168 results about "Laplacian matrix" patented technology

The invention relates to the technical field of temperature control, in particular to a heating and ventilation heating power constant-temperature variable-flow control method, and provides the following scheme that circulating pump frequency instructions and water temperature response data of multiple historical regulation and control periods are collected, and a dynamic track of the current water temperature in a temperature inertia field is constructed; calculating an attraction potential energy state between the current water temperature and a target constant temperature value in combination with the disturbance residual matrix; and outputting the frequency bandwidth interval of the circulating pump based on the state to realize dynamic adjustment. An adjacency matrix and a Laplacian matrix are constructed by introducing a hydraulic communication relation, spatial convolution and weighted updating are performed on disturbance residues in combination with a topological diffusion core, and accurate propagation and suppression of disturbance are realized. The method can effectively solve the problems of nonlinear response, delay effect and disturbance coupling of the water temperature to the pump frequency, and the temperature stability and the energy utilization efficiency of a heating system are improved.

The invention discloses a short-term wind speed prediction method for multiple offshore wind power plants, relates to the technical field of power system intellectualization, and constructs a dynamic graph structure fusing the correlation between geographic distance and wind speed according to the correlation between the geographic position of a target area and the wind speed, namely a multi-wind-plant connected graph, and represents the spatial topological relation of a wind power plant group. A wind power plant group is mapped into a node network by constructing a dynamic graph structure fusing geographic distance and wind speed correlation, spatial dependence intensity between nodes is quantized by using a weighted adjacent matrix, spectral domain convolution operation is carried out by a graph convolution network based on a normalized Laplacian matrix, and the spatial dependence intensity between nodes is quantized by using a normalized Laplacian matrix. Efficient neighborhood feature aggregation is achieved through Chebyshev polynomial approximation, complex spatial association caused by geographic position difference and meteorological condition interaction can be accurately captured, the defect of non-Euclidean spatial relationship modeling in a traditional method is overcome, and the representation capacity of the spatial dependency relationship in the multi-wind-power-plant environment is remarkably improved.

The invention belongs to the field of energy storage monitoring, and provides an energy storage facility safety early warning protection method and system, and the method comprises the steps: obtaining the temperature, heating rate and equivalent impedance data of each monitoring unit of a battery cluster; constructing a physical adjacency matrix based on the physical space arrangement relation of the monitoring units; fusing the physical adjacency matrix and the electrical adjacency matrix according to a preset weight; constructing a graph structure based on the comprehensive adjacent matrix, and calculating a Laplacian matrix of the graph; calculating a node neighborhood residual error for the temperature and the heating rate based on a Laplacian matrix to obtain a temperature consistency residual error; calculating a node neighborhood residual error for the equivalent impedance to obtain an electrical consistency residual error; calculating a temperature prediction residual error under the time sequence prediction model with graph regularization constraint; and constructing an early abnormal score of the node, adaptively setting a quantile threshold according to the healthy operation data distribution, and carrying out graded early warning judgment on the early abnormal score. The accuracy of safety early warning of the energy storage facility can be improved.

The invention relates to the technical field of natural language processing, and discloses a generated text quality processing method and system based on a large language model, and the method comprises the steps: constructing a logic topological graph and a Laplacian matrix of an original generated text, and extracting a feature value sequence; recognizing a logic bearing wall based on the attention gradient and generating an anti-fact contrast text; calculating a logic collapse index by combining the difference between the map characteristic values of the original text and the contrast text and the logic polarity overturning condition; and judging the quality of the generated text according to the logic collapse index, blocking the text which does not pass the judgment, and shaping and resampling the output probability value of the error position by utilizing the spectrum difference information to generate a new text. The method does not need to depend on an external knowledge base, quantifies the stability of text logic through anti-fact interference and spectrum analysis, and identifies a high-risk illusion text; and a wrong logic path is automatically corrected through Logits shaping, so that the logic self-consistency of the generated content is improved on the premise of ensuring the semantic smoothness of the text language.

