25results about How to "Improve modeling accuracy" patented technology

This invention discloses a modeling method, electronic device, and storage medium for hydrogen energy systems. The method includes: establishing a hydrogen production mass flow model for an electrolyzer, and using a multi-condition linearization method to handle the nonlinear operating efficiency of the electrolyzer; establishing a storage mass flow model for a hydrogen storage tank, dynamically characterizing its effective capacity based on the tank's physical volume, operating pressure, and ambient temperature parameters; constructing a hydrogen mass flow bus, linking the electrolyzer's hydrogen production variable and the storage tank's hydrogen filling variable to this bus, and establishing hydrogen mass flow balance constraints among the models to characterize that at any given time, the total amount of hydrogen flowing into the bus equals the total amount flowing out. This invention solves the problem of insufficient accuracy in traditional models, achieving a balance between high accuracy of the equipment model and high solution efficiency for the optimization problem, providing a reliable modeling foundation for the optimal scheduling of hydrogen-containing integrated energy systems.

This invention relates to a method for modeling the capillary evolution of humidity-related adhesion forces under micrometer-scale probes. This method establishes a physical relationship between ambient relative humidity and Kelvin radius, and calculates the Kelvin radius in conjunction with the initial volume of the liquid bridge, thereby constructing a parameterized model of capillary forces that characterizes the change in liquid bridge volume with ambient humidity. The influence of the liquid film on the solid surface is considered during model construction, enabling the model to more realistically reflect the formation and evolution of liquid bridges in actual micro / nano interfaces. As the two surfaces gradually move away, the relationship between the curvature of the liquid bridge interface and the Kelvin radius is established under the condition that the liquid bridge volume remains constant, thus achieving a continuous description of the evolution of capillary and adhesion forces during the separation process between the micrometer-scale probe and the sample interface. Furthermore, experimental measurements and model verification of micrometer-scale interface forces are conducted using a self-developed fiber Bragg grating probe.

This application discloses a method, apparatus, device, and medium for processing data related to neuropsychiatric diseases, relating to the field of data processing technology. The method includes: acquiring gut microbial species characteristic data of a target sample, including relative abundance characteristics or species presence / absence status characteristics; acquiring the target disease and its corresponding decision tree binary classification model, which is trained using the CatBoost algorithm, and the model's output categories include a target category and a null category; inputting the gut microbial species characteristic data into the model, and outputting a probability value indicating that the target sample belongs to the target category; and generating a data processing result corresponding to the target sample based on the probability value. Using the neuropsychiatric disease data processing method of this application can improve the accuracy and cross-disease generalization ability of neuropsychiatric disease data processing models.

This invention discloses a light projector and a sensing system used in conjunction with the light projector. The light projector includes a light source and a diffraction optical element. The light source has a first light-emitting area and a second light-emitting area, and the diffraction optical element is disposed on the light source. The diffraction optical element has a first diffraction structure and a second diffraction structure, which overlap the first light-emitting area and the second light-emitting area, respectively. The first diffraction structure is different from the second diffraction structure.

The application provides a positive and negative sequence four-port impedance modeling method and device of a mixed frequency power transmission system, and relates to the technical field of impedance modeling. The method comprises the following steps: obtaining a positive and negative sequence voltage coupling matrix according to a power frequency side PCC steady-state q-axis voltage, a frequency division side PCC steady-state q-axis voltage, a power frequency side PLL transfer function and a frequency division side PLL transfer function; obtaining the positive and negative sequence four-port impedance according to a pre-obtained power frequency side current inner loop controller matrix, a frequency division side current inner loop controller matrix, a power frequency side voltage outer loop controller matrix, a frequency division side voltage outer loop controller matrix, a power frequency side decoupling matrix, a frequency division side decoupling matrix, a voltage calculation matrix, a current calculation matrix and the positive and negative sequence voltage coupling matrix. The method and device provided in the application embodiment can improve the modeling accuracy of the positive and negative sequence four-port impedance of the mixed frequency power transmission system.

