50 results about "Adaptive mesh refinement" patented technology

The invention relates to the technical field of oil exploitation, and provides a CFD-based electric submersible pump oil-gas-water three-phase flow adaptability analysis method and system, and the method comprises the steps: constructing a three-dimensional geometric model of an electric submersible pump, carrying out the initial grid division through CFD software, and optimizing the grid density through a self-adaptive grid refinement algorithm; setting oil well boundary conditions, a turbulence model and cavitation model parameters; steady-state and unsteady-state simulation calculation is carried out, and a flow field distribution diagram in the electric submersible pump is obtained; key performance parameters are extracted; simulating flow characteristics of oil-gas-water three-phase flow by adopting a turbulence model optimization algorithm based on multiphase flow coupling and combining a turbulence model, a multiphase flow model, an interphase turbulence correction term and an interphase interaction force model; and carrying out adaptability analysis on the oil-gas-water three-phase flow of the electric submersible pump. All-around technical support can be provided for design, operation and maintenance of the electric submersible pump, and technical progress and development of the oil exploitation industry are powerfully promoted.

The invention relates to the technical field of mine safety and geological environment evaluation, in particular to a mine goaf geological environment dynamic evaluation method based on digital twinning. The method comprises the following steps: constructing a digital twinborn model of a mine goaf, designing a data structure and a virtual-real mapping rule, collecting multi-source heterogeneous data, and performing spatio-temporal data indexing processing to form a spatio-temporal associated data stream; microseismic travel time data are extracted from the space-time associated data stream; performing parallel ray tracing and rapid inversion on the microseismic travel time data to obtain a wave velocity distribution initial value; carrying out wave velocity field self-adaptive grid refinement on the micro-seismic travel time data by utilizing the wave velocity distribution initial value to obtain a multi-resolution wave velocity field; and carrying out wave velocity change rate analysis on the multi-resolution wave velocity field to obtain a wave velocity variability tensor. According to the method, the abstract energy transfer process is concrete into a data field capable of being quantitatively analyzed, so that the geological environment of the goaf is accurately and dynamically evaluated.

The invention discloses a pressure reconstruction method and system based on error correction and an adaptive grid, and a storage medium. The system comprises a PIV experiment test subsystem, a processing module (comprising a PIV post-processing unit and a pressure reconstruction unit) and the like. The method comprises the following steps: firstly, controlling a laser emitting sheet to irradiate tracer particles in a water tank through a synchronizer, and simultaneously triggering a high-speed camera to capture a particle image sequence at a microsecond-to-millisecond-level time interval; the PIV post-processing unit performs cross-correlation analysis on the image to generate a gridding velocity field; a pressure reconstruction unit solves a pressure Poisson equation based on the velocity field to obtain a preliminary pressure field, then iteratively corrects an error through an error correction mechanism and gradient correlation optimization, and triggers self-adaptive mesh refinement / coarsening according to a pressure / velocity gradient amplitude; and finally, outputting a high-precision pressure field on the optimized grid. The method can be applied to flow field tests and hydrodynamic performance analysis in ship and ocean engineering, especially pressure field reconstruction in a complex flow field, and has a wide application prospect.

The invention relates to the cross technical field of calculation of solid mechanics and deep learning, in particular to a self-adaptive mesh refining method, system and equipment for a physical information neural network, and the method comprises the steps: employing a coarse mesh physical information neural network model to discretize a mechanical problem, and building a mapping relation between node coordinates and node displacement; constructing a physical information function and a total loss function of the solid mechanics problem, and training the coarse mesh physical information neural network model to obtain a total loss function value and a change rate of the solid mechanics problem; if the coarse grid physical information neural network model meets a solving condition or the number of training times is greater than a preset number of times, outputting a displacement solution of the solid mechanics problem, otherwise, selecting a grid unit with a relatively large strain energy density value in the coarse grid physical information neural network model for refining to obtain a refined grid model; and migrating parameters of the coarse mesh physical information neural network model to the refined mesh model, and repeatedly executing the steps until a high-precision solution of the solid mechanics problem is output.