The invention relates to the technical field of data processing, in particular to a multi-modal sensor data real-time fusion processing method for a smart community. According to the method, a sensor network topological graph is constructed, a connection weight is optimized, distributed clock synchronization is realized by using a graph Laplacian matrix, and clock drift prediction and compensation are performed in combination with a fractional Brownian motion model; performing wavelet transform decomposition on the sensor data after time sequence alignment, calculating each scale Hurst index, predicting a load trend through a fractal prediction model, and outputting an optimal resource allocation scheme through hybrid evolution calculation; the method comprises the following steps: constructing multi-modal sensor data into a graph structure, extracting node features by using a graph convolutional neural network, obtaining global feature representation by using a self-attention mechanism, and performing anomaly detection classification in combination with a resource utilization rate and a prediction error; an anomaly detection feedback mechanism is established, and Laplacian matrix eigenvalues and weight parameters are dynamically adjusted; the real-time performance, the accuracy and the robustness of data fusion processing are improved.

The invention discloses a dual-channel label-guided multi-label feature selection method and system, and belongs to a feature engineering technology. The method mainly comprises the steps of obtaining a feature matrix and a positive label matrix of a multi-label data set, performing logic negation on the positive label matrix to generate a mirror image negative label matrix, and constructing a graph Laplacian matrix based on the feature matrix; constructing a multi-label model based on the preprocessed data, wherein an objective function of the multi-label model at least comprises a positive label regression loss item, a negative label regression loss item, a label alignment constraint item, a graph regularization item and a sparse constraint item; constructing an optimization function through relaxation processing constraint and in combination with a Lagrangian multiplier method, iteratively solving the objective function according to a KKT condition, and evaluating feature importance based on a projection matrix for associating features and positive tags after iterative convergence; according to the method, the requirement of multi-label learning for accurate and efficient feature screening is met, label information can be comprehensively utilized, the anti-interference capability is enhanced, and the efficiency is considered.

The invention discloses a data privacy protection method based on multi-party security computing and a block chain, and the method comprises the following steps: 1, creating a multi-party security computing task request, and distributing the multi-party security computing task request to each participant node; 2, performing Kronecker perturbation coding on local original data, and performing Hash processing through an SHA-256 Hash algorithm to generate a perturbation Hash value; 3, recording access behaviors, and generating an access behavior track sequence; step 4, calculating a Soft-DTW distance between the access behavior track sequence and the historical access track sequence; 5, constructing an adjacent matrix and a Laplacian matrix, and carrying out eigenvalue decomposition; step 6, calculating the Euclidean distance between the access embedding vector and the legal access embedding vector; and 7, generating a result abstract and writing the result abstract into a task recording module. According to the method, the Kronecker perturbation coding and the Hash algorithm are combined, so that the data privacy protection, the calculation transparency and the compliance verification capability are improved.

The invention provides a flexible interconnection device site selection method and system based on distributed resource clustering, and the method comprises the steps: constructing a photovoltaic output feature map Laplacian matrix and a load feature map Laplacian matrix according to the photovoltaic output and load data of a planning region, decomposing the constructed matrixes through DGRNMF, and obtaining a distributed resource cluster; solving a photovoltaic clustering center curve and a load clustering center curve according to a decomposition result through a k-means clustering algorithm, and dividing the photovoltaic clustering center curve and the load clustering center curve into an urban region and a rural region; constructing constraint conditions of the power supply grid segmentation model to obtain the power supply grid segmentation model; and solving the power supply grid segmentation model to obtain an optimal power grid partitioning scheme, and determining a candidate installation position of the flexible interconnection device according to the optimal power grid partitioning scheme. According to the invention, the precision and reliability of the flexible interconnection device site selection planning decision are improved.

