661 results about "Discretization" patented technology

The invention relates to a quick river flood forecasting method based on a physical information neural network, and belongs to the field of river flood forecasting. The method comprises the following steps: on the basis of a traditional physical information neural network (PINN), introducing a boundary condition parameter as an input variable, and enabling the PINN to learn a flood wave propagation rule under different boundary conditions. Furthermore, on this basis, a physical information neural network flood fast forecasting framework (RFF-PINN) integrated with a hydrodynamic method is provided, a numerical solution based on grid discretization is reconstructed into a continuous function in a time-space domain through a piecewise polynomial interpolation method, residual error loss between network output and a hydrodynamic model simulation value is constructed, and therefore, a flood fast forecasting result is obtained. The network parameters are optimized in cooperation with the PDE loss, and the problem that the network parameter optimization effect is reduced due to the fact that the PDE loss of the complex flow state area is difficult to converge is solved. The method has the beneficial effect that the water depth change process of each section of the river channel under any boundary condition can be accurately and quickly predicted.

The invention relates to the technical field of detection and identification, and discloses a mobile ship dynamic tracking and locking method and system based on target detection and identification, and the method comprises the steps: constructing a manifold embedded network, and mapping the multi-modal features of a ship to a Riemannian manifold space; constructing a continuous evolution modeler of the Shenchang differential equation model, and predicting a continuous evolution trajectory of the features under time, view angle and scale changes; a multi-scale bridging network is constructed, and bidirectional conversion between long and short distance feature representations is realized; constructing a differential geometric attitude encoder, and representing a ship attitude in an SO (3) Lie group space; a feature memory and reconstruction mechanism is realized, and the problem of short-time disappearance target recovery is solved; the components are integrated to construct a unified tracking and locking system. According to the method, the problems of discretization processing limitation, feature representation fragmentation, cross-condition consistency deficiency and the like in the prior art are solved, remarkable technical effects are achieved in the aspects of cross-condition recognition capability, continuous feature evolution capability, short-time disappearance target recovery capability and the like, and the method is suitable for the scenes of Yangtze River shipping monitoring, maritime affair safety supervision and the like.

The invention relates to a thermal simulation numerical simulation method of a multi-physics field architecture with CFD as a core, and belongs to the technical field of cross of computational fluid mechanics and heat transfer science. The method comprises the following steps: constructing a geometry and a conformal / non-conformal grid, presetting a model library, establishing association between a region and a physical field through physical group division, confirming a boundary and initial parameters, and screening a core simulation model; dividing an independent control volume based on a core numerical model, converting an area component and a volume component through integral operation, discretizing a boundary flux to construct a parameter matrix, then solving an equation set, and synchronously optimizing the grid quality; calculating a boundary flux and matching a discrete format, a turbulence model and a corresponding multi-physical field coupling model; original data are screened and processed, and cloud picture results such as temperature and flow field are generated through visualization. Accurate simulation of a multi-physics coupling scene is achieved, the precision and stability of thermal simulation under complex working conditions are improved, and efficient and reliable technical support is provided for engineering design optimization and performance evaluation.

The invention discloses a simulation method, device and equipment for predicting flow and deformation of a porous medium and the medium. The method comprises the steps that a computational domain is selected, a corresponding computational grid is generated, and porosity initialization is conducted on the computational domain; determining a time step length and a discrete format, and setting a non-uniform factor to solve a control equation set which comprises a continuity equation, a fluid dynamics equation, a skeleton mechanics equation and a porosity updating equation; and analyzing and displaying a solving result. According to the porous medium flow and deformation simulation method independent of hypothesis and empirical parameters, the method is based on the basic principle of the microcosmic fluid-solid coupling effect, a porous medium fluid-solid coupling body kinetic equation is adopted, the situation that a traditional modeling method depends on volume averaging hypothesis and depends on a multi-phase interaction model is avoided, and the modeling efficiency is improved. The method has cross-scale applicability, and provides a new theoretical basis and prediction means for the development of technologies related to porous medium multiphase dynamics and multi-field coupling problems.