The application discloses a method for intelligently modeling rock mechanics properties of a tight sandstone reservoir, comprising the following steps: obtaining three-dimensional seismic interpretation data, well logging data and rock mechanics experiment data; establishing a three-dimensional geological structure model of the tight sandstone reservoir which fuses a fault structure; calculating a quantitative fitting relationship between static rock mechanics parameters and a sand ratio; extracting a lithology factor data body from the three-dimensional seismic interpretation data, discriminating the lithology factor data body, generating a three-dimensional spatial distribution of sandstone and mudstone, and thus obtaining a three-dimensional sand ratio distribution model; and based on the three-dimensional sand ratio distribution model and the quantitative fitting relationship, calculating initial values of rock mechanics parameters of each grid unit, and assigning values to the three-dimensional geological structure model and correcting the fault, so as to obtain a final three-dimensional rock mechanics property model. The application realizes multi-source data fusion and quantitative modeling of fault disturbance, and solves the problems of insufficient description of reservoir heterogeneity, lack of physical basis of the model and low reliability of modeling of a fracture zone in the prior art.

The invention discloses a Dynamo-based municipal drainage pipe network automatic modeling method, and relates to the technical field of building information model and municipal pipe network modeling. An external data file containing inspection well and pipeline information is created, automatic data analysis and processing are achieved through a Dynamo environment, and a parameterized data set is generated. And automatically generating an inspection well family instance based on the inspection well space coordinate attribute subset, and generating a pipeline family instance in combination with the pipeline connection relation attribute subset to realize endpoint space association. According to the method, classified generation of the sewage pipe network and the rainwater pipe network is supported, batch processing is achieved through List.FilterByBoolMask nodes, the modeling time is obviously shortened, the modeling efficiency is remarkably improved, and dynamic synchronization of the model is supported by updating external data files.

This invention discloses a method, apparatus, terminal equipment, and storage medium for flexible load aggregation modeling based on price response, belonging to the field of load dispatching. The method includes: constructing a physical aggregation model and corresponding aggregation model constraints based on load data, energy storage operation status data, distributed generation capacity data, and grid topology data; solving the model under constraints based on the acquired electricity purchase price sequence and predicted electricity purchase price, with the objective of minimizing electricity purchase cost; obtaining the optimal model parameters by inverse optimization of the physical aggregation model based on the solved initial aggregation response power trajectory, initial power allocation results, and measured response power data from the load data, with the objective of minimizing the mean square error between the observed response and the simulated response; and performing grid dispatching based on the target aggregation response trajectory, target power allocation results, and grid topology obtained from the updated physical aggregation model. By implementing this invention, the problem of prior art neglecting the dynamics and real-time rolling of flexible loads is solved.

The application discloses a kind of based on CatBoost regression model's chlorine ion diffusion coefficient prediction method.The method first collects measured sample data and carries out data cleaning, ensure the quality of data.Subsequently, by LASSO regression feature selection method, the feature that has significant contribution to target value is screened out, and the optimal feature coefficient is selected using cross validation.Then, define regression model evaluation function.Then, using CatBoost regression model is trained, and the hyperparameter is optimized by grid search, and the best model parameter is obtained.Prediction is carried out on training set and test set, and model evaluation index is calculated and output, finally, fitting effect diagram and residual plot are drawn, show the prediction ability and error distribution of model.By combining LASSO feature selection and CatBoost regression model, the application improves the prediction accuracy and stability of chlorine ion diffusion coefficient, and can effectively handle complex regression problems.

The present application relates to the technical field of industrial big data processing, artificial intelligence and industrial process control, and particularly relates to a sintering state prediction method based on dynamic graph and topological perception hierarchical interaction, which comprises the following steps: S1, constructing an initial sample matrix based on original time series data of a sintering industrial site; S2, constructing a dynamic adjacency matrix in combination with a historical state feature matrix; S3, grouping the time feature and the dynamic graph adjacency matrix, and sequentially performing in-group convolution and inter-group interaction convolution to extract a current time granularity spatiotemporal representation; S4, predicting future parameters according to the spatiotemporal representation and a control instruction feature matrix. The method of the present application can respond to non-stationary working conditions such as raw material property changes and equipment state fluctuations in a timely manner, and effectively overcomes the problem of sudden drop in prediction performance of traditional static graph models when the working condition changes.