The invention provides a non-stationary industrial process monitoring method and system. The method comprises an offline training stage: calculating a time Laplacian matrix and a space Laplacian matrix based on a historical data matrix; constructing an objective function of the stationary subspace analysis method, and adding a time constraint term of a time Laplacian matrix and a space constraint term of a space Laplacian matrix into the objective function; solving the objective function to obtain a stable projection matrix; calculating a stationary component and a monitoring index of each sample in the data matrix X in sequence; determining a control limit by using a kernel density estimation method; an online monitoring stage: based on the real-time operation data x, calculating a stationary component of the real-time operation data x and a corresponding real-time monitoring index according to the stationary projection matrix, and if the real-time monitoring index is greater than a control limit, judging that the operation of the non-stationary process has a fault; the monitoring accuracy can be improved.

The invention discloses a distributed photovoltaic cluster elastic grid division method based on space-time dynamic association. The method comprises the following steps: constructing a space-time high-dimensional feature tensor for describing the running state of a power distribution network; a comprehensive space-time incidence matrix is obtained; a Laplacian matrix reflecting source-load interaction characteristics is constructed; performing characteristic decomposition and dimension reduction mapping on the Laplacian matrix to obtain initial sub-grids of a plurality of distributed photovoltaic clusters; verifying whether the frequency change rate caused by the maximum power shortage in the isolated island operation mode is out of limit or not; calculating the critical clearing time of the system through time domain simulation, and taking the critical clearing time as a quantitative index of transient stability; and calculating a migration priority index of the boundary node, carrying out iterative updating until the whole network meets the multi-dimensional constraint, and outputting a final division scheme. The method can adapt to the spatial-temporal fluctuation characteristics of high-proportion distributed photovoltaic, effectively guarantees the frequency safety and transient stability of the power distribution network, and achieves the dynamic balance of physical topology, spatial-temporal association and safe operation.

The invention discloses an impedance mismatch design method for a gas turbine combustion chamber multi-flame-tube cross-flame system, which comprises the following steps: forming a combustion chamber by 12 flame tubes in a gas turbine through cross-flame tubes on the basis of a graph theory, modeling as a cyclic graph, and constructing an adjacent matrix and a Laplacian matrix; constructing a system matrix based on the angular frequency of the flame tube and the coupling strength of the cross flame tube; calculating a characteristic value and a characteristic vector of a system matrix, determining an oscillation frequency and a vibration mode, and identifying a degenerate mode; by modifying the coupling strength of at least one cross flame tube and recalculating the eigenvalue and eigenvector of the system matrix, the degenerate frequency is split, and the symmetry is broken; and according to the frequency and the mode after splitting, parameters of the combustion chamber are adjusted to optimize the combustion stability. Modeling of a complex system is simplified based on a graph theory method, the calculation cost is reduced, the symmetry of the system is broken by modifying the coupling strength of the cross flame tubes, the degenerate frequency is split, and combustion instability is relieved.

The invention relates to the technical field of remote sensing image processing, in particular to a multi-temporal remote sensing image change detection method and system based on a space-time diagram neural network, and the method comprises the steps: carrying out the geometric registration, radiation correction and superpixel segmentation of an image, and outputting a segmented superpixel region set; taking the superpixel of each time phase as a node, constructing a space-time diagram structure fusing the spatial adjacency relation and the multi-step time association, and generating a normalized space-time Laplacian matrix; spatial structure features and multi-temporal evolution features are extracted through a double-branch graph neural network, spatial branch and time branch output are fused based on a gating mechanism, and node embedding is generated; constructing positive and negative sample pairs based on node embedding, and optimizing a feature space by comparing a loss function; and calculating the Euclidean distance of node embedding at adjacent moments, and outputting a change detection result in combination with a dynamic threshold. The method effectively reduces the false alarm rate and omission rate, greatly improves the reasoning speed, and is suitable for the scenes of urban expansion monitoring, dynamic disaster evaluation and the like.