The invention provides an integrated feature selection method and product based on quantum computing and Bayesian optimization, and relates to the technical field of data processing. According to the embodiment of the invention, an original mathematical integration model is converted into a QUBO model which can be solved by quantum calculation, and the QUBO model is decomposed into a joint optimization discretization step length optimization sub-problem and a parameterization QUBO sub-problem. In a mixed quantum classical optimization algorithm framework, a self-adaptive Q learning model is designed on the upper layer, and proper sub-problems can be dynamically selected in the search process. In the lower layer, a dropout Bayesian optimization algorithm is provided for effectively optimizing the high-dimensional discretization step length in each iteration. A CIM-based quantum computing method is adopted, and a parameterized QUBO sub-problem under the given discretization step length is efficiently solved. According to the method provided by the embodiment of the invention, the feature selection problem can be successfully and efficiently solved, and the selected features of the credit classification problem and the credit classification model for classification based on the selected features can be obtained.

The invention discloses an adaptive edge detection iterative denoising point cloud reconstruction method and system. A regularization parameter and a kernel bandwidth are automatically set through point cloud normalization and local covariance noise estimation; an implicit curved surface is fitted based on a parameterized Gaussian field, and initial curved surface construction without a normal direction is realized. Edge detection is carried out by utilizing Voronoi covariance measurement, and the edge detection is continuously mapped into a projection weight, so that feature details of the point cloud are reserved in an acute angle region, and smooth denoising is realized in a flat region. The regularization intensity is dynamically adjusted through residual trend analysis to enhance detail representation, and a triangular mesh is extracted and output in combination with octree discretization and contour surface extraction. The method has unsupervised adaptivity, has outstanding performance in the aspects of edge fidelity, noise robustness and complex structure adaptability, can effectively process point cloud data with high noise, missing normal direction or irregular structure, and is suitable for the fields of engineering survey, reverse modeling, cultural relic digitization and the like.

The invention discloses a multi-dimensional training method and device for a support vector machine, electronic equipment and a computer readable storage medium, and the method comprises the steps: carrying out the discretization of a training sample data set, and obtaining a discretized data set, the discretized data set comprises a plurality of different attributes, and each attribute corresponds to a plurality of feature vectors; calculating a classification contribution parameter of each attribute feature vector to obtain a plurality of classification contribution parameters; performing data mapping on the plurality of classification contribution degrees by using a kernel function to obtain a target function; and optimizing and training the objective function by using a gradient descent algorithm to obtain a support vector machine model. According to the method, different dimension data are mapped through the kernel function, the data gain weight can be determined, the linear separable effect of the mapped dimension data can be improved, and then the classification and analysis capability of the SVM can be improved.

The invention provides a semi-active suspension vibration control method and system based on preview fusion and game decision, and the method comprises the following steps: obtaining pavement preview information through a forward sensor, carrying out the signal processing, and generating a future pavement preview sequence in real time; establishing a two-degree-of-freedom 1 / 4 whole vehicle vertical dynamical model; under a model prediction control framework, participants of a non-cooperative Nash game model are defined, a corresponding cost function is designed, a damping coefficient is discretized by adopting mixed integer programming, Nash equilibrium is solved by adopting a branch and bound method, and an optimal damping coefficient is obtained; mapping the optimal damping coefficient into optimal damping force meeting constraint conditions; hysteresis force compensation is conducted through a hysteresis model, and target control current is obtained through damping force obtained after hysteresis compensation according to the mapping relation between the damping force of the magnetorheological damper and the control current; and outputting the target damping force according to the instruction of the target control current. According to the invention, autonomous balance adjustment of multiple control targets is realized.