The invention discloses a multi-objective optimization solution method for structural topological optimization software, which relates to the technical field of structural topological optimization, and comprises the following steps of: importing a geometric model of an engineering structure into topological optimization software, discretizing the geometric model through a self-adaptive mesh refinement technology, and solving the discretized geometric model; solving the constructed multi-objective optimization model by adopting an efficient finite element solver to obtain stress distribution data and displacement distribution data, and carrying out optimization solution through a parallel computing multi-population particle swarm algorithm to obtain a Pareto optimal solution set; and performing multi-criterion decision analysis on the Pareto optimal solution set by adopting an analytic hierarchy process, determining a topological structure scheme, and feeding back the topological structure scheme to the structure topological optimization software to complete multi-objective optimization solution. Through the above steps, objective discrimination and scientific screening of the optimal topological structure scheme are realized, the optimal solution is transmitted back to topological optimization software through a feedback mechanism, and the practicability of the optimization result and the engineering implementability are improved.

The invention provides a switch cabinet insulation structure optimization design and simulation method based on thermal resistance analysis, and relates to the technical field of power equipment, and the method comprises the steps: recognizing a heat transfer path through thermal resistance analysis, decoupling temperature rise control and an insulation distance target into temperature field uniformity and insulation margin indexes, constructing a dual-index agent value through employing a self-adaptive fitting method, and carrying out the simulation of the dual-index agent value. Optimal structure parameters are obtained through optimization of the improved ant colony algorithm, local encryption is carried out on the high-gradient area in combination with the self-adaptive mesh refinement technology, and accurate thermal field simulation is achieved. The design efficiency, safety and reliability of the insulation structure of the switch cabinet are improved.

The invention discloses a three-dimensional fire modeling method based on a lattice Boltzmann method, and the method comprises the following steps: building a fire model which comprises a calculation module and a self-adaptive mesh refinement module, the calculation module comprises a lattice Boltzmann flow module, a buoyancy driving module, a combustion module based on gradient dynamic triggering, an energy equation module, a flue gas transportation module containing particle sedimentation and visibility calculation, and a multi-component transportation module. According to the fire modeling method based on the lattice Boltzmann method, LBM flow field data serves as core drive and is transmitted to the combustion module, the flue gas transportation module and the multi-component transportation module in real time, accurate convection and diffusion information is provided, meanwhile, the self-adaptive grid refining module dynamically adjusts the grid resolution according to the gradient of key physical quantities, and therefore the fire modeling accuracy is improved. The accuracy can be ensured, the calculation amount is reduced, and the calculation speed is increased. And the prediction precision, robustness and real fire complexity capturing capability of the model are obviously improved.

The invention provides a hyperbolic variable cross-section construction process optimization method, and belongs to the technical field of building construction.The method includes the steps that precise geometric data of a hyperbolic variable cross-section structure is obtained to establish a geometric parameter matrix, and a high-precision finite element analysis model is established through a self-adaptive mesh refinement algorithm; establishing a double-layer game grid optimization model to realize coordination and balance of calculation precision and efficiency, establishing a prestressed tendon three-dimensional path planning model, optimizing a spatial arrangement path of the prestressed tendon in a hyperboloid by adopting a genetic algorithm, accurately calculating a curvature radius and a friction loss coefficient of the prestressed tendon, and adjusting arrangement parameters according to a prestress loss rate. And a real-time monitoring system is established to dynamically adjust construction parameters, and the technical problem of low prestress transmission efficiency caused by complex spatial path planning of the prestressed tendons in the construction process of the hyperbolic variable cross-section structure is solved.