The invention relates to a production line beat tuning method and system, and the method comprises the steps: firstly building a station efficiency evaluation model of a hypergraph structure through collecting the production data of each station in real time, and capturing a complex interaction relation between stations through hyperedges; then, a weighted adjacent matrix is constructed based on the model, and the station cooperation efficiency and the information flow intensity are analyzed through a Laplacian matrix; secondly, identifying bottleneck stations influencing the whole production rhythm by calculating the spectrum energy and the spectrum energy difference of each station; and when a bottleneck station is detected, the system starts a beat adjustment and optimization mechanism, and dynamically adjusts beat configuration with the purpose of minimizing the sum of the production entropy, the waiting entropy and the collaborative entropy, thereby finally realizing the maximization of the overall efficiency of the production line and the optimal allocation of resources. The method has high dynamic adaptability, can effectively deal with the problem of rhythm imbalance in a complex production environment, and has a wide application prospect.

The invention relates to the field of robot path planning, in particular to a multi-robot path planning method based on group control, which comprises the following steps: on the basis of a visibility graph, generating a communication undirected weighted graph, introducing a Laplacian matrix, and reflecting the connectivity of the graph according to a second small feature value of the Laplacian matrix; the arrival time of the robots is controlled by adjusting the side weights, space-time conflicts are avoided, the robot path planning sequence is determined according to the contribution value of the second small feature value, and the overall performance of multi-robot cooperation is optimized; according to the method, data in robot path planning are integrated through the Transform model, the comprehensiveness of environmental understanding is improved, spatial constraints of the path are optimized through a message passing mechanism of the graph neural network, the robot dynamically adapts to environmental changes, the path is generated through the generative adversarial network, diversity and high efficiency are achieved, local optimum is avoided, and the method is suitable for being popularized and applied. And dynamic obstacle avoidance and real-time path optimization are realized through cooperation of multiple models.

The invention relates to the technical field of unmanned aerial vehicle autonomous control, and discloses an unmanned aerial vehicle cluster formation control method and system based on a multi-level intelligent algorithm, and the method comprises the steps: obtaining the external environment monitoring data and motion state data of an unmanned aerial vehicle cluster, constructing a multi-target comprehensive cost function, and carrying out the optimization of a cluster formation initial track, obtaining an optimal trajectory set meeting dynamic constraints; executing collision risk assessment on the sampling points on the optimal trajectory set, and outputting a trajectory safety judgment result; performing mathematical characterization and eigenvalue modulation on the formation topological structure through a Laplacian matrix, and generating target formation parameters adapted to environmental constraints; and finally, converting the optimal track set and the target formation parameters into a bottom layer flight control instruction, driving the unmanned aerial vehicle to execute a flight action, and collecting real-time motion state data at the same time. Therefore, the problems that in the prior art, trajectory optimization convergence is slow, collision detection real-time performance is poor, formation dynamic adaptation is insufficient, and system module collaboration is weak are solved.

The invention discloses a low-rank multi-modal remote sensing image clustering method and device on a superpixel manifold and a storage medium, and relates to the technical field of multi-modal remote sensing image clustering. The method comprises the following steps: splicing a multi-modal remote sensing image along a channel direction, segmenting the spliced image into a plurality of sub-regions through superpixel segmentation, and solving a mean value for each sub-region to obtain a multi-modal superpixel; embedding the Laplacian matrix of the multi-modal superpixels into manifold regularization about the superpixel clustering matrix, and capturing a local manifold structure of the multi-modal superpixels; under the constraint of manifold regularization, constructing a low-rank reconstruction model of a product of a single-mode clustering matrix and a unified clustering matrix; initializing and alternately optimizing the single-mode clustering matrix and the unified clustering matrix by using fuzzy clustering; and analyzing the super-pixel clustering result, and mapping the super-pixel clustering result into a clustering result of the original image. According to the invention, the accuracy and efficiency of remote sensing image clustering are improved.