This invention belongs to the field of cutting parameter optimization technology, specifically relating to a multi-objective cutting parameter optimization method, including: acquiring cutting parameters and machine tool spindle power signals, calculating cutting specific energy, and simultaneously acquiring surface roughness; constructing a BP neural network model, optimizing the connection weights and thresholds of the BP neural network model using the Harris Eagle optimization algorithm to minimize the fitness function, and establishing a cutting specific energy prediction model; constructing a surface roughness prediction model with the encoded variables of cutting parameters as input and surface roughness as output, and establishing a multi-objective cutting parameter optimization model; solving the multi-objective cutting parameter optimization model to obtain the Pareto non-dominated solution set, using the entropy weight method for comprehensive decision-making, and selecting the solution with the largest comprehensive evaluation value as the optimal cutting parameter combination. This invention, through methods such as the BP neural network improved by the Harris Eagle optimization algorithm, reduces machining energy consumption, improves surface roughness, and enhances overall machining performance.

The application provides a table tennis competition point management system and method, which comprises a multi-modal data acquisition module, a hierarchical calculation processing module, a dynamic point generation module, a blockchain storage module, an intelligent decision module and an elastic service module working cooperatively; the multi-modal data acquisition module is configured to receive physical motion signals and environmental state signals of a competition subject through a distributed sensor network, output time-space correlated multi-source heterogeneous data streams, and comprises a biomechanics feature acquisition unit, an instrument motion trajectory capture unit and an environmental parameter perception unit; the hierarchical calculation processing module receives original data streams from the acquisition module, performs data cleaning through an edge node, implements feature fusion through a fog calculation node, conducts deep correlation analysis through a cloud platform, and outputs a standardized feature data set; and the dynamic point generation module receives the standardized feature data set and historical competition data.

The present application relates to the technical field of computer modeling, and particularly relates to a three-dimensional geological model automatic generation method, system and electronic equipment, the method comprises the following steps: obtaining three-dimensional geological modeling requirement data and modeling data input by a user, and processing the modeling data; generating a modeling instruction stream through a large model after reinforcement learning, and performing feasibility verification on the generated modeling instruction stream; if the verification result is feasible, the modeling instruction stream is sent to a three-dimensional geological modeling software to generate a three-dimensional geological model; obtaining feedback information of a modeling expert adjusting the three-dimensional geological model and diagnosing and analyzing the modeling instruction stream generated by the large model, and storing the feedback information in a knowledge base; if the verification result is unfeasible, problem matching is performed in the knowledge base, and the result is input to the large model, after analysis, a modeling instruction stream is regenerated to drive the three-dimensional modeling software to model; thus, the present application can automatically generate a three-dimensional geological model.

This invention discloses a multi-objective optimization method for motors that integrates neural networks and NSGA-III, relating to the fields of motor body design optimization and neural learning technology. The method first clarifies the optimization objectives and design variable constraints of maximizing average torque, efficiency, and minimizing torque ripple. Samples are collected and preprocessed through central composite design and finite element simulation. An FNN-CNN hybrid mapping model is constructed, simultaneously extracting local and global features, and establishing a high-precision mapping relationship with a MAE less than 3% as the training termination condition. Based on NSGA-III, θ dominance relations, density indices, and FC functions are introduced to optimize Nadir point estimation and genetic operations, achieving multi-objective collaborative optimization. Finally, the optimization effect is verified through finite element simulation. This invention improves mapping accuracy and the uniformity of the optimized solution distribution, has a faster convergence speed, and strong robustness, making it applicable to various permanent magnet motors and effectively solving the problems of inaccurate modeling and inefficient convergence in traditional methods.

This invention discloses a high-order all-drive motor modeling and control method based on a strict feedback structure, belonging to the technical field of motor control. The method includes: establishing fundamental electrical-mechanical-force coupling equations to describe the coupling relationship between the electrical subsystem, mechanical motion subsystem, and electromagnetic thrust subsystem; performing Taylor expansion on the fundamental coupling equations and constructing coupled dynamic equations by combining instantaneous electric power; converting the coupled dynamic equations into a high-order strict feedback nonlinear system model; converting the high-order strict feedback nonlinear system model into a high-order all-drive model; constructing a trajectory tracking control law; solving for the feedback gain vector using a parameterized configuration method; and performing Lyapunov stability analysis on the all-drive closed-loop LSRM system to ensure system stability and tracking performance. This invention provides a unified modeling of the high-order nonlinear coupling relationship of linear switched reluctance motors, effectively preserving the physical characteristics of the system, improving model accuracy, and expanding the applicable operating conditions of the system.