The invention discloses a water flow numerical simulation self-adaptive grid block encryption method based on a quadtree spatial data structure, and relates to the field of simulation and analysis of hydrodynamic problems in water conservancy projects, and the simulation of water flow movement is calculated through a grid Boltzmann method. The self-adaptive encryption is realized by a self-adaptive grid refining technology based on a quadtree spatial data structure. The dam break wave flowing phenomenon in the dam break process is simulated, the water volume of a calculation area is constant, the upstream and downstream water levels are different before complete dam break, the upstream and downstream water flow flowing and self-adaptive grid encryption conditions in the instantaneous complete dam break process are simulated and analyzed, the real-time display function of the calculation result is achieved, and the calculation accuracy is improved. Through reasonable parameter setting, the speed field, the pressure field, the grid number and other conditions obtained through calculation can be displayed on a screen in real time under the condition that the calculation speed is not affected. A user can conveniently adjust calculation parameters according to a calculation intermediate result, and simulation calculation can be continued according to a preset target.

The invention discloses a side slope excavation dynamic monitoring and early warning method based on digital twinning. The method comprises the following steps that terrain point cloud, rock and soil layering and construction progress are obtained and preprocessed; three-dimensional geometric twinning and physical twinning are established, finite element simulation is adopted, boundary push-back and pore pressure rebalance are completed, and a twinning base line is formed; an observation residual error and a physical conservation residual error are calculated, and a high-consistency twin and assimilation sample set is obtained through joint updating; applying a topological constraint sliding surface evolver and implementing adaptive mesh refinement and interface reconstruction of a critical zone; and constructing a displacement and osmotic pressure uncertainty conical domain based on the assimilation sample set and the critical region strengthening model, and outputting a grading early warning result. According to the method, a double-consistency assimilation and topological constraint sliding surface evolution method is adopted, high-consistency simulation and quantitative early warning of the whole slope excavation process are achieved, and the method has the advantages of being high in accuracy, high in stability and good in traceability.

This application discloses a method and system for estimating the direction of arrival (DOA) of weak targets based on Riemannian manifold background suppression and adaptive sparse Bayesian learning. The method includes: acquiring time-series signals using a sensor array to construct a series of sample covariance matrices, mapping them to a point sequence on a Hermitian positive definite matrix manifold space; iteratively calculating the background interference covariance matrix using the non-Euclidean geometric properties and logarithmic shielding effect of the Riemannian metric; mapping the background interference covariance matrix back to Euclidean space, adaptively performing background subtraction based on an energy decision mechanism to reconstruct a positive definite covariance matrix to be measured; inputting the covariance matrix to be measured into a sparse Bayesian learning framework, first iteratively recovering the signal power through adaptive mesh refinement sparse Bayesian learning, then performing a closed-loop iteration of subspace noise cleaning while keeping the mesh fixed to recover the sparse spatial spectrum of the target signal; and finally, using local analytical interpolation techniques to eliminate mesh quantization errors and calculate the precise DOA of the target.

The application relates to a large eddy simulation-based atmospheric flow field reconstruction method, which comprises the following steps: acquiring turbulent flow data of a target region and performing normalization processing; inputting the normalized turbulent flow data into a resolution promotion module to extract high-resolution time characteristics and predict time characteristics at the next moment, and acquiring reconstructed high-resolution data, wherein the resolution promotion module is obtained through training of turbulent flow data and corresponding resolution-promoted turbulent flow data; performing numerical simulation on the reconstructed high-resolution data by using a large eddy simulation model to construct a multi-layer resolution grid structure; and if the resolution accuracy of a local region in the multi-layer resolution grid structure is lower than a preset threshold, adjusting the grid density through an adaptive grid refinement technique to acquire a three-dimensional atmospheric flow field reconstruction result. The application improves the accuracy and time continuity of flow field simulation under complex terrain.

