12 results about "Conservation law" patented technology

The invention discloses a flow field pressure gradient prediction method based on a graph attention network guided by a physical operator, and aims to rapidly and accurately predict the gradient distribution of a flow field. According to the method, a grid computational domain of computational fluid dynamics (CFD) is expressed as a graph structure, a graph neural network is adopted to carry out feature learning on grid nodes and adjacency relations, and a physical operator (such as a fluid control equation difference operator) is introduced in a model training process to carry out guide constraint on the network. Through the above technical scheme, the method can greatly improve the efficiency of flow field gradient prediction on the premise of ensuring the prediction precision and physical consistency, has the advantages of fast model calculation speed, strong adaptability to complex boundary conditions, and the prediction result meets the law of conservation of fluid mechanics, and is suitable for popularization and application. The method can be used for rapidly predicting the flow field gradient in the fields of aerospace fluid simulation, wind engineering and the like.

The invention provides an electric power system carbon potential tracking and predicting method based on physical and data dual drive. The method comprises the following steps: constructing a time sequence characteristic data set; constructing a source-side unit double-layer collaboration map, and generating a normalized adjacent matrix; constructing a physically guided source side graph neural network model, and outputting a predicted source side dynamic carbon emission factor and a node carbon injection amount sequence; constructing a high-dimensional input tensor through a space-time diagram attention mechanism; constructing a whole-network space-time diagram neural network model to realize dynamic tracking of carbon potential; the method comprises the following steps: introducing a Kirchhoff carbon flow conservation law as a physical regularization term in a space-time law deduction process, constructing a mixed loss function containing the physical regularization term, and executing an optimal sentinel mechanism in a back propagation process of whole-network space-time diagram neural network model training, and finally, outputting a dynamic carbon potential prediction result of the load side after physical verification. According to the invention, the precision and credibility of carbon potential prediction can be improved.

The application provides a physical mechanism coupling data-driven fast prediction method, and belongs to the technical field of disaster prediction.The application extracts an invariant topological feature vector through topological data analysis driven phase space reconstruction and inputs a physical constraint identification model to obtain a conservation law deviation vector and a physical feasible region boundary parameter, adopts a multi-resolution adaptive grid technology to perform multi-scale decomposition, extracts long-term evolution trend features and spatial local features through a full connection layer network and a convolution layer network, combines the physical feasible region boundary parameter to perform projection gradient descent iterative optimization to generate a preliminary prediction field, performs Bayesian uncertainty quantization, adjusts an artificial intelligence model regularization coefficient according to a disaster precursor identification function value, and outputs a graded early warning result, so that the technical problem of insufficient prediction accuracy caused by insufficient coupling of a physical constraint and a data-driven model in a disaster prediction system is solved.

This application relates to a satellite orbit prediction method. The method includes: constructing an expression for the gravitational force acting on the satellite based on its mass and distance from Earth; constructing a differential equation for the satellite's acceleration based on Newton's second law and the expression for the gravitational force; solving the differential equation based on the conservative force field characteristics of mechanics, calculating the gravitational potential energy function and its derivative with respect to time; transforming the derivative expression based on the mechanical conservation law that the sum of the derivatives of kinetic energy and gravitational potential energy with respect to time is zero, to obtain the derivative expression for acceleration; converting the time relationship in the derivative expression of acceleration to a radial position relationship based on the principle of conservation of angular momentum; integrating the radial position relationship to obtain the satellite's orbit equation. This method enables satellite orbit prediction.

The invention discloses a multi-modal fusion and reasoning enhanced boiler four-tube intelligent monitoring method and system, and the method comprises the steps: collecting unstructured asynchronous event flow data through a sensor group disposed on the inner wall of a boiler, carrying out the processing of the asynchronous event flow data, generating a sparse pulse feature tensor, and carrying out the processing of the sparse pulse feature tensor; the sparse pulse feature tensor is mapped to a high-dimensional potential space, the future physical state of the boiler system is dynamically predicted in the high-dimensional potential space, the physical conservation law constraint is introduced in the training process of an industrial world model, it is ensured that prediction of the future physical state conforms to hydromechanics and thermodynamic equations, and the prediction accuracy is improved. By predicting a set of states in a physical state parameter set, a future physical state is comprehensively predicted, so that an artificial intelligence system conforms to the pre-judgment capability of a physical law and obtains key dynamic details for multiple physical states, a perception-decision-action closed loop is realized, and the boiler monitoring efficiency is greatly improved.

The invention relates to the field of boiler prediction, and discloses a boiler real-time data prediction system which is used for improving the accuracy of boiler fault prediction. According to the boiler real-time data prediction system, intrinsic characteristics of combustion fluctuation, convective heat transfer, heat conduction and the like of a flow field in a boiler are extracted through computational fluid mechanics characteristic modal decomposition, a dynamic state evolution model is built by combining mass, energy and momentum conservation laws, and coupling prediction of a temperature field, a pressure field and a flow velocity field is achieved; and a non-equilibrium thermodynamic entropy generation principle is introduced to carry out physical consistency correction on a prediction result, so that high-precision prediction under the constraint of a thermodynamic law is ensured. And finally, mapping the risk indexes to a three-dimensional boiler model through a digital twinning technology, and generating an interactive early warning interface with spatial positioning and risk level visualization. According to the invention, a complete technical chain of data acquisition, feature analysis, state prediction and risk early warning is formed, and early accurate early warning and space positioning capabilities are provided for safe operation of the boiler of the power plant.

