Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

51 results about "Sparse regression" patented technology

Physics-based method for discovering control equation from scarce and noise data

ActiveCN120105365AKnowledge representationNeural learning methodsComplex dynamic systemsNoisy data
The invention discloses a physics-based method for discovering a control equation from scarce and noise data. According to the method, a physical information neural network and a sparse regression method are combined, and a partial differential control equation of a dynamic system is found from scarce and noisy data. Firstly, the size of a candidate function library is effectively reduced through a dimension verification method. And then modeling a physical system and calculating a candidate function by using the strong nonlinear fitting capability and automatic differential characteristic of the deep neural network. And finally, the form of a control equation and the coefficient of the equation can be obtained through sparse regression, and the coefficient is finely adjusted through DNN. The method not only can discover the control equation from the data, but also can obtain the network model to realize the response prediction of the dynamic system. The method is easy to implement, efficient, high in precision and high in universality. The method can be widely applied to physical knowledge mining, modeling and reasoning of a complex dynamic system.
Owner:ZHEJIANG UNIV

Multi-source ecological factor-based comprehensive evaluation method and system for biodiversity influence

The invention provides a biodiversity influence comprehensive evaluation method and system based on multi-source ecological factors, and relates to the technical field of biodiversity analysis, and the method comprises the steps: obtaining multi-source ecological factor data of a geological disaster control influence region, and carrying out the maximum likelihood estimation and kernel function learning processing of the multi-source ecological factor data to obtain a space-time intensity field; performing adaptive time-frequency decomposition processing based on the time-space intensity field and the time sequence of the community functional traits to obtain a multi-scale modal set; performing low-rank dynamic recognition processing on the multi-scale modal set to obtain an approximate operator and control gain pair; performing sparse regression modeling processing to obtain a nonlinear state equation set; and carrying out distribution displacement measurement processing based on the nonlinear state equation set and preset quadrat baseline distribution to obtain a comprehensive evaluation result of the biodiversity influence. According to the method, the problems of single index, rough modeling and lack of multi-scene prediction capability in the prior art are solved.
Owner:SICHUAN FORESTRY RES INST (SICHUAN FORESTRY IND RES & DESIGN INST)

Model predictive control method and system for seismic simulation shaking table, medium, and terminal

A model predictive control method for a seismic simulation shaking table, comprising: S1: acquiring offline data of input signals and output signals of a seismic simulation shaking table; S2, on the basis of the acquired offline data, using sparse regression to identify a kinetic equation of the seismic simulation shaking table; and S3, on the basis of the kinetic equation, predicting output signals of the seismic simulation shaking table under different input signals, and using an optimization model to correct the input signals; moreover, on the basis of real-time data of the input signals and the output signals collected by the seismic simulation shaking table in the control process, performing parameter updating on the kinetic equation; and repeating the processes of prediction, optimization of input signals, and parameter updating until a preset target signal has been reproduced. The method has good precision and stability, and can effectively improve the control performance of the seismic simulation shaking table. The present invention further relates to a model predictive control system for a seismic simulation shaking table, a medium, and a terminal.
Owner:CENT SOUTH UNIV +1

Intelligent monitoring system and method for realizing dynamic performance adjustment for distributed photovoltaic network

The invention discloses an intelligent monitoring system and method for realizing dynamic performance adjustment for a distributed photovoltaic network, and relates to the technical field of intelligent power grids. The system comprises a data acquisition module used for acquiring image data of a photovoltaic panel and operation data of a photovoltaic power station in real time; the task scheduling module is used for coordinating the dynamic transmission of the image data and the operation data between the edge computing node and the cloud; the data processing module is used for carrying out preprocessing and feature extraction on the image data and the operation data; the dynamic estimation module is used for inputting the extracted features into a sparse regression model and predicting the dynamic state of the photovoltaic network by using an unscented Kalman filter; and the monitoring feedback module is used for remotely and dynamically adjusting the angle or position of the photovoltaic panel, starting and stopping equipment and / or adjusting parameters according to the prediction result of the dynamic state. The overall operation efficiency of the photovoltaic network can be remarkably improved, the operation and maintenance cost can be reduced, and the service life of equipment is prolonged.
Owner:NANJING SHENDA ENG TECH CO LTD