The present application relates to the technical field of safety monitoring and industrial automation, and discloses a silo safety monitoring method and system, which synchronously acquires coal pile surface point cloud, lateral pressure, vertical load and environmental data through the arrangement of a laser scanning unit, an annular pressure sensing belt, a weighing sensor and a temperature and humidity node; a high-precision three-dimensional morphology is reconstructed based on an adaptive grid refinement algorithm, and a joint representation model is constructed by fusing mechanical data; a stability discrimination function is established in combination with coal body physical parameters to identify the instability critical region; and through pattern matching with a historical stability morphology library, hierarchical early warning of abnormal working conditions such as suspended arch, bonding and blockage and generation of blockage removal strategies are realized. The present application constructs a joint representation model of coal pile three-dimensional morphology and mechanical state by fusing laser point cloud, pressure distribution and load sensing data, breaks through the limitation of traditional single-point material level measurement which cannot reflect the overall coal pile structure, and realizes high-precision measurement with a storage calculation error of less than 2%.

This invention discloses a method and system for imaging complex geological bodies based on adaptive grid partitioning. The method includes: intelligent sensors acquiring multi-physics observation data of the target area and outputting time stamps, data quality markers, and discrete information on observation directions; constructing an initial geological skeleton and an initial grid model based on the multi-physics observation data; generating interface protection zones and identifying boundary offset corridors; establishing evidence accumulation states to filter cells to be updated; determining the grid partitioning direction and performing adaptive grid partitioning; performing topology locking review; and then performing grid updates and imaging updates to obtain the imaging results of the complex geological body. This method solves the problems of unstable grid updates near complex interfaces, misalignment of multi-physics boundaries, and easy loss of deep anomalies in multi-scale imaging.

The invention provides a numerical simulation method for fluid-structure interaction vibration of wind tunnel airflow and a tunnel wall structure, and belongs to the technical field of wind tunnels. Geometric parameters of a wind tunnel section are obtained, a multi-layer grid sequence is established, an implicit Runge-Kutta format and an explicit Newmak method are adopted to solve a flow field and a structure field respectively, and the wind tunnel airflow and tunnel wall structure fluid-structure interaction vibration is obtained. Mesh deformation is realized by using tightly-supported radial basis function interpolation, a volume conservation constraint is introduced, an interface projection matrix is constructed based on a virtual work equivalence principle to transmit interface data, a force balance and displacement coordination constraint is introduced, and coupling iteration convergence is accelerated through an inverse Broyden formula. A density gradient shock wave sensor is adopted to drive grid adaptive encryption and a mixed indicator function to control turbulence model switching, and the technical problems that wind tunnel section airflow and wall surface structure fluid-structure interaction vibration numerical simulation calculation is low in efficiency and insufficient in precision are solved.

The invention discloses an adaptive mesh refinement method for a non-uniform soft tissue mesh model, which belongs to the technical field of computer graphics, and comprises: generating an eye movement heat map according to eye movement data of a user gazing at a primary tetrahedral mesh model; extracting a primary refinement area of ​​the primary tetrahedral mesh model according to the eye movement heat map, and obtaining a vertex density value of the primary tetrahedral mesh model; performing primary refinement on the primary refinement area based on the vertex density value to obtain a secondary tetrahedral mesh model; extracting a secondary refinement area from the secondary tetrahedral mesh model according to the conduction range of an external force; performing secondary refinement on the secondary refinement area based on the vertex density value to obtain a third tetrahedral mesh model adapted to the non-uniform soft tissue; wherein the primary tetrahedral mesh model is constructed and obtained through a CT image of the non-uniform soft tissue; and the authenticity and real-time performance of the simulation are improved.