The invention provides a method for calculating the vibration velocity of surrounding rock of a water-sealed cave depot, and relates to the technical field of water-sealed cave depot engineering construction. The method comprises the following steps: constructing a vibration velocity calculation formula of water-sealed cave depot surrounding rock under the action of supercritical CO2 phase change fracturing on the basis of a momentum conservation law and an elastic mechanics theory; carrying out a similar model test of the underground water-sealed cave depot to obtain vibration speed response data; and a water-seal cave depot surrounding rock vibration speed calculation formula under the supercritical CO2 phase change fracturing action is compiled into Matlab for parameter analysis, optimal parameters are summarized and analyzed, and calculation of the vibration speed is completed. The method aims at solving the problems that in the prior art, under the supercritical CO2 phase change fracturing effect, the vibration speed of the surrounding rock of the underground water-sealed cave depot is difficult to accurately calculate, and parameters influencing the vibration speed cannot be effectively optimized.

An embodiment of the present invention provides a method for unconditionally maintaining the strong stability of a hyperbolic conservation law in computational fluid dynamics, comprising: spatially discretizing the hyperbolic conservation law equation in computational fluid dynamics using a preset spatial discretization method to obtain a semi-discrete system of the hyperbolic conservation law; introducing a stabilization term into the semi-discrete system of the hyperbolic conservation law to form a stabilized semi-discrete system of the hyperbolic conservation law; temporally discretizing the stabilized semi-discrete system of the hyperbolic conservation law using an explicit integrating factor Runge-Kutta method to obtain a fully discrete system of the hyperbolic conservation law; and reasonably approximating the exponential function in the explicit integrating factor Runge-Kutta method of the fully discrete system of the hyperbolic conservation law to obtain a fully discrete format of the hyperbolic conservation law that unconditionally maintains strong stability. This method ensures that the hyperbolic conservation law can maintain strong stability at any time step.

This invention discloses a rapid simulation and evaluation method and system for large-scale shock wave damage. The method includes: S1, calculation of shock wave power parameters; S2, rapid acquisition of explosion shock wave power field parameters; S3, establishment of target data structure, and establishment of a target data description structure suitable for explosion effect analysis based on mirrored virtual sources; S4, damage analysis of the target by the explosion shock wave; S5, visualization of the target damage scene; S6, using a dynamic physics engine, integrating the engine with the explosion shock damage algorithm, combining graphical interface technology and graphical visualization technology to practically integrate the explosion shock damage algorithm, forming a rapid simulation analysis of explosion wave damage. This invention approximates the linearization of the propagation law of explosion waves, introduces virtual sources to handle the propagation, reflection, and diffraction problems of explosion waves in complex scenes, and forms a source effect superposition algorithm based on the basic conservation laws of fluid mechanics to calculate the nonlinear superposition effect of explosion waves from multiple virtual sources.

The application provides a neural network prediction method and device for two-dimensional quantized vortex dynamics, and the method comprises the following steps: S1. For a flow field to be predicted, a two-dimensional vortex dynamics equation of the flow field is determined, and the two-dimensional vortex dynamics equation is converted into a Schrodinger equation of a quantum system; S2. Wave function evolution data of the quantum system is obtained according to the Schrodinger equation, and a wave function data set is generated; S3. A full connection layer neural network model is constructed, and phase integration is performed on the full connection layer neural network model; S4. A normalization factor is introduced into the full connection neural network model; S5. The full connection layer neural network model is trained by using the wave function data set; S6. The trained full connection layer neural network model is used to predict a wave function according to a given initial value; and S7. The predicted wave function is converted into flow field evolution. According to the method provided by the application, the conservation law of the system can be ensured in the prediction of the Schrodinger system, so that the accuracy of the flow field prediction can be ensured in a long period.

This paper discloses a method for predicting Arctic sea ice based on a multi-scale graph neural network and conservation laws. The method selects surface, ocean, and atmospheric variable data, grids the Arctic region at different spatial resolutions, and constructs a multi-scale graph. The method updates the features of each node in the graph and integrates the spatial information of the different-scale graphs to obtain a sea ice density prediction model. The method adds humidity conservation terms and potential vorticity conservation terms to the loss function of the sea ice density prediction model, constructing a composite loss function to ensure that the model adheres to the physical conservation laws of humidity and potential vorticity during the prediction process. The sea ice density prediction model is trained using historical data and the composite loss function, and the future sea ice state is predicted using the daily updated data and the trained model. This method significantly improves the accuracy and efficiency of sea ice density prediction, providing a scientific basis for climate change research, resource development, and shipping route planning.

The application provides a method and system for predicting particle settling, collision and flow characteristics in a pump, comprising: constructing a multi-phase flow training sample library of real-time collected environment and virtual-real fusion; after pre-processing the sample library data, using a double-flow deep neural network architecture to extract multiple source feature vectors in parallel; cross-fusing the multiple source feature vectors to obtain a multi-modal fusion feature vector, and establishing an Euler-Lagrange two-way coupled partial differential control equation; inputting the multi-modal fusion feature vector into a physical information neural network based on a gradient-aware adaptive physical loss function, and outputting particle instantaneous settling velocity, particle collision contact force, local entropy production rate and global reconstructed three-dimensional flow field; based on the output calculation index, hierarchical regulation is executed accordingly. The application breaks the feature mismatch problem, replaces the complex grid iteration with millisecond-level forward reasoning conforming to the conservation law of fluid mechanics, and realizes high-precision intelligent online prediction of complex flow field characteristics.