Mechanism information analysis method of nonlinear dynamic control system based on causal diagram

PendingCN119106220AFacilitate parameter optimizationFacilitate sparse learningBiological modelsComplex mathematical operationsRobustificationTheoretical computer science
The invention relates to a mechanism information analysis method of a nonlinear control system, which combines a space-time diagram convolutional network with mechanism information modeling, and learns a control equation of the system from observation data with noise. Firstly, causal graph coding based on transfer entropy is designed to obtain an adjacent matrix with causal significance between variables. Then, a space-time diagram convolutional network is used to obtain an approximate solution of a system variable. On this basis, automatic differentiation is applied to obtain a basic derivative, a dictionary of candidate algebraic terms is formed, and sparse regression is used to obtain a coefficient matrix to determine an explicit formula of the control equation. The invention also designs a new cross combination optimization strategy to learn heterogeneous parameters including neural network parameters and control equation coefficients. According to the method, the advantages of causal representation, deep learning, automatic differentiation and sparse regression are fused aiming at the limitation of numerical differentiation and redundancy calculation of an existing control equation learning method, the robustness and universality are high, and theoretical and practical significance is achieved for mechanism information discovery of a nonlinear dynamic system.
Owner:SHENYANG INST OF AUTOMATION - CHINESE ACAD OF SCI

Neutron transport constitutive relation and low-dimensional control equation modeling method based on sparse regression

The invention discloses a neutron transport constitutive relation and low-dimensional control equation modeling method based on sparse regression, and the method comprises the steps: enabling a state parameter of each state and a physical parameter calculated through a neutron transport program to form a neutron current matrix under a plurality of state parameters, and filling a sparse regression physical parameter matrix with the built fitting physical parameter, calculating an inverse matrix by using a least square method to obtain a coefficient matrix, and assembling to obtain a neutron transport constitutive relationship under the plurality of state parameters; according to the method, the characteristic that the calculation result of the neutron transport equation is accurate is utilized, the reason that the calculation result of the neutron diffusion equation is large in deviation in a small-scale range is grasped, the neutron transport constitutive relation is obtained in combination with sparse regression fitting, then the low-dimensional macroscopic neutron transport control equation is obtained, the theoretical derivation difficulty is small, the research progress is easier, and the method has good application prospects. And the effect of improving the calculation precision is obvious, the applicability is better, the difference from the situation under the actual reactor core working condition is small, and certain practicability is achieved.
Owner:SUN YAT SEN UNIV

Light beam transmission characteristic model discovery and reconstruction method based on sparse physical information neural network

The invention discloses a physical model discovery and reconstruction method for light beam transmission characteristics based on a sparse physical information neural network, and belongs to the technical field of nonlinear optics and deep learning intersection, and the method comprises the following steps: S1, making a sparse and low-fidelity data set; s2, establishing a candidate function library of items possibly existing in a data back model by using automatic differentiation; S3, constructing a physical information neural network model with sparse regression to discover a physical model structure; s4, coefficient fine tuning is carried out on the found physical model, and high-fidelity data is reconstructed.The method is suitable for model discovery of light beam transmission data supported by a constant coefficient or variable coefficient partial differential equation, scarce and noisy low-fidelity data can be utilized, and high-fidelity data can be obtained. A simple physical model is found according to a physical information neural network and sparse regression, and a light beam transmission dynamic process is reconstructed, so that subsequent theoretical analysis and research on light beam transmission characteristics are facilitated.
Owner:ZHEJIANG FORESTRY UNIVERSITY

Low-complexity random access channel coding and decoding method based on sparse regression code