The application discloses a rope driving mechanical tracking control method based on TDE and recursive non-singular terminal sliding mode, and relates to the technical field of mechanical hand control. The application comprises the following steps: through mechanical dynamics analysis, the dynamics standard equation of the mechanical hand is determined and is converted into a TDE simplified equation, and a tracking error is defined; an FNTSM function containing a proportional control parameter is constructed, a recursive non-singular terminal sliding mode variable is formed in combination with a recursive integral term; an interference adaptive law is designed to dynamically adjust the proportional control parameter, the convergence speed and the chattering characteristics of the tracking error are analyzed, and a composite reaching law is generated; based on the composite reaching law and the updated TDE simplified equation, a generalized output vector of a joint controller containing a fluctuation parameter is extracted, a complete control law is formed, the generalized vector is output in real time through the control law, the motion trajectory of the mechanical hand is simulated, and high-precision tracking control is realized.

The application belongs to the technical field of satellite ionospheric electron density modeling system, and particularly relates to a Zhangheng No.1 satellite ionospheric electron density modeling system based on a CNN-LSTM hybrid architecture, which is realized through the cooperation of four modules, namely a data preprocessing module, a feature engineering module, a model construction and training module, and a model verification and application module. Each module is adaptively designed in view of the 97° high-inclination orbit characteristics of the Zhangheng No.1 satellite, the EDP data characteristics of the L03 level ionospheric occultation, and special application requirements, and sequentially completes the processing of satellite ionospheric data, feature extraction and optimization, construction and training of the CNN-LSTM hybrid model, model verification, and special scene application, thereby performing high-precision modeling of the ionospheric electron density of the Zhangheng No.1 satellite.

This invention discloses a hierarchical hypergraph neural network method and system for rainfall prediction in the field of meteorological station rainfall prediction technology. The method includes: extracting features from acquired meteorological index data and meteorological text data of each meteorological station to obtain meteorological features for each station; performing relationship interaction on the meteorological features of each meteorological station to obtain relationship features for each station; and concatenating the meteorological features and relationship features of each station and inputting them into a feedforward neural network for rainfall prediction to obtain the predicted rainfall for each meteorological station. This invention, by introducing an attention mechanism with time decay and suppression effect capture capabilities and a dynamically adaptive hypergraph message passing mechanism, makes the modeling of the temporal dynamics of rainfall processes more accurate, adapts to dynamic weather environments, and achieves precise prediction of rainfall at meteorological stations.

The application discloses a high-precision tunnel three-dimensional reconstruction method and device based on a CP III network, and the reconstruction method comprises the following steps: S1, establishing a CP III network, and making the coordinates of each base mark known under the CP III network, and numbering each base mark; S2, providing a holder trolley, and making the holder trolley drive along a tunnel to acquire point cloud data of the tunnel, detection coordinates of each base mark and a detection running track of the holder trolley; S3, sticking the acquired detection coordinates of the base mark into a coordinate system of the CP III network, and comparing the detection coordinates of the base mark with the coordinates of the control points, correcting the detection running track section at the base mark where an error is found, and correcting corresponding point cloud data; and S4, generating a three-dimensional model of the tunnel according to the point cloud data. The control points of the CP III network with known coordinates in the tunnel are taken as reference points, the detected tunnel holder data is corrected, more accurate tunnel holder data is obtained, and the modeling precision of the three-dimensional reconstruction of the tunnel is improved.

The application provides a SAR constellation area effective coverage evaluation method and system, the method comprises the following steps: S1, a constellation comprising a plurality of satellites is constructed to periodically cover a target area; S2, for the target area, the blind area state of each ground point under single satellite transit is calculated based on a digital elevation model and satellite orbit parameters, and a binary blind area matrix is output; S3, the single effective coverage is calculated based on the binary blind area matrix; S4, the cumulative effective coverage is further calculated and output according to the calculation result of the single effective coverage. The application combines the digital elevation model and the slant range monotonicity analysis, solves the technical problem that the complex terrain area overlap / shadow blind area is difficult to accurately quantify, realizes a sub-pixel level terrain blind area detection system, and significantly improves the modeling accuracy of SAR observation in a complex deformation environment.