The invention discloses a synchronous generator excitation method based on mathematical modeling and gain scheduling, and relates to the technical field of synchronous generator control, and the method comprises the steps: building a synchronous generator global dynamic mathematical model comprising the electromagnetic dynamic characteristics of a stator winding, a rotor winding and a damping winding through a modeling module; a discretization simulation platform is constructed in a Matlab / Simulink environment by utilizing a simulation module, model parameters are corrected, dynamic adjustment is realized through a gain scheduling excitation controller, an automatic voltage regulator (AVR) adopts PID (Proportion Integration Differentiation) control and comprises a voltage deviation integral separation strategy, and a power system stabilizer (PSS) dynamically adjusts parameters of a lead-lag compensator based on active power. According to the method, the deviation between a model and an actual system can be reduced by more than 30%, the electromechanical oscillation damping ratio within the rated power range of 10%-150% is increased by 25%-40%, the harmonic content of exciting current is reduced to be lower than 5%, the stability of an electric power system and the electric energy quality are remarkably improved, the engineering practicability is high, and the method is easy to transform and upgrade on an existing synchronous generator.

The invention provides an intelligent control method and system for discrete event simulation of an aluminum template production line, and relates to the technical field of data processing.The method comprises the steps that a material transfer path instruction and a cutting process key point coordinate sequence are analyzed from a dynamic anti-interference scheduling strategy, the key point coordinate sequence is input into a curve driving algorithm, and a curve driving algorithm is obtained; a parameterized continuous cutting path curve is generated according to the aluminum template contour precision requirement and the cutting head dynamics constraint; the method comprises the following steps: performing equal-bow-height error discretization processing on a continuous cutting path curve to generate a discrete pose point set with a timestamp, superposing AGV dynamic constraints to generate a speed look-ahead control instruction set, splitting the instruction set into independent control units according to process sections, packaging the independent control units into a time trigger message through an industrial bus protocol, and sending the time trigger message to a server; and the simulation clock synchronization signal is injected into a physical execution unit. According to the invention, intelligent and accurate control of the aluminum template production process is realized, and the production efficiency is effectively improved.

The invention relates to the field of rotating machinery fault diagnosis, and discloses a bearing fault diagnosis method based on a physical perception KAM network, and the method comprises the steps: carrying out the discretization of a bearing vibration signal through a Gabor filter group based on physical prior initialization, and generating modal feature lexical elements with physical frequency band meanings; and inputting the lexical elements into a PC-KAM backbone network, calculating a hidden state vector by using a state space model branch, and dynamically adjusting the position of a primary function node of a Kolmogov-Arnod network branch to realize collaborative dynamic feature extraction. In the training stage, an orthogonal subspace constraint and physical perception low-rank adaptation fine tuning mechanism is introduced. And finally, searching a historical fault case, performing multi-modal fusion on the historical fault case and the deep feature sequence, mapping a fusion representation into a soft prompt, and inputting the soft prompt into a large language model to generate a diagnosis report. According to the method, the problems of poor physical interpretability of characteristics and few-sample diagnosis under variable working conditions are effectively solved, and the generalization and decision-making ability of a diagnosis system are improved.

The invention discloses a complex medium controllable source electromagnetic multi-parameter three-dimensional inversion method and system, and the method comprises the steps: carrying out the three-dimensional space discretization of an exploration region, and carrying out the multi-time iteration inversion of a discrete exploration region through initializing a multi-parameter geologic model based on a multi-parameter inversion target function, and obtaining a multi-parameter three-dimensional inversion result; wherein the multi-parameter inversion objective function comprises a data fitting constraint, a model constraint and a cross gradient constraint, and the multi-parameter model comprises resistivity, magnetic conductivity, a charging rate, a frequency correlation coefficient and a time constant; and a hexahedral mesh discrete calculation region is adopted in inversion, and a finite element linear equation set is obtained through a vector finite element method, so that multi-parameter three-dimensional forward modeling calculation is realized. According to the method, the induced polarization effect and the magnetic conductivity influence are considered at the same time, multi-parameter three-dimensional inversion is achieved, the application precision and reliability of the controllable source electromagnetic method under the complex geological condition are remarkably improved, and rich geophysical basis is provided.