The invention discloses a low-complexity random access channel coding and decoding method based on a sparse regression code, which belongs to the technical field of random access channel coding and decoding, and comprises the following steps of: acquiring to-be-sent information of all active users through a sending end; wherein the information to be sent comprises identity identification information and message information; encoding and transmitting the to-be-sent information, and repeating a plurality of stages; receiving signals are obtained through a receiving end, and whether the whole transmission process is finished after the tth stage is completed or not is judged; wherein the obtained receiving signal transmitted in the first stage is the superposition of a plurality of identity identification sub-code words and noise; the received signals transmitted from the second stage to the t-th stage are all superposition of a plurality of message sub-code words and noise; if the transmission is finished, determining the number of active users according to the number of the transmission stages, and decoding the received signals of the first stage to obtain identity information of all the active users; and decoding the received signals transmitted from the second stage to the t-th stage to obtain message information of all active users.
Owner:BEIHANG UNIV

Power failure information monitoring method and system based on real-time position

The invention discloses a power failure information monitoring method and system based on a real-time position, and relates to the technical field of power failure information monitoring, and the method comprises the steps: collecting the data of an intelligent electric meter, calculating the power flow of a line, carrying out the sparse regression through an LASSO algorithm, and recognizing a key line which is strongly related to a power failure event; window sliding processing is carried out based on the key line, and an event matrix is generated according to statistical characteristics; the TF-IDF value of each feature is calculated based on the time frequency of the statistical features in the event matrix. According to the method, power flow data and voltage phase angle data are used as feature matrixes, the sparsity of LASSO regression is beneficial to eliminating non-critical line data with large noise or interference, the complexity and redundancy of a model are reduced, feature importance of a critical line in different time windows is enhanced through a TF-IDF weighting matrix, and the accuracy of the LASSO regression is improved. According to the method, the SVM model is helped to recognize the power failure event more accurately, and after SVM preliminary classification, isolated misjudgment can be corrected through MDC auxiliary recognition and clustering analysis.
Owner:YUXI POWER SUPPLY BUREAU OF YUNNAN POWER GRID

Nonlinear multi-scale dynamic system control mechanism inference method based on Koopman operator theory

The invention discloses a nonlinear multi-scale dynamic system control mechanism inference method based on a Koopman operator theory, and the method comprises the steps: building an order reduction mechanism model of a dynamic system from observed time series data based on the Koopman operator theory; firstly, dynamic sampling is carried out on observation time sequence data based on a self-adaptive sampling strategy designed by the method, the sampling rate is adjusted according to data changes, and information of different time scales is effectively separated. Thirdly, performing local data smoothing and numerical differentiation on the discrete sampling points by applying cubic B-spline interpolation; on the basis, a Block-Hankel matrix is innovatively constructed by using the nonlinear modeling capability of a Koopman operator theory so as to capture the multivariate dynamic characteristics of the system; and a main dynamic mode is extracted by using a kernel singular value decomposition technology, and system identification is expanded to a nonlinear power system. And finally, the SINDy algorithm is used for identifying the reduced-order model, and accurate modeling of the complex nonlinear multi-scale system is realized. According to the method, the advantages of Koopman embedding and sparse regression are fused on the basis of the multi-scale data processing capability of adaptive sampling, so that the method has relatively high robustness and universality, and has theoretical and practical significance for mechanism modeling of a nonlinear and multi-scale system.
Owner:SHENYANG INST OF AUTOMATION - CHINESE ACAD OF SCI

Dam safety state intelligent prediction method and system based on regression modeling

The invention relates to a dam safety state intelligent prediction method and system based on regression modeling, and the method comprises the steps: carrying out the self-adaptive wavelet packet decomposition of structural data collected by a sensor, so as to extract a time sequence feature, and carrying out the differential coding of an environment variable, so as to generate a covariable feature; the method comprises the following steps: mapping a text event into vector features through a knowledge graph, inputting the features into a mixed primary function library formed by a physical primary function and a data primary function, and constructing a sparse regression model through a dynamic gating mechanism; a physical constraint term based on a stiffness matrix and an external load matching error is introduced into the model, a causal graph is dynamically updated through an incremental PC algorithm, and a causal regularization term is added into a loss function to improve the adaptability of the model. The method is suitable for dam structure health monitoring, intelligent early warning, long-term trend analysis and other application scenes, and has good real-time performance, interpretability and small sample generalization ability.
Owner:HUANENG SICHUAN HYDROPOWER CO LTD +2

