16 results about "Spectral radius" patented technology

The invention discloses a video monitoring abnormal behavior real-time detection method based on a graph neural network, and the method comprises the following steps: collecting a video frame sequence, extracting a detection frame, a key point and an optical flow feature, generating a node feature matrix, and constructing a dynamic graph structure; establishing a dynamic graph neural network model based on EvolveGCN, and updating a convolution weight by using a gating circulation unit; calculating event intensity and change rate according to the motion abrupt change signal, generating a time delay parameter and adjusting a weight modeling step length; performing low-rank decomposition and spectral radius projection on the convolution weight matrix, and adjusting a spectral constraint threshold according to an abnormal score; inputting a weight matrix to generate graph branch and hypergraph branch embedded representation; exchanging topology correction information based on a mutual generation mechanism and updating model parameters; and inputting the dynamic graph structure and the node feature matrix in real-time reasoning, calculating an abnormal score and outputting a detection result. According to the invention, adaptive evolution and high-precision anomaly detection of dynamic graph modeling are realized.

The invention discloses a multi-model game method for multi-scene intelligent safety management. The method comprises the following steps: acquiring multi-domain original data and generating information structure features; the heterogeneous indexes are uniformly mapped into information loss components based on the features, and stable routing of the game template is carried out in combination with sequential statistics and hysteretic decision; performing game payment-oriented decision sensitive subspace compression on the unified semantic state output after routing; a dynamic risk-to-go (dynamic risk surplus) time difference method is adopted to generate step risk signals meeting total amount conservation, and the step risk signals are injected into a compression space; and through online estimation of the spectral radius of the Jacobian matrix of the optimal response operator, an equilibrium vulnerability index is generated, and stability management and closed-loop adjustment of an equilibrium point are realized. According to the invention, the decision adaptability, stability and robustness of the intelligent system in a dynamic confrontation environment can be improved.

The invention discloses a hydraulic engineering operation and management cooperative control system based on digital twinning, and the system comprises the following steps: a data assimilation module which collects observation station, remote sensing and meteorological data and assimilates the data to obtain an assimilation estimator; the dictionary and lifting module is used for generating a lifting variable sequence according to hydrodynamic force prior; the structure identification and uncertainty module is used for identifying a lifting linear model and recursively predicting uncertainty under the constraints of conservation, monotonicity, dissipation and spectral radius; the terminal security module constructs a positive invariant terminal security set according to a support function; the opportunity constraint module is used for setting a total out-of-limit probability budget and generating an opportunity constraint substitution set; the Koopman rolling optimization module is used for solving the control sequence and executing the first control quantity; and the online updating and correcting and deformation and verification module is used for executing residual error triggering low-rank correction, projection return and support surface parameter online deformation. According to the invention, risk-controllable and stable and efficient collaborative scheduling is realized.

The invention relates to an adaptive multi-scale decomposition echo state network chaotic time sequence prediction method suitable for meteorological prediction, and belongs to the field of chaotic time sequence prediction. Comprising the following steps: performing multi-scale decomposition on an input sunspot number sequence by using an HP filter; each sub-sequence obtained through decomposition is distributed to an ESN sub-network; inputting a weight in the ESN sub-network by adopting a Xavier weight initialization method, introducing a composite activation function in reservoir nonlinear updating, and performing state splicing by adopting a state splicing strategy; a sparse random matrix is constructed, a unit matrix scale factor is introduced, and the spectral radius is strictly controlled; each ESN sub-network independently trains an output weight, ridge regression is adopted for solving, integration and reconstruction are carried out after corresponding component prediction is completed, and if a target sequence is a sunspot activity index, the model finally outputs a sunspot number prediction value in a future time step; according to the method, high-precision, stability and robustness prediction of the chaotic time sequence can be realized.

The application discloses a stability evaluation method suitable for a multiple-input multiple-output system and belongs to the technical field of stability control, which obtains the amplitude margin and the phase margin of the multiple-input multiple-output system by calculating the gain solution of the system at different frequencies when the system is critically stable. The application comprises the following steps: calculating the maximum range of the amplitude margin corresponding frequency according to the spectral radius of the transfer function of the multiple-input multiple-output system; calculating whether the amplitude gain solution exists at different frequencies within the frequency range, and if the amplitude gain solution exists, calculating the corresponding amplitude margin; drawing all the amplitude gains and solving the intersection points with the gain curve to obtain the amplitude margin; calculating the maximum range of the phase margin corresponding frequency according to the spectral radius of the transfer function of the multiple-input multiple-output system; calculating whether the phase gain solution exists at different frequencies within the frequency range, and if the phase gain solution exists, calculating the corresponding phase margin; and drawing all the phase gains and solving the intersection points with the phase gain curve to obtain the phase margin.

The present application relates to the technical field of data processing, in particular to a kind of spectrum radius feature extraction and Hamiltonity determination method for mixed graph structure.It includes the following steps: obtaining mixed graph data and constructing adjacency matrix, Laplacian matrix and unsigned Laplacian matrix;Three spectrum radii are extracted using power iteration acceleration algorithm, and a spectrum radius pyramid is constructed by multi-layer coarsening, forming a multi-scale spectrum feature vector;The feature vector is input into the classification model to output the Hamiltonity probability value;The confidence score is calculated by multiple random dropout, and it is dynamically decided whether to activate the accurate determination engine;Finally, the determination result is output.The present application combines multi-scale spectrum features and statistical features, uses graph neural networks to achieve high-precision probability prediction, and improves the reliability of boundary samples through confidence evaluation and cascade determination architecture, suitable for graph data analysis in the fields of communication networks and bioinformatics.