The present invention discloses a flood control safety evaluation system for dike projects, which relates to the technical field of flood control safety. The initial position and boundary conditions of dike breaches are set through a simulation scenario construction module, and the MIKE 21 model is used to simulate the process of dike failure and breach formation; the data acquisition module records the breach boundary and water flow dynamic data in real time to provide support for numerical stability analysis; the numerical stability analysis module evaluates the stability of the model based on the dynamic data and identifies unstable areas; the instability adjustment module adopts an adaptive grid refinement technique to optimize the model calculation by dynamically adjusting the grid division; finally, the optimized model conducts a risk assessment of dike breaches through a risk analysis report generation module to generate an accurate risk analysis report. This technology improves the prediction accuracy and numerical stability in the process of dike breach simulation, and significantly enhances the reliability of risk assessment.

This invention belongs to the field of laser processing and numerical simulation, and particularly relates to a finite element modeling and analysis method for nanosecond pulsed laser-impacted target materials. The method includes: establishing and loading a laser heat source model; establishing and meshing a finite element model of the target material; establishing and solving the transient heat conduction equation; multi-period iterative solution and post-processing; model verification and parameter analysis. By establishing a laser heat source model, employing an adaptive mesh refinement strategy optimized for temperature gradients, and combining a transient solution algorithm, this invention significantly improves the prediction accuracy of the spatiotemporal evolution of the target material temperature field under nanosecond pulsed laser irradiation compared to traditional simplified analytical models or general commercial software simulations. This invention can accurately predict the temperature distribution and evolution of the target material under laser irradiation, and has advantages such as high modeling accuracy, good computational efficiency, and wide applicability. It is suitable for engineering applications such as laser processing and damage assessment of optical and composite materials.

This invention discloses a multi-feature adaptive mesh generation method based on GPU-accelerated multi-scale FDTD. By constructing a two-layer feature representation framework and combining it with GPU-accelerated BVH ray tracing technology, this invention can intelligently identify key electromagnetic regions and achieve adaptive mesh refinement. Compared with existing uniform mesh methods, while maintaining high-fidelity simulation accuracy, it can significantly reduce the number of meshes by 33% to 51% and reduce computation time by 52% to 70.1%. Simultaneously, a four-stage spatial connectivity algorithm ensures the robustness and numerical stability of non-uniform meshes, effectively avoiding numerical reflections caused by mesh abrupt changes, and solving the problem of balancing efficiency and accuracy in electromagnetic simulation of complex industrial models.

This invention provides a method and system for predicting the state of a substation affected by stray current fluctuations. Based on train operation data, a digital twin traction power supply system model integrating train dynamics and traction arc characteristics is used to generate a stray current disturbance signal containing transient pulse components. Based on earth resistivity distribution data and the stray current disturbance signal, a three-dimensional finite element underground stray current diffusion model with adaptive mesh refinement is used to obtain the current component flowing into the substation grounding grid. Based on the current component and substation operation data, a digital twin model of the substation's electromagnetic-thermal coupling, explicitly modeled with the core lamination structure and oil-cooled convection channel, is used to make predictions, obtaining the substation's bias magnetization prediction results and temperature rise prediction results. This invention comprehensively considers multi-source heterogeneous data and combines simulation technology for the entire chain of subway traction, earth, and power grid substations, which can improve the risk prediction results of substations.

In order to deeply explore the dynamic response mechanism and the elasticity resistance of the composite target plate under the impact of the projectile, the missile-target action finite element model of the composite target plate is based on all components of the composite target plate, combines the dynamic characteristics of the impact of the projectile, reasonably divides grid units and selects an adaptive material constitutive model, so that the missile-target action finite element model of the composite target plate is obtained. And deformation, damage and failure behaviors of different materials under the high-speed impact load are accurately described. In a model building process, a nonlinear behavior of a missile target contact interface is emphatically considered, a penalty function method and a Lagrangian multiplier method are combined to process contact state change in a projectile penetration process, and meanwhile, an adaptive mesh refinement technology is introduced, so that the operation efficiency is improved on the premise of ensuring the calculation precision; through comparison with related experimental data (such as projectile residual velocity, armor damage mode and the like), the calculation result of the model is highly matched with an experimental value, which indicates that the dynamic mechanical response rule of the composite target plate under the impact of the projectile can be effectively reflected.