Network dynamics symbol regression method based on LLM and Bayesian optimization

The invention is suitable for the technical field of computers, and provides a network dynamics symbol regression method based on LLM and Bayesian optimization. According to the method, a primary function library does not need to be predefined, natural language is fused into physical priori to guide LLM to generate an equation, the quality of the generated equation is further improved, and the search process is accelerated; and meanwhile, a Bayesian optimization serialization optimization method is utilized to efficiently search a globally optimal solution, so that the accuracy of the generated equation is improved, and an efficient and intelligent solution is provided for symbol model search. Experiments in various network dynamics scenes show that under the condition that a primary function library does not need to be additionally predefined, candidate equation generation through LLM is combined with efficient search of Bayesian optimization, prior knowledge is effectively fused, equation discovery precision and search efficiency are greatly improved, and the method is suitable for large-scale popularization and application. The accuracy of the obtained equation is far higher than that of the existing methods based on combination of GNN and GP (GG), predefined function library sparse regression (TPSINDy) and the like.
Owner:JILIN UNIVERSITY

A physics-based method for discovering governing equations from scarce and noisy data

The application discloses a method for discovering control equations from scarce and noisy data based on physics. The application combines a physical information neural network and a sparse regression method to discover the partial differential control equations of a dynamic system from scarce and noisy data. Firstly, the size of the candidate function library is effectively reduced through a dimension verification method. Then, the powerful nonlinear fitting capability and automatic differentiation characteristics of a deep neural network are used to model a physical system and calculate candidate functions. Finally, the form of the control equation and the coefficients of the equation are obtained through sparse regression, and the coefficients are fine-tuned through a DNN. The method can not only discover control equations from data, but also obtain a network model to realize response prediction of a dynamic system. The method is simple, efficient, high-precision and highly universal, and can be widely applied to physical knowledge mining, modeling and reasoning of complex dynamic systems.
Owner:ZHEJIANG UNIV

Self-adaptive ambient light matching method and system for color temperature of electronic whiteboard

PendingCN121686970ACathode-ray tube indicatorsWhiteboardColor adaptation
The invention discloses a self-adaptive ambient light matching method and system for the color temperature of an electronic whiteboard, ambient light information is collected through a common camera, the color characteristics of ambient light are sensed in real time in combination with an intelligent spectrum reconstruction algorithm and a color temperature matching strategy, and the color temperature displayed by the electronic whiteboard can be dynamically matched with the ambient light, so that the color temperature of the electronic whiteboard can be dynamically matched with the ambient light. Comprising the following steps: 1, pre-constructing a strategy library in which a plurality of groups of color temperature-brightness parameters and matched scenes are stored; 2, acquiring an ambient light image through a wide-angle camera, and performing multi-frame average noise reduction and exposure calibration on the ambient light image to obtain an RGB measured value of ambient light; 3, based on a preset PCA spectrum basis vector library, reconstructing the RGB measured value of the ambient light into an ambient spectrum in a range of 380-780 nm by adopting a non-negative sparse regression algorithm; and 4, calculating an environment color temperature according to the reconstructed environment spectrum, selecting a color temperature-brightness parameter matched with the environment color temperature from a strategy library in combination with a CIE CAM02 color adaptation transformation model, and adjusting the electronic whiteboard.
Owner:XIAMEN PRIMA TECH

Low-voltage distribution network voltage prediction method, system and device and storage medium

