficiency while ensuring the accuracy, and provides an efficient technical path for engineering-level aerodynamic robust design under high-dimensional uncertainty conditions.

The invention discloses a three-dimensional object deformation method and system based on semantic driving, and belongs to the technical field of computer graphics and three-dimensional shape processing. The method comprises the following steps: firstly, acquiring a three-dimensional model and a target text, extracting features by using a cross-modal semantic model, and driving a grid to deform; and then binding the Gaussian point cloud to a deformation grid, dynamically optimizing geometry by means of a diffusion model multi-view rendering signal, adjusting the Gaussian color through a texture repair model, and finally outputting a three-dimensional shape with consistent geometry and texture. The whole process does not need manual annotation, semantic deformation of any object can be achieved, and the universality is high. According to the method, the structural integrity and multi-view consistency of the object can be kept, and the texture can be automatically repaired, so that the three-dimensional content editing efficiency and the reality sense of an output result are remarkably improved.

A method and associated apparatuses for optimizing a sampling scheme which defines sampling locations on a bonded substrate, having undergone a wafer to wafer bonding process. The method includes determining a sampling scheme for a metrology process and optimizing the sampling scheme with respect to a singularity defined by a large overlay error and / or grid deformation at a central location on the bonded substrate to obtain a modified sampling scheme.

The present invention is a blade parameterization method based on fluid-solid mesh collaborative deformation, which belongs to the technical field of aero-engine blade shape parameterization. In view of the problems of large computational complexity and low computational efficiency in current mesh deformation methods, the method of the present invention includes constructing an initial fluid domain mesh and a solid domain mesh adapted to the blade model, introducing an O-type domain in the solid domain mesh, using a set of blade points at different blade heights on the blade as blade control points, and using Hicks-Henne type functions, circumferential offsets combined with radial basis functions to obtain the contour changes and stacking rule changes of the blade respectively, and evenly spreading the offset increments of the control points to the entire blade surface. Based on the idea of ​​structural mesh orthogonality, the deformation of the blade surface mesh is transmitted to the far end along the orthogonal direction. Compared with the traditional blade parameterization method, the present invention not only realizes the two deformation modes of fluid-solid domain blade contour and stacking rule at the same time, but also improves the deformation efficiency.

The invention discloses a tensioning type carrier for an electrode grid production line, which comprises fixed parts, movable parts, elastic parts and adjusting assemblies, the fixed parts and the movable parts are arranged in a layered manner, the movable parts and the fixed parts are connected in a sliding manner, the fixed parts are provided with limiting assemblies and a plurality of convex blocks, and the elastic parts are connected with the limiting assemblies. The limiting assembly is used for limiting the periphery of the grid; a plurality of stand columns are arranged on the movable part, and clamping grooves used for containing arc-shaped single lines are formed between the stand columns and the corresponding protruding blocks. Two ends of the elastic member are respectively connected with the fixed member and the movable member. The carrier adopts a double-layer panel structure, a clamping groove gap is reduced by switching a moving part in an adjusting assembly from a narrow part to a wide part and utilizing pulling force generated when an elastic part resets, so that the effect of pressing a grid is achieved, and during pressing, the moving part makes contact with the inner wall of the outer side of the wide part, the movement amount of the moving part is limited, the pressing force is guaranteed, and meanwhile the clamping groove gap is reduced. And grid deformation or damage caused by overlarge elastic force is avoided.

The invention provides an additive manufacturing-oriented shape follow-up honeycomb structure optimization design method. The design method comprises the following steps of (1) honeycomb lattice parameterization design: carrying out honeycomb lattice design based on additive manufacturing cantilever angle constraint and powder cleaning requirements; (2) honeycomb filling design is conducted, specifically, based on additive manufacturing cantilever angle constraint and powder clearing requirements, plane cutting is conducted on a regularly-arranged honeycomb structure, and honeycomb filling layout design is achieved; (3) grid deformation intelligent modeling: mapping the plane honeycomb filling structure to a shape follow-up honeycomb filling structure by using an RBF interpolation algorithm to obtain a shape follow-up honeycomb finite element model; and (4) optimizing the design variables: constructing a target and constraint function, updating the design variables by using a global or local optimization algorithm, and obtaining an optimal conformal honeycomb configuration. According to the technical scheme, the technical problem that the honeycomb structure design cannot adapt to the complex special-shaped structure of the aircraft in the prior art can be solved.

The application discloses a kind of grid deformation data enhancement methods based on WGAN-GP model, it is related to image processing, data enhancement technical field, including: constructing training dataset and test dataset;WGAN-GP model is constructed;WGAN-GP model is trained based on training dataset;The performance of model is evaluated using test dataset, determine grid deformation data enhancement model;Random noise is respectively used as input with medical endoscope image, output corresponding deformation grid, respectively with medical endoscope image and deformation grid warp operation obtains the medical endoscope image after data enhancement.This application constructs WGAN-GP model and automatically generates deformation grid, by deformation grid to medical endoscope image is carried out deformation enhancement, enhanced medical endoscope image is real and has certain diversity, effectively solve the problem of data sample scarcity and data sample unbalance, to improve the precision and generalization of artificial intelligence algorithm model based on medical endoscope image.

The invention provides a local updating system and method for a high-precision map of a strip mine, and the system comprises six modules: a multi-source data collection module, an intelligent change detection module, an incremental three-dimensional reconstruction module, a space-time database management module, a visual verification module and a mining and stripping amount measurement module. PointNet + + and U-Net networks are fused to realize change region extraction with the precision less than or equal to 1m, only local reconstruction is carried out on the change region, centimeter-level seamless fusion is realized by adopting graph optimization and Laplacian grid deformation, and updating is completed in a short time. The model supports version second-level rollback, WebGL30FPS rendering and TIN method automatic amount calculation, the error is less than 1.2%, the bandwidth is saved by 90%, the GPU occupation is reduced by 75%, the requirement of an unmanned mine card for the current trend less than or equal to 30 minutes under the pace of daily propulsion of a strip mine is met, mining while updating is realized, and a real-time data base is provided for intelligent scheduling and digital twin mines.

The invention discloses a machine learning-based low-altitude air route time slot distribution method and system. The method comprises the steps of constructing a four-dimensional space-time grid, dividing voxel units, collecting an airspace state and forming a feature sequence; based on the state feature sequence, constructing a voxel space-time diagram containing space and time edges; fusing intrusion time coding, inputting the improved STSGCN model, and generating a risk and time sequence; constructing a reversible reasoning link, and executing reverse time sequence reasoning to obtain a reverse time sequence; generating oscillation parameters by using the time sequence, and forming a dynamically adjusted time slot adjustment sequence; and constructing an airspace elastic field, executing grid deformation, generating an elastic grid and outputting a scheduling result. According to the invention, through combination of improved STSGCN prediction and reversible predicted intrusion time reasoning with time slot self-adjustment and airspace elastic grid updating, dynamic safe distribution of low-altitude airway time slots and airspace capacity improvement are realized.