The present application provides a rail transit material combustion smoke density dynamic monitoring method and system, relating to the technical field of rail transit, comprising collecting smoke concentration data of multiple sensors in a combustion test box, using a three-dimensional space interpolation algorithm of adaptive grid refinement for three-dimensional reconstruction, combining material combustion characteristic parameters to construct a combustion kinetics model, fusing high-precision three-dimensional distribution map and the kinetics model, constructing a multi-scale diffusion prediction model based on graph neural network, generating smoke density distribution and diffusion path prediction data at future time, so as to realize accurate monitoring and early warning of rail transit carriage smoke density.

The invention relates to the technical field of safety monitoring and industrial automation, and discloses a silo safety monitoring method and system.The method comprises the steps that a laser scanning unit, an annular pressure sensing belt, a weighing sensor and a temperature and humidity node are arranged, and coal pile surface point cloud, lateral pressure, vertical load and environment data are synchronously obtained; a high-precision three-dimensional form is reconstructed based on an adaptive mesh refinement algorithm, and mechanical data are fused to construct a joint representation model; establishing a stability discrimination function in combination with coal physical parameters, and identifying an instability critical region; through mode matching with a historical stable form library, graded early warning and blockage clearing strategy generation of abnormal working conditions such as suspension arching, bonding and blockage are realized. According to the method, the laser point cloud, the pressure distribution and the load sensing data are fused, a combined characterization model of the three-dimensional form and the mechanical state of the coal pile is constructed, the limitation that traditional single-point material level measurement cannot reflect the whole coal pile structure is broken through, and high-precision measurement with the reserve calculation error smaller than 2% is achieved.

The present application belongs to the field of computational materials science and scientific computing technology, and relates to a self-adaptive grid generation method in material simulation, comprising: S1, generating an atomic region grid and a continuous region grid of a three-dimensional material, and splicing the grids to obtain a coupled grid; S2, optimizing the grid by using a local self-adaptive grid refinement strategy to obtain a grid; S3, detecting whether the calculation accuracy of the material under the grid meets a preset, and if not, executing step S4; S4, expanding the atomic region grid of the grid to obtain a grid; S5, eliminating elements intersecting with the grid in the continuous region grid of the grid to obtain a grid; S6, constructing a cavity transition grid, splicing the grid with the cavity transition grid to obtain a grid; S7, judging whether a convergence condition is reached, and if yes, obtaining a final coupled grid; otherwise, optimizing the grid and returning to step S4; the present application automatically identifies and connects the atomic region and the continuous region with high quality in three-dimensional geometry.

The application relates to the technical field of oil exploitation, and provides a CFD-based adaptability analysis method and system for oil-gas-water three-phase flow of a submersible electric pump, which comprises the following steps: constructing a three-dimensional geometric model of the submersible electric pump and performing initial mesh division through CFD software, optimizing the grid density through a self-adaptive mesh refinement algorithm; setting oil well boundary conditions, a turbulent flow model and cavitation model parameters; performing steady-state and non-steady-state simulation calculation to obtain a flow field distribution diagram of the submersible electric pump; extracting key performance parameters; adopting a turbulent flow model optimization algorithm based on multiphase flow coupling, combining the turbulent flow model, the multiphase flow model and an interphase turbulent flow correction term and an interphase interaction force model to simulate the flow characteristics of oil-gas-water three-phase flow; and performing adaptability analysis on the oil-gas-water three-phase flow of the submersible electric pump. The application can provide all-round technical support for the design, operation and maintenance of the submersible electric pump, and powerfully promotes the technical progress and development of the oil exploitation industry.