The invention relates to the technical field of voltage prediction, and discloses a low-voltage power distribution network voltage prediction method, system and device and a storage medium, and the method comprises the steps: screening historical measurement data from the measurement data of historical moments according to the power data of a current moment and each historical moment; performing equidistant division on the historical measurement data to obtain a plurality of data subsets, and obtaining historical measurement change submatrixes of the data subsets according to the historical measurement data; performing local sparse regression on the historical measurement change sub-matrix to obtain a sensitivity matrix of the data subset; performing similarity matching on the power data variation at the current moment and each data subset to obtain an optimal matching data subset; and according to the voltage data variation at the current moment and the sensitivity matrix of the optimal matching data subset, obtaining voltage prediction data of each node. According to the invention, accurate and efficient voltage prediction can be realized, so that the operation safety and stability of the low-voltage power distribution network are ensured.
Owner:STATE GRID ZHEJIANG ELECTRIC POWER CO LTD HANGZHOU POWER SUPPLY CO +1

Neutron transport constitutive relation and low-dimensional control equation modeling method based on sparse regression

The application discloses a neutron transport constitutive relation and low-dimensional control equation modeling method based on sparse regression, which comprises the following steps: a neutron flow matrix under multiple state parameters is formed by state parameters of each state and physical quantity parameters calculated through a neutron transport program; fitting physical quantity parameters are filled into a sparse regression physical quantity matrix; a coefficient matrix is obtained by using a least square method to calculate an inverse matrix; and a neutron transport constitutive relation under multiple state parameters is obtained by assembling. The application utilizes the accurate characteristics of the calculation result of the neutron transport equation, grasps the cause of the larger deviation of the calculation result of the neutron diffusion equation in a small scale range, combines sparse regression fitting to obtain the neutron transport constitutive relation, and further obtains a low-dimensional macroscopic neutron transport control equation. The theoretical derivation is less difficult, the research progress is easier, the effect of improving the calculation precision is more obvious, the applicability is better, the gap with the situation under the actual reactor core working condition is not large, and the application has certain practicality.
Owner:SUN YAT SEN UNIV

Blood type and microbiological detection data correlation analysis method based on statistical model

The invention discloses a blood type and microbiological detection data correlation analysis method based on a statistical model, and belongs to the technical field of medical treatment. Through data preprocessing and feature fusion, standardized processing and dimensionality reduction expression of blood type information and microbiological detection high-dimensional features are realized, and a joint feature matrix is constructed; an integrated Gaussian mixture network is adopted to carry out multi-level distribution modeling, and high-resolution clustering of microbial spectrums of different blood group populations is realized in combination with a category cascade discrimination mechanism. Further, a complex nonlinear relation between the blood type and the microorganism data is dynamically analyzed in a multi-scale space by utilizing a self-adaptive multi-scale correlation mining algorithm. Then, screening and analyzing key causal variable pairs of the blood type-microbiome through sparse regression and causal inference by adopting an interpretable correlation modeling and variable screening method; the accuracy and generalization ability of correlation analysis are improved, and accurate prediction and high-reliability risk judgment are realized.
Owner:CHONGQING ACAD OF METROLOGY & QUALITY INST

Sparse regression code encoding and decoding method for reaching capacity region of MIMO multiple access channel

The application discloses a sparse regression code encoding and decoding method for MIMO multiple access channel capacity domain reachability, which adopts a sparse regression code as a forward error correction code of a system. A random transformation matrix is introduced at an encoding end to perform encoding processing on a block sparse signal, so that an equivalent channel matrix is converted into a universal matrix, thereby ensuring decoding reliability of an MA-OAMP receiver under any channel matrix condition, and overcoming the dependence of a traditional AMP receiver on statistical characteristics of an observation matrix. Encoding of the sparse regression code is performed according to an optimal encoding criterion based on the MA-OAMP receiver, power distribution is reasonably performed on sparse blocks, and a time division technique is combined, so that approximation to the whole MIMO-MAC capacity domain is realized.
Owner:ZHEJIANG UNIV

Nonlinear multi-scale dynamic system control mechanism inference method based on koopman operator theory