The invention relates to the technical field of knowledge graph construction, and discloses a knowledge graph-based agent behavior description method, which comprises the following steps of: firstly, analyzing an unstructured agent operation log data stream through a processor, and mapping the unstructured agent operation log data stream into a discretized module access sequence by utilizing a topological structure of a code knowledge graph so as to remove text redundancy; performing multi-dimensional time sequence analysis on the sequence, and calculating resident distribution data representing operation continuity, frequency domain fluctuation data representing switching rhythm and multi-scale coverage extension data representing a traversal range; and then aggregating the data into a fixed-length multi-dimensional index vector, constructing an index node associated with a session entity in a graph database, and writing the vector as a binary structured attribute into a storage field. According to the method, massive unstructured logs are converted into compact graph structured indexes, and quick positioning and direct retrieval of a complex agent operation mode are supported while the storage space is remarkably saved.

The invention discloses a method for predicting fluid excitation characteristics of a propeller, and relates to the field of propeller simulation. The method comprises the following steps: modeling a propeller flow field according to attribute information corresponding to the propeller flow field to obtain a propeller flow field grid model; determining a propeller flow field control equation corresponding to the propeller flow field grid model, and discretizing the propeller flow field control equation by adopting a finite volume method to obtain a discretized propeller flow field control equation; performing propeller hydrodynamic numerical simulation on the discrete propeller flow field control equation based on a delay separation vortex simulation method to obtain speed spatial-temporal distribution and pressure spatial-temporal distribution, and performing decoupling processing on the speed spatial-temporal distribution and the pressure spatial-temporal distribution; and according to the whirling vibration data, the vortex motion data, the decoupled pressure field and the decoupled speed field, the exciting force of the propeller flow field is obtained. The method can accurately reflect the influence of different vortex motions on the excitation characteristics of the propeller and the fluctuation of the velocity field of the propeller.

The invention provides a numerical simulation method and system constructed based on a neural network and an iterative approximation method. The method comprises the steps that a to-be-solved partial differential equation is determined; constructing a precise solution data set; establishing a deep learning objective function network; constructing a loss function; training the network through the random training set; and outputting the prediction solution. According to the method, the time variable is used as network input, continuous time prediction under single-round training is realized, and the problem that a traditional discrete model needs multiple times of iterative training is solved; tasks such as target detection and image classification can be executed, screen and back images of the mobile terminal / tablet personal computer are input in product defect detection, and defect existence, position or type are output; and inputting a sea surface image in sea surface ship identification, and outputting existence, position or type of the ship. According to the method, the number of training iterations in a Burgers equation experiment is reduced by 19.3%, and full-time-domain repeated training of a traditional discrete model is eliminated; internal point sampling is avoided through time semi-discretization, and the calculation efficiency is remarkably improved.

The invention discloses a flood peak enhanced physical base flow and residual error correction collaborative runoff prediction method, which belongs to the field of hydrological prediction, and comprises the following steps of: dividing a training set and a verification set according to a proportion, performing oversampling processing on flood peak samples, and constructing a time sequence window; a Xinanjiang model is discretized and expressed by adopting an ordinary differential equation, rainfall and potential evaporation data are input, and intermediate variables are obtained. A physical base flow and residual error correction dual-channel module is constructed, and physical base flow and residual error correction is calculated through two full-connection networks. And calculating a final runoff predicted value by adopting a residual connection structure, taking basic NSE loss as a core, superposing a flood peak sample error weighted item, strengthening flood peak fitting precision, and updating physical parameters and neural network weight through a back propagation algorithm. And verifying the model, and respectively calculating prediction indexes of the training set and the verification set. According to the method, fusion of a traditional hydrological model and a deep learning method is realized, the physical interpretation of the model is enhanced, and the basin runoff prediction precision is improved.

The invention discloses a carbon dioxide geological sequestration quantity prediction method, system and device and a storage medium, and the method comprises the steps: obtaining a mathematical expression of carbon dioxide sequestration capacity through combining a coupling model of a flow field, a mechanical field and a thermal field with random parameters, and introducing a leakage risk coefficient; taking the weighted sum of the expected value of the storage capacity and the variance as an optimization objective function, and according to the mathematical expression of the carbon dioxide storage capacity, carrying out optimal solution solving on the carbon dioxide storage capacity to obtain an optimal storage strategy; performing spatial discretization on the control equation by adopting a finite element method, discretizing time by adopting a semi-implicit difference method, discretizing random variables by combining a random Galkin method, and solving a coupling model of a flow field, a mechanical field and a thermal field to obtain a prediction result of the carbon dioxide sequestration amount; and according to a prediction result of the carbon dioxide sequestration amount, outputting an expected value and a variance of the sequestration capacity, and generating a sequestration amount distribution diagram, a risk assessment diagram and a sensitivity analysis diagram.