The application discloses a kind of double-channel label guide multi-label feature selection method and system, belong to feature engineering technique.Method mainly includes: obtaining the feature matrix and positive label matrix of multi-label data set, by performing logical negation to positive label matrix, generate mirror negative label matrix, and construct graph Laplacian matrix based on feature matrix;Based on the data after pre-processing, a multi-label model is constructed, and the objective function of the multi-label model includes at least positive label regression loss term, negative label regression loss term, label alignment constraint term, graph regularization term and sparse constraint term;The constraint is processed by relaxation, and the optimization function is constructed by combining the Lagrange multiplier method, and then the objective function is iteratively solved according to the KKT condition, and after iterative convergence, the feature importance is evaluated based on the projection matrix used to associate features and positive labels;The application meets the demand of multi-label learning for accurate and efficient feature selection, can fully utilize label information, enhance anti-interference ability and consider efficiency.

The invention provides a power supply unit intelligent division method and system, and belongs to the technical field of power grid operation management. The method comprises the following steps: acquiring related data in a target division region, constructing a cross-administrative region penalty model, a cross-road network penalty model and a dynamic load balancing target function based on the related data, further constructing a constraint perception Laplacian matrix, and decomposing to obtain a node deep feature matrix; obtaining an initial division scheme based on the node deep feature matrix; constructing a multi-objective optimization model based on a preset user number balancing objective, a geographic constraint violation penalty, a connectivity guarantee objective and a dynamic load balancing objective function; and performing boundary adjustment on the initial division scheme based on the multi-objective optimization model to obtain a final division scheme. According to the intelligent division method provided by the invention, the division efficiency and quality are remarkably improved on the premise that constraint conditions such as fitting with administrative regions and not crossing a road network as far as possible are met, and technical support is provided for fine management of the power distribution network.

The invention discloses a graph signal blind source separation method in a white noise environment, and belongs to the technical field of signal processing. Aiming at the problem that the traditional blind source separation performance is reduced due to additive white noise, the invention provides a joint optimization scheme combining blind compression nonlinear noise suppression and image smoothing regularization. Firstly, a blind compression function is applied to a noisy observation image signal for preprocessing; secondly, carrying out mean value removal and whitening processing on the compressed signal; then, constructing a graph Laplacian matrix as a smoothing operator, establishing a joint diagonalization objective function fusing graph autocorrelation, a FastICA item and a graph smoothing regular item, and adopting a Givens rotation algorithm to iteratively optimize and solve a separation matrix; and finally, reconstructing a source signal through multiplication of the separation matrix and the whitening data. According to the method, noise interference is effectively suppressed through blind compression, signal structure consistency is maintained by using graph smoothing prior, and separation robustness, precision and convergence stability in a strong noise environment are remarkably improved.

The invention belongs to the technical field of machine learning and data mining, and particularly relates to a large-scale high-dimensional image data set-oriented compressible subspace clustering method, which comprises the following steps of: firstly, designing a dictionary representation learning model based on a partitioning mechanism to select part of samples to construct a small-scale dictionary to replace the whole original data; a bipartite graph construction method is ingeniously introduced by utilizing the thought of joint clustering, the problem that the bipartite graph cannot be directly constructed due to the fact that a coefficient matrix is not a square matrix is solved, and the relevance between a dictionary sample and a new input data sample can be fully considered. Under the Laplacian matrix rank constraint of the combination graph, the method can directly learn to obtain an optimal structured bipartite graph, can directly obtain a final clustering result, and does not need any post-processing process. In addition, an efficient optimization algorithm based on alternate iteration is further designed in combination with an augmented Lagrangian multiplier method to solve the model.