The application discloses a nonlinear multi-scale dynamic system control mechanism inference method based on Koopman operator theory, which is based on the Koopman operator theory and establishes a reduced-order mechanism model of a dynamic system from observed time series data. First, the observed time series data is dynamically sampled based on the adaptive sampling strategy designed by the application, the sampling rate is adjusted according to the data change, and the information of different time scales is effectively separated. Then, cubic B-spline interpolation is applied to locally smooth and numerically differentiate the discrete sampling points. On this basis, the nonlinear modeling capability of the Koopman operator theory is used to innovatively construct a Block-Hankel matrix to capture the multi-element dynamic characteristics of the system; and the kernel singular value decomposition technology is used to extract the main dynamic mode, so that the system identification is extended to a nonlinear dynamic system. Finally, the SINDy algorithm is used to identify the reduced-order model, so as to realize accurate modeling of a complex nonlinear multi-scale system. On the basis of the adaptive sampling and multi-scale data processing capability, the application combines the advantages of Koopman embedding and sparse regression, has strong robustness and universality, and has theoretical and practical significance for mechanism modeling of nonlinear and multi-scale systems.
Owner:SHENYANG INST OF AUTOMATION - CHINESE ACAD OF SCI

A wave field reconstruction method

The application discloses a wave field reconstruction method, which comprises the following steps: firstly, obtaining a standardized space-time data set, the data set containing effective wave height, average wave period and wave direction; then, constructing a low-rank joint wave feature subspace according to the data set; after obtaining sparse observation data and constructing an observation matrix and an observation vector based on the observation data, performing sparse regression on each variable of the data set based on the feature subspace, the matrix and the vector, obtaining a preliminary spatial distribution result of each variable, then performing Gaussian weighted fusion smoothing according to the distribution result, obtaining a preliminary reconstruction result, then determining the dynamic weight of each variable based on the reconstruction result, and optimizing the preliminary reconstruction result according to the weight, obtaining a target reconstruction result. The embodiment of the application can effectively reconstruct a high-resolution wave field under limited observation conditions, improve the reconstruction accuracy, simplify the calculation process, adapt to complex data scenarios, and quickly output high-quality results.
Owner:JINAN UNIVERSITY

Method and apparatus for determining control equations for high noise granular flow systems

The application discloses a method and device for determining a control equation of a high-noise particle flow system, which comprises the following steps: determining the noise intensity of each data sample in a data sample set of the particle flow system; screening out high-noise data samples based on the noise intensity and a preset high-noise intensity range; randomly sampling the high-noise data samples to obtain a plurality of sub-sample sets; for each sub-sample set, firstly performing denoising processing, and then learning and mining the sub-sample set after the denoising processing based on a normalized sparse regression method to determine a control equation; and determining the control equation corresponding to the particle flow system in the high-noise intensity range based on the collective average of the control equations corresponding to the sub-sample sets. The application realizes the target of stably and accurately identifying the control equation in the high-noise particle flow system, and overcomes the problem that it is difficult to ensure the accuracy of the control equation under the condition of serious noise interference in the prior art.
Owner:CHINA UNIV OF PETROLEUM (BEIJING)

Adaptive bayesian identification method for long-term prediction of industrial chain power load

The application provides an adaptive Bayesian identification method for long-term prediction of industrial chain electric load, comprising: aggregating active power data into daily frozen power time series; performing discrete Fourier transform on the daily frozen power time series, obtaining amplitude spectrum, and calculating spectral complexity index; constructing an adaptive basis function library containing sine and cosine functions according to the screened dominant frequency; establishing a sparse regression model containing regularization, representing the daily frozen power time series as a linear combination of basis functions in the adaptive basis function library, and obtaining a sparse coefficient vector by solving an optimization problem; and using the identified sparse coefficient vector and adaptive basis function library, calculating the daily frozen power prediction value of the future time step through time index extrapolation. The method can plan the characteristics of upstream and downstream industrial chains, accurately capture multi-scale periodic structures, and has low calculation cost and high physical interpretability.
Owner:WUXI UNIV

Partial differential equation data processing method, system, storage medium, device and application