The invention relates to the technical field of metal composite materials, in particular to a metal composite material production process dynamic simulation method based on digital twinning. Comprising the following steps: constructing a digital twin model comprising a physical entity layer, a virtual model layer and a connection layer, and establishing a Fourier neural operator model embedded with a Navier Stokes equation and a heat conduction equation constraint; real-time molten pool monitoring data are mapped into model input through a connecting layer, and advanced time step deduction is executed by using a deep learning model to predict the state of a molten pool; based on the differentiable characteristic of a physical information neural operator, calculating the gradient of a prediction state and an ideal target by using an automatic differential algorithm, generating a process parameter field through reverse iteration updating, and discretizing the process parameter field into a galvanometer control instruction; and when the risk of aggregation of the reinforced phase particles is predicted to exist, automatically generating a swing scanning strategy to realize active suppression. The method gives consideration to simulation precision and real-time performance.

The invention discloses a coal-series gas yield increasing and CO2 geological sequestration integrated simulation method and system, and relates to the technical field of gas exploitation of each layer, and the method comprises the following specific steps: building a geological model representing a multi-lithology superposed coal-series reservoir; the coal series reservoir is a heterogeneous reservoir and comprises shale, a coal seam and sandstone, the shale and the coal seam are double-hole double-permeability media, and the sandstone is a single-hole single-permeability medium; based on the geologic model, establishing a multi-physical field coupling model of a coupling temperature field, a seepage field and a stress field; performing discretization solution on the multi-physics field coupling model to obtain a numerical simulation result; constructing a numerical model, and performing inversion correction on model parameters of the numerical model based on experimental data and the numerical simulation result; and utilizing the corrected numerical model to simulate different injection-production working conditions so as to carry out double-target evaluation and scheme optimization of coal-series gas yield increase and CO2 geological sequestration. The method provides a reliable decision basis for engineering scheme design.

The invention belongs to the technical field of underwater propeller design, and discloses an annular propeller modeling and hydrodynamic performance evaluation method and system. The method comprises the following steps: defining a distribution rule of blade parameters along a wheelbase based on a segmented cubic Hermite interpolation method, and realizing parametric modeling; automatically generating a propeller three-dimensional geometric model and a flow field computational domain; carrying out discretization processing on the rotating domain and the outer watershed by adopting a hybrid grid technology; hydrodynamic performance data are obtained through numerical simulation calculation, and thrust, torque and efficiency coefficients are obtained through processing; and completing multi-working-condition performance calculation and drawing an open water curve by automatically and cyclically updating a speed coefficient, and finally storing design variables and performance response data into a database. According to the method, full-process automation from modeling, grid division, simulation calculation to post-processing is achieved, the efficiency of annular propeller design and hydrodynamic performance analysis is remarkably improved, and the labor cost and the time cost are greatly reduced.

The invention discloses a rock blasting load transfer and throwing simulation method which comprises the following steps: S1, acquiring field parameters of a target blasting project, the field parameters comprising rock mass mechanical parameters, explosive parameters, blast hole parameters and charging structure parameters; and S2, based on the field parameters, constructing a three-dimensional explosive-rock calculation model by adopting pretreatment software, dividing grids, importing the grids into finite element software, setting a material model, a state equation and boundary conditions, and simulating an explosive detonation process. The limitation of a single numerical value method is solved; quantitative analysis and accurate prediction of rock blasting lumpiness are achieved, inclined joints are adopted in the discretization process, actual rock mass joint distribution is better fitted, and the reliability of a simulation result is improved; bulk rocks and excessive crushing phenomena are effectively reduced, the explosion energy utilization rate is improved, the engineering cost is reduced, and a scientific basis is provided for explosion design of projects such as mining and water conservancy and hydropower.