The invention discloses a smart pipeline abnormal data sequence prediction method based on a Transform network. The method comprises the following steps: collecting pipeline data, constructing a time sequence, and dividing a monitoring window; a pipe network topology Laplacian matrix is constructed based on the pipe section connection relation, and a normal reference set is extracted; separating a source signal by using topology constraint SOBI and establishing a source space security domain model; constructing a time sequence fusion Transform prediction network to output abnormal sequence prediction; calculating an abnormal weight according to the deviation between the monitoring window and the normal reference set and between the monitoring window and the source space security domain, and updating source signal features; based on the abnormal weight, the source signal feature and the prediction residual error, constructing an improved SOBI optimization target to jointly optimize a blind source separation parameter and a prediction network parameter; and constructing an event triggering criterion according to the source space security domain default index and the prediction residual error to adaptively update the abnormal weight and the joint model, and outputting a pipeline abnormality prediction result. According to the invention, the reliability of pipeline abnormity prediction is improved.

The present application relates to the technical field of agricultural robot control, and particularly provides an anti-interference convergence method for an agricultural robot cluster based on interaction link dynamics. The method comprises obtaining state information and speed information of each interaction link, determining an interaction topology structure and a connection weight matrix between the interaction links; constructing an edge Laplacian matrix according to the connection weight matrix of the interaction links, and determining a control gain based on all non-zero eigenvalues of the edge Laplacian matrix; determining a control amount of each interaction link according to the state information and speed information of each interaction link, the state information and speed information of a neighbor interaction link, the control gain, and a random interference term; and controlling the state and speed of the corresponding interaction link according to the control amount of each interaction link. The method effectively suppresses the influence of external interference on the link state, guarantees the convergence of the link state, and improves the robustness and cooperative control ability in a complex airspace environment.

The invention discloses a power distribution network dynamic partitioning method based on a composite electrical distance. The method comprises the steps that the composite electrical distance between each pair of buses in a power distribution network is calculated; constructing an affinity matrix according to the composite electrical distance, and constructing an undirected weighted graph based on the affinity matrix; calculating a Laplacian matrix for the undirected weighted graph to obtain a preliminary partition of the power distribution network; a mixed integer linear programming method is adopted to optimize the preliminary partitioning result, it is ensured that each sub-region meets set constraints, and optimized partitions are output; topological change of a power distribution network and output fluctuation of a distributed power supply are monitored in real time, and when the change exceeds a preset threshold value, a dynamic adjustment mechanism is triggered, and a partition adjustment strategy is executed. According to the method, on the basis of comprehensively describing the electrical similarity of the bus, the engineering feasibility and the running state adaptability of the partitioning result can be realized by combining constraint optimization and a dynamic adjustment mechanism.

The invention provides an abnormal transaction identification method and device, equipment and a storage medium, and relates to the technical field of financial data identification. The method comprises the following steps: acquiring entities in transaction data to be identified and a transaction relationship between the entities; and generating a weighted financial association network graph according to the entities and the transaction relationship between the entities. And calculating a Laplacian matrix of the weighted financial association network diagram, and carrying out eigendecomposition on the Laplacian matrix to obtain eigenvectors corresponding to the first k minimum eigenvalues. Generating a feature matrix based on the k feature vectors; and clustering nodes in the feature matrix to obtain k node communities, and determining a risk score of each node community. And if it is determined that the target node community with the risk score exceeding the preset risk threshold exists, determining that the target node community is an abnormal object. The method is used for achieving the effect of improving the recognition capability of abnormal transaction structures of complex and irregular network communities.

The application discloses a recommendation method based on a guided diffusion model enhanced graph encoder, and specifically comprises the following steps: S1, initializing a user-item interaction adjacency matrix and performing normalization processing to generate a Laplacian matrix; S2, adaptively calculating a mask probability for each click behavior, generating a partially visible mask bipartite graph as a mask graph encoder input in a model training process, and enhancing a supervision signal from a structural level; S3, generating user and item representations by using the mask graph encoder, and performing denoising processing by adding directed noise and introducing a collaborative signal guided diffusion model, and enhancing the supervision signal from a semantic level; and S4, generating user and item representations by using the graph encoder, optimizing semantic information of the representations, and finally predicting possible clicked items for each user. By combining the structural and semantic double enhancement strategies, the application significantly improves the accuracy and recall rate of the recommendation system, and improves the robustness of the system to noise information.