The present invention belongs to the technical field of partial differential equation processing for learning system implicit from data, and discloses a partial differential equation data processing method, system, storage medium, device and application, using attention mechanism to sample time-space data; using prior knowledge and basis function representation to construct a differential term candidate library; using multiple single-layer regression networks to form a deep network to learn the corresponding coefficients of differential terms in the candidate library; adding regularization to the loss function, reducing interference terms through sparse regression, and obtaining the final partial differential equation. The present invention can make full use of the spatial and temporal information of the observed data, reduce the amount of data required for the model, and improve the accuracy of the learned partial differential equation while ensuring efficiency; solve the problem that the traditional method of learning the partial differential equation implicit in the system through artificial experience and experimental verification is too inefficient, easily affected by human subjective factors, and difficult to find the mechanism behind the complex system.
Owner:XIDIAN UNIV

Coupling separation and state collaborative prediction system and method based on power supply characteristic spectrum

The invention belongs to the technical field of equipment state detection, and discloses a coupling separation and state collaborative prediction system and method based on a power supply characteristic spectrum. Comprising the following steps: collecting current and voltage signals of all equipment in the shared power supply topology and extracting a spectrum feature vector; carrying out random Fourier feature mapping on the spectrum feature vectors of other devices except the target device, and training a group sparse regression model to obtain a group regression coefficient set; calculating an external interference component, and obtaining a pure feature vector of the target equipment based on the external interference component; performing significance index calculation on the pure feature vector to obtain a high-dominance window set; obtaining a window abnormal value of the high-dominance window, averaging to obtain an abnormal evaluation index, comparing the abnormal evaluation index with an alarm threshold, and judging whether the state of the current target equipment is normal or not; constructing an anomaly evaluation sequence, establishing a trend function, and predicting residual safe operation time; accurate judgment and prediction of the state of the power supply equipment are realized through external coupling separation.
Owner:SHENZHEN GREAT ENERGY TECH

A symbolic regression method for network dynamics based on LLM and Bayesian optimization

The present invention is applicable to the field of computer technology and provides a network dynamics symbolic regression method based on LLM and Bayesian optimization. This method does not require a predefined basis function library, but instead uses natural language to integrate physical priors to guide the LLM to generate equations, further improving the quality of the generated equations and accelerating the search process. At the same time, it utilizes Bayesian optimization, a serialized optimization method, to efficiently search for global optimal solutions, improving the accuracy of the generated equations and providing an efficient and intelligent solution for symbolic model search. Experiments in various network dynamics scenarios have demonstrated that, without the need for an additional predefined basis function library, the method effectively incorporates prior knowledge by generating candidate equations through LLM and combining it with the efficient search of Bayesian optimization, significantly improving the accuracy of the discovered equations and the efficiency of the search. The resulting equations are far more accurate than current methods based on GNN combined with GP (GG) and sparse regression with a predefined function library (TPSINDy).
Owner:JILIN UNIVERSITY

Channel flow heat transfer prediction method and device based on machine learning and medium

The invention discloses a channel flow heat transfer prediction method and device based on machine learning and a medium, and relates to the technical field of parameter prediction.Supplementary data is generated based on an existing flow heat transfer relational expression, the working condition of rare experimental data is effectively covered, and key dimensionless parameters influencing flow resistance and heat transfer coefficients are determined; therefore, the model can focus on the most critical factors influencing the flow heat transfer relation, the ability of the model to grasp the nature of the flow heat transfer phenomenon is improved, and the flow heat transfer coefficient and the frictional resistance coefficient can be predicted more accurately. A neural network model is constructed to effectively capture a complex nonlinear relationship between an input parameter and a flow heat transfer coefficient or a frictional resistance coefficient, a sparse regression method is adopted to fit experimental data, and relational prediction data and a large amount of prediction data generated by the neural network model are explicit mathematical expressions. The problems that an existing relational expression is narrow in prediction range and poor in universality can be solved, and the prediction precision of the channel flow heat transfer relation is effectively improved.
Owner:NUCLEAR POWER INSTITUTE OF CHINA