The invention provides a hydrological flow long sequence prediction method and system of an improved state space model. The method comprises the steps of collecting multi-source data of a drainage basin to be predicted; constructing a time sequence sample pair by the processed multi-source data through a sliding window method, wherein the time sequence sample pair comprises an input sequence and a target sequence; a HydroMama model is constructed according to the time sequence sample pair, a HydroMama prediction model is trained, and an optimal hyper-parameter combination is searched for; based on the trained HydroMama model, traffic prediction and result restoration are realized by adopting an autoregression mechanism; the calculation efficiency is high, and the continuous state equation discretization of the state space model is utilized, so that the long-distance meteorological-hydrological hysteresis effect which is difficult to capture by a traditional cycle model can be captured; according to the method, skewed distribution of hydrological data is processed through logarithmic transformation, and an anti-overfitting objective function is combined, so that the prediction performance of the model on unseen data is remarkably improved, and the practical application value is high.

The invention discloses a surface material point embedding method for a multi-body contact problem, and belongs to the technical field of computational mechanics and numerical simulation. The method comprises the following steps: 1, performing discretization and parameter setting on a physical model of a multi-body contact system; 2, calculating a contact force; 3, transmitting the physical quantity of the material points of the contact body to background grid nodes to obtain node mass, momentum and force; 4, updating the grids; 5, updating the speed, the displacement, the stress and the density of the material point; 6, resetting the physical quantity of the background grid; seventhly, repeating the first step, the fourth step, the fifth step and the sixth step, sequentially updating the physical quantities of the material points of other contact bodies and the positions and the speeds of the slave nodes, and distributing background grid node memories of the areas where the corresponding contact bodies are located; and 8, after all contact bodies are updated, updating the current time to obtain a dynamic simulation result. According to the method, the boundary of the object can be accurately represented, complicated problems such as impact and multi-body contact can be effectively simulated, and meanwhile, the method has good stability and relatively low memory consumption.

The invention provides an aircraft component fault intelligent diagnosis and maintenance control system, and relates to the technical field of aircraft component diagnosis. A stall modeling module of the system establishes an aeroelastic coupling equation for quantifying stall behaviors on the basis of collected pressure disturbance data of an air inlet flow field of an air compressor and blade row vibration response information. The elastic response deduction module discretizes the equation through a state conversion operator and then embeds the equation into a graph neural network node update function. The graph embedding coupling recognition module recognizes blade tip gap mutation behaviors induced by local stall and time sequence precursor characteristics of the blade tip gap mutation behaviors by using a node state evolution trajectory output by the graph neural network, and constructs a surge precursor evolution path. And the early warning output module carries out instability trend judgment on the surge precursor evolution path, and outputs a maintenance suggestion and an early warning identifier aiming at a blade fracture fault, so that the structure failure risk caused by the surge of the gas compressor is effectively predicted, and the accuracy and timeliness of fault diagnosis of aircraft parts are remarkably improved.

The invention discloses a model discretization method and device, electronic equipment and a computer readable storage medium. Comprising the steps that a to-be-discretized CAD model is obtained, the to-be-discretized CAD model is analyzed, geometric features of the to-be-discretized CAD model are obtained, and the geometric features at least comprise faces and edges; for any adjacent edge A and edge B, discretizing the edge A and the edge B according to the relationship between the edge A and the edge B, and discretizing the edge A and the edge B respectively to obtain a discrete point column of the edge A and a discrete point column of the edge B; and for any adjacent surface C and surface D, discretizing the surface C and the surface D according to the relationship between the surface C and the surface D, and discretizing the surface C and the surface D respectively to obtain a discrete grid of the surface C and a discrete grid of the surface D. According to the method, when the model is converted into the net-shaped structure, it is guaranteed that selfing of the grids does not occur.