A multi-agent system encirclement control method triggered by intermittent dynamic events under hybrid attacks comprises the following steps: S1, constructing a directed communication topology and a Laplacian matrix of a multi-agent system according to an actual task; s2, constructing a multi-agent system model including uncertain disturbance, nonlinear dynamics and false data injection attacks; s3, designing a sliding-mode observer and a double-layer attack detection mechanism for denial of service attack and false data injection attack; s4, designing a fixed time encircling controller of the intermittent dynamic event triggered multi-agent system; s5, the controller designed in the step S4 is used for achieving encircling control within fixed time, and the upper bound of convergence time is obtained; and S6, continuously operating until the encircling control of the multi-agent system is completed. The invention aims to solve the problem of safe, efficient and rapid encircling control when a multi-agent system is subjected to DoS attack, FDI attack and external disturbance at the same time.

The invention discloses a credible drug target prediction method based on function space regularization, and relates to the technical field of drug research and development, and the method comprises the steps: collecting interaction data and feature data of drug molecules and EGFR mutants; constructing a drug similarity graph and a target similarity graph; defining a bilinear prediction function and a loss function; a function space regularization term based on double-graph convolution is introduced, and the regularization term forms a regularization term of the smoothness of the constraint prediction function in the joint similarity space through Laplacian matrix construction of the joint drug graph and the target graph; combining the loss function with the regularization item to construct an objective function, and performing optimization solution through a gradient descent method; and finally, predicting the interaction between the new drug and the EGFR mutant by using the optimized model. By introducing a mechanism based on double-graph convolution function space regularization, combined similarity information of a pharmaceutical chemical structure and a target sequence can be fused in drug target prediction of non-small cell lung cancer EGFR mutants.

The invention discloses a deep multi-label classification method based on a graph structure, which can perform efficient compression and feature reconstruction on input data through a deep network layer under the condition that noise exists in original data, break through redundant information interference of the original data, accurately capture essential feature representation contained in the data, and perform classification on the basis of the compressed feature data. And constructing a graph Laplacian matrix capable of completely retaining key node associated information, and finally performing label prediction classification on the obtained graph Laplacian matrix. The method can be applied to a classification experiment of dimension reduction data, the lowest rank expression of a graph structure is finally obtained, interference of data noise on the classification process is better resisted, and therefore the classification performance is improved.

This invention discloses a robust fusion method for large models based on semantically aligned fuzzy clustering. The method includes: first, obtaining a sequence of probability distribution vectors from the outputs of multiple heterogeneous large-scale pre-trained models for the samples to be processed; then, introducing non-negative reliability weights to weight this sequence to obtain an aggregated probability matrix, and constructing a graph Laplacian matrix accordingly; second, approximating the aggregated probability matrix into a fuzzy membership matrix and a semantic prototype matrix, constructing a cost objective function by combining the graph Laplacian matrix, and iteratively updating each matrix and weight until convergence; finally, solving for the maximum value of the converged fuzzy membership matrix to obtain the sample prediction category result. This invention can effectively suppress the influence of inferior models in unsupervised environments, significantly improving the accuracy and robustness of fusion prediction.

The application provides a distributed offshore wind turbine control method based on a consistency algorithm, comprising: taking the distributed offshore wind turbine as a distributed intelligent agent, and obtaining total active power shortage of a plurality of intelligent agents; establishing an initial first adjacency matrix according to a communication relationship between adjacent intelligent agents, and adjusting the first adjacency matrix to obtain a second adjacency matrix; modifying an initial first Laplacian matrix according to the second adjacency matrix to obtain a second Laplacian matrix; establishing a target function taking maximization of a wind energy utilization coefficient of each distributed offshore wind turbine as an objective, and obtaining an output power increment allocated to each distributed offshore wind turbine according to the second Laplacian matrix, so that the distributed offshore wind turbine outputs corresponding output power; and the application can improve power quality and improve power control accuracy of the distributed offshore wind turbine.