This invention discloses a method and apparatus for predicting and controlling error propagation in computational fluid dynamics (CFD), belonging to the field of CFD numerical simulation technology. Addressing the problems of existing technologies' difficulty in identifying error propagation paths in advance and the lack of proactive control measures, this invention constructs an error propagation state characterization model, representing the propagation behavior of numerical errors under physical mechanisms such as convection, diffusion, and reaction as a combination of multiple error propagation modes; it predicts the error propagation velocity, direction, and propagation length by combining local flow field physical parameters; it constructs an error propagation risk criterion to identify potential error growth regions; and it implements adaptive strategies such as local time step adjustment, adaptive mesh refinement, enhanced numerical scheme stability, and rigid control of the reaction zone. This invention can identify and proactively control unstable regions before significant error amplification, significantly improving the stability and reliability of numerical simulations of complex flows and combustion.

This invention discloses a finite element method, storage medium, and computer program product for the microstructural evolution of rock masses based on material heterogeneity mapping. The method includes data acquisition; parameter inversion; mesh construction; coordinate registration; weighted statistics; assigning equivalent parameters to elements; adaptive mesh refinement in high-gradient regions; model initialization and solution: initial field loading and boundary condition definition are completed, and elastoplastic or multi-field coupled calculations are performed. The advantage of this invention is that it improves the predictive ability of the finite element model for the non-uniform deformation and microstructural evolution of rock masses.

This invention discloses a numerical simulation method for chloride ion diffusion in recycled aggregate concrete based on a five-phase random aggregate model, belonging to the field of numerical simulation technology for building material durability. This invention solves the problems of low modeling accuracy of the microstructure of recycled concrete, fragmented modeling and assignment processes, and the inability to dynamically optimize in existing technologies. The technical solution includes: parametrically defining a five-phase model consisting of a core-old ITZ-old mortar-new ITZ-new mortar; using a piecewise integration method to achieve precise control of aggregate gradation; automatically constructing multi-layered annular geometry through Boolean operations; synchronously generating explicit selection sets for each phase; performing transient simulation of chloride ion diffusion, identifying high-concentration gradient regions based on error analysis; automatically backtracking and dynamically adjusting the selection set partitioning, assigning diffusion coefficients based on local diffusion characteristics, and performing adaptive mesh refinement; and continuing subsequent solving. This invention achieves an intelligent closed loop of "modeling-solving-refinement-resolving," significantly improving simulation accuracy and efficiency, and providing an efficient and accurate numerical tool for the durability assessment of recycled concrete.

The invention provides a rail transit material combustion smoke density dynamic monitoring method and system, and relates to the technical field of rail transit, and the method comprises the steps: collecting the smoke concentration data of a plurality of sensors in a combustion test box, and carrying out the three-dimensional reconstruction through a self-adaptive grid refined three-dimensional space interpolation algorithm, the method comprises the following steps: constructing a combustion kinetic model by combining material combustion characteristic parameters, fusing a high-precision three-dimensional distribution diagram with the kinetic model, constructing a multi-scale diffusion prediction model based on a graph neural network, and generating smoke density distribution and diffusion path prediction data at future moments, thereby realizing accurate monitoring and early warning of the smoke density of a rail transit carriage.

The invention provides a thermal expansion stress monitoring method for a thermal energy storage system, and belongs to the technical field of thermal energy storage systems. A multi-physics field coupling sensor array is deployed, and a thermal expansion stress component and other load stress components are separated by adopting an empirical mode decomposition algorithm; the method comprises the following steps: establishing a self-adaptive grid refined finite difference thermal stress prediction model, combining with a Kalman filtering algorithm to realize transient thermal shock stress prediction, simplifying stress component matrix calculation through spectral decomposition and a diagonalization frame, and establishing a stress threshold calculation equation set; intelligent switching of a standard operation mode, a stress monitoring enhancement mode and a stress adaptive adjustment mechanism is realized according to a comparison result of the stress concentration coefficient and a threshold value, and the technical problem that thermal expansion stress and other load coupling stress cannot be accurately separated and predicted in the operation process of the thermal energy storage system is solved.