The invention discloses a UE scene localization migration and reconstruction method and system based on Gaussian sputtering, and belongs to the technical field of digital scene construction and virtual reality. Track planning is carried out through a virtual camera matrix, a high-definition image sequence is output, camera parameters are extracted synchronously, left and right hand coordinate system conversion is realized by adopting a coordinate system mapping algorithm, and the real-time migration and reconstruction of the UE scene are realized. An initial point cloud is generated in combination with denoising sharpening processing and the depth map, and three-dimensional reconstruction convergence is accelerated; constructing an initial model by using a spherical harmonic function and a covariance matrix, reducing errors by optimizing renderer iteration, exporting a point cloud format, and constructing a real-time rendering framework by means of block sorting and an Alpha mixing technology; and parameters are dynamically adjusted for performance fluctuation, point cloud density information is fused, light and shadow consistency is ensured, and finally a stable high-fidelity scene model is formed. According to the method, the modeling precision and the rendering efficiency of the virtual reality scene are remarkably improved, and reliable support is provided for interactive application.

This invention relates to the field of parameter field cascade reconstruction technology, and discloses a method for cascade reconstruction of three-dimensional water vapor parameter fields based on sparse GNSS station network data. The method includes: 1. Collecting observation data from m GNSS stations sparsely distributed within a target area; 2. Obtaining the slant path water vapor content of all GNSS signals corresponding to each GNSS station; 3. Establishing a water vapor tomography equation set for each GNSS station, using a single GNSS station as the basic unit; 4. Obtaining the local three-dimensional water vapor parameter field of each GNSS station; 5. Dividing the entire target area into spatial units with the same spatial resolution as the local three-dimensional water vapor parameter field of each station, and calculating the spatial correlation coefficient between each spatial unit and the local three-dimensional water vapor parameter field of each station at each height level; 6. Constructing a recursive model to obtain the three-dimensional water vapor parameter field of the target area. This invention achieves high-precision reconstruction of the three-dimensional water vapor parameter field of the target area under sparse GNSS station network conditions, improving the inversion accuracy of three-dimensional water vapor parameter fields based on sparse GNSS station networks.

This invention discloses a method and system for neuroimaging feature extraction and toxicity assessment based on scale-adaptive self-supervised learning, comprising: acquiring confocal microscopic images and structured morphological quantification indicators; employing a flexible patch dynamic embedding strategy and a fixed-length masking strategy to perform self-supervised pre-training of an image-scale-adaptive mask autoencoder; extracting multi-scale latent representations from the images using a multi-scale fusion strategy, and fine-tuning the pre-trained feature encoder for downstream tasks; using the fine-tuned feature encoder to extract a visual data stream containing multi-scale latent representations from the images, and semantically embedding the structured morphological quantification indicators to extract a semantic label stream, fusing the two types of feature streams; inputting the fused joint features into a prediction network to obtain predicted values ​​of behavioral indicators of the target organism and determine the neurotoxicity quantification assessment result. This invention enables high-throughput, high-precision, and low-cost automatic neurotoxicity assessment.

The application provides a robot dynamic error compensation method based on a capsule neural network, adopts Newton-Euler method to establish a robot dynamics model, and performs linearization and dynamics parameter identification; uses joint motion data as input to obtain predicted torque for model training by using the linearized dynamics model after the dynamics parameter identification, and calculates error between the predicted torque and actual torque; regards normalized parameters as vectors in a vector space, calculates inner products of the vectors to form a Gram matrix and converts the Gram matrix into a two-dimensional image; learns error between the predicted torque and the actual torque by using a capsule neural network, and obtains a trained capsule neural network; obtains real-time prediction error by using the trained capsule neural network, and compensates real-time predicted torque obtained by the linearized dynamics model after the parameter identification by using the real-time prediction error. The application realizes high-precision and adaptive dynamic error compensation.