The invention provides a stability analysis method and system for a tracking and networking inverter hybrid system and a medium, and the method comprises the steps: selecting the active power of a tracking inverter and the reactive power of a networking inverter as core variables through working condition variable dimensionality reduction; determining a working condition range in combination with capacity constraint, performing discretization processing, and calculating dq components of voltage and current of an inverter port based on a discrete working condition; discrete impedance models of the two types of inverters are respectively constructed; the influence of a phase-locked loop controller, a current loop controller and a power loop controller is considered in a following network type, and the effect of a virtual synchronous controller, a current loop controller and a voltage loop controller is included in a construction network type; and calculating a closed-loop transfer admittance matrix of the power grid through the discretized impedance of the inverter to realize stability analysis of the hybrid system. According to the method, the modeling complexity is effectively reduced, the stable working areas of the system under different working conditions can be quantified, and an accurate reference basis is provided for stable operation of the hybrid inverter system.

The invention provides a dam safety monitoring data analysis method and device based on space-time gradient. Constructing a space-time fusion coordinate system, discretizing a three-dimensional space coordinate of the dam into finite element grid nodes, and mapping each node to a normalized parameter space through cubic spline interpolation to form a space parameter; the method comprises the following steps: dynamically expanding monitoring data on a time axis, generating time-frequency joint representation by adopting a sliding time window and wavelet transform, and forming a high-dimensional space-time curved surface with a time-space double-continuous axis; a first-order space-time gradient field and a second-order curvature tensor are calculated based on the space-time curved surface, a finite element balance equation is combined as a regularization item, and mathematical modeling and dynamic recognition of an abnormal propagation path are achieved; and according to the characteristics of the gradient field and the curvature tensor, constructing a space-time attention mechanism, automatically focusing a high-risk area and generating a graded early warning signal. According to the invention, the abnormity identification precision and early warning timeliness can be effectively improved, the false alarm rate is reduced, and the system expansibility is enhanced.

The invention relates to a primary frequency modulation control method for a pumped storage coupled compressed air energy storage system, and the method comprises the steps: building a frequency modulation control model based on the core component information and operation constraints of the pumped storage coupled compressed air energy storage frequency modulation system, building a state space model, and building a discretization state equation, establishing a comprehensive optimization model by taking the minimum power grid frequency deviation and the system operation cost as targets; and establishing a multi-parameter planning model based on the reservoir level, the gas storage pressure, the frequency deviation and the comprehensive optimization model, and converting a hierarchical model prediction control problem into multi-parameter quadratic programming to form a self-healing frequency modulation strategy to perform power output tracking control. Therefore, the problem that the frequency modulation performance is limited due to the fact that a water pressure system and an air pressure system in the hybrid power station are complex in coupling and large in response difference, and a related linear control strategy or a single control framework is difficult to achieve rapid response and accurate coordination of the hybrid system in the primary frequency modulation process is solved.

The invention provides an optimal primary frequency control method based on reinforcement learning. The method comprises the following steps: establishing a dynamic model of a power system; designing a frequency control strategy based on a deep reinforcement learning algorithm; the system frequency deviation is used as a reward signal, and a control strategy is trained through interaction with a power system; lyapunov function stability constraints are introduced into the reinforcement learning algorithm; a neural network is adopted to carry out parameterization design on the controller; discretizing the frequency dynamic equation, and proposing a multi-node collaborative optimization target; and deploying the trained control strategy to an actual power system, and adjusting control parameters in real time through an online optimization technology to adapt to the dynamic change of the system operation state. By means of the method, the frequency control accuracy and stability of the power system are improved, and the global optimal performance of frequency control of the whole power grid is achieved in actual deployment.

The invention discloses a text-to-action generation method and system based on body part fine-grained representation, and belongs to the field of human body action generation. Body part-level fine-grained discretization modeling is performed on human body actions to obtain an action encoder and an action decoder which can stably reconstruct a continuous action sequence; the method comprises the following steps: taking a part fine-grained discrete action space as a producible space, introducing video potential representation as space-time dynamic priori, establishing an alignment modeling mechanism of the video potential representation and the action discrete space, and training to obtain a conditional generation network, so as to realize stable prediction and iterative completion of an action discrete index sequence in a reasoning stage; and a continuous human body action sequence is reconstructed through an action decoder, and finally a human body action sequence which is consistent with text semantics, richer in details and more coherent in time sequence is generated. According to the method, the technical threshold and data dependence of action content production can be reduced, and the controllability and generalization ability of action generation are improved.