Information processing apparatus, information processing method, and program

To provide a technique that can suitably prevent overtraining.SOLUTION: An information processing apparatus comprises: an acquisition unit that acquires target data and the number of a plurality of target models; a regression coefficient calculation unit that refers to the target data and a ratio parameter defining the ratio between the plurality of target models, and calculates a sparse regression coefficient for each of the plurality of target models; a ratio parameter calculation unit that refers to the target data and the regression coefficients to calculate the ratio parameter; and an output unit that outputs the regression coefficients calculated by the regression coefficient calculation means and the ratio parameter calculated by the ratio parameter calculation means.SELECTED DRAWING: Figure 1
Owner:NEC CORP

Breeding screening method and system based on phenotypic performance

PendingCN122637875AAlleleLinear model
The application provides a breeding screening method and system based on phenotypic expression, and relates to the technical field of plant breeding screening. The application extracts a pre-planting population from a population to be screened, obtains SNP allele codes of each plant, and determines phenotypic indexes at multiple time nodes; a genome matrix is constructed, a mixed linear model is fitted, a two-dimensional residual matrix is constructed by extracting time sequence residuals, and information entropy of each plant is calculated; key marker sites are screened and genetic stability scores are calculated by performing sparse regression; the key sites are classified according to gene functions, yield potential values are calculated, a screening index calculation rule is constructed by combining the stability scores and the population yield potential median, and finally, the screening index of the pre-planting population and the remaining plants to be tested is calculated to complete the screening of high-quality plants. The application does not need to determine the field phenotypes of all plants to be screened in the whole growth period, and solves the problems of heavy workload, long breeding cycle, and high labor and test cost in the prior art.
Owner:INNER MONGOLIA AUTONOMOUS REGION ACAD OF AGRI & ANIMAL HUSBANDRY SCI

Electronic nose array optimization method based on information theory and weighted sparse regression

PendingCN121389055ASensor arrayEngineering
The invention belongs to the technical field of gas sensors, and relates to an electronic nose array optimization method based on an information theory and weighted sparse regression, which comprises the following steps: acquiring response data of an electronic nose sensor array, and constructing an original feature matrix and a target category vector thereof according to the response data; calculating a comprehensive information theory score of each sensor in the electronic nose sensor array according to the original feature matrix and the target category vector thereof; solving the original feature matrix by using a weighted sparse optimization model according to the comprehensive information theory score of each sensor to obtain a sparse coefficient of each sensor; determining an optimal sensor subset according to the sparse coefficient of each sensor and the comprehensive information theory score; according to the method, a global optimization framework based on a weighted sparse optimization model is adopted, comprehensive information theory scores are used as prior guidance, the model can be quantized in advance, and the sensor subset approaching global optimum can be efficiently and deterministically found by utilizing the cooperation and redundancy effect between the sensors.
Owner:CHONGQING UNIV OF POSTS & TELECOMM

A machine learning based method, apparatus, and medium for predicting channel flow heat transfer

The application discloses a channel flow heat transfer prediction method and device based on machine learning and a medium, relates to the technical field of parameter prediction, generates supplementary data based on an existing flow heat transfer formula, effectively covers the working conditions with few experimental data, determines key dimensionless parameters affecting flow resistance and heat transfer coefficients, enables the model to focus on the most critical factors affecting the flow heat transfer relationship, improves the model's ability to grasp the nature of the flow heat transfer phenomenon, and can more accurately predict the flow heat transfer coefficient and friction resistance coefficient, effectively captures the complex nonlinear relationship between the input parameters and the flow heat transfer coefficient or the friction resistance coefficient by constructing a neural network model, and uses a sparse regression method to fit the experimental data, the relational expression prediction data and a large amount of prediction data generated by the neural network model into an explicit mathematical expression, so that the problem of narrow prediction range and poor universality of the existing relational expression can be solved, and the prediction accuracy of the channel flow heat transfer relationship is effectively improved.
Owner:NUCLEAR POWER INSTITUTE OF CHINA