This invention discloses a method for collaborative delivery path planning using unmanned aerial vehicles (UAVs) and unmanned vehicles (UAVs), specifically relating to the field of unmanned delivery path planning technology. It establishes a unified spatiotemporal evolution model by constructing a cross-domain environmental parameter set and a cross-coupling relationship graph, and introduces temporal phase synchronization constraints to achieve unified modeling of air-to-ground collaborative paths. Furthermore, it constructs a mismatch propagation function to quantify the impact of UAV trajectory deviation on the UAV path, and calculates the path-level mismatch gain coefficient by combining the spectral radius index and the total graph variation. Simultaneously, it calculates the path phase entropy value based on the information entropy method to identify mismatch accumulation segments. Based on this, it introduces path recoverability coding to redundantly reconstruct unstable regions, and iteratively optimizes the objective function through joint optimization, ultimately obtaining a collaborative delivery path scheme that meets energy consumption constraints and possesses temporal convergence characteristics. This invention can effectively suppress mismatch propagation and improve path stability and collaborative delivery reliability.

This invention relates to a distributed accelerated optimization scheduling method, device, and medium based on time-varying directed topology. The method includes: constructing a topology model and an optimization scheduling objective function for a distributed network system, and initializing variables; enabling local information interaction between nodes in the topology model and their neighboring nodes, and performing distributed iterative updates including Nesterov momentum acceleration terms; after completing one iteration, calculating key errors based on the updated agent variables, constructing an error linear system matrix and verifying the spectral radius of the system matrix, determining whether the distributed iterative update converges linearly, and if it converges, outputting the globally optimal scheduling solution of the optimization scheduling objective function; otherwise, performing distributed iterative updates including Nesterov momentum acceleration terms again; and obtaining scheduling instructions for the distributed network system based on the globally optimal scheduling solution. Compared with existing technologies, this invention effectively suppresses iterative oscillations, overcomes the convergence speed bottleneck of traditional algorithms, and achieves fast and robust approximation of the globally optimal solution.

The invention belongs to the cross technical field of artificial intelligence large model reasoning acceleration and rigid differential equation numerical solution, and discloses a matrix-free rigid equation solution adaptation method and system based on large model reasoning acceleration. According to the method, a core calculation module without matrix rigid solution and a large model reasoning acceleration framework are subjected to lightweight adaptation, a spectral radius estimation link driven by quantum annealing is optimized through tensor rearrangement and operator fusion, and a hybrid precision calculation process is optimized in combination with a video memory dynamic scheduling strategy of a large model; resource sharing and collaborative acceleration of rigid equation solving and large model reasoning are achieved; and meanwhile, the rigidity ratio pre-judgment logic is optimized based on the feature extraction capability of the large model, and the real-time performance of adaptive step length control is improved. According to the method, the problem of resource competition when traditional matrix-free rigid solution and large model reasoning coexist is solved, cross-scene calculation efficiency is greatly improved, and the method can be widely applied to the fields of large model and numerical solution fusion such as intelligent equipment real-time simulation, aerospace intelligent deduction and chemical process intelligent regulation and control.

The invention belongs to the technical field of non-linear dynamic numerical calculation and rigid differential equation solution, and discloses a non-matrix rigid self-adaptive non-linear dynamic numerical calculation method and a non-matrix rigid self-adaptive non-linear dynamic numerical calculation system. According to the method, equivalent product calculation of the Jacobian matrix and the vector is realized based on the directional derivative through a matrix-free calculation framework, and a complete Jacobian matrix does not need to be stored; high-efficiency and high-precision estimation of the spectral radius is realized through quantum annealing driven Krofft power iteration; constructing an adaptive mixing precision calculation engine based on a spectral radius result to reduce memory occupation; and realizing stable solution of the extreme rigid system through rigid self-adaptive step length control. According to the method, the solution convergence rate and the calculation efficiency of the extreme rigid nonlinear system can be effectively improved, the memory occupation and the long-term integral error are greatly reduced, and the method can be widely applied to the fields of aerospace, mechanical engineering, intelligent equipment, chemical reaction kinetics and the like.

The invention discloses a Chebyshev KAN dimension raising-based Koopman time sequence prediction system and a Chebyshev KAN dimension raising-based Koopman time sequence prediction method, and the system comprises a time sequence data preprocessing module which synchronously collects multi-source original time sequence signals of each device, and carries out the preprocessing of the multi-source original time sequence signals to obtain an input time sequence; the linearization embedding module is used for acquiring an input time sequence and mapping the input time sequence to a linear space based on a Chebyshev KAN encoder; the dynamic operator generation module is used for generating a dynamic operator according to the obtained time sequence data mapped into the linear space; the segmentation prediction module is used for dividing a prediction time domain into a plurality of continuous time periods, and performing state recursion and prediction through a dynamic operator in each time period; the training module is used for training the system in an end-to-end mode and sharing an encoder and a decoder, and a weighted combination of reconstruction loss, prediction loss and spectral radius constraint loss is adopted in the training process.

The invention provides a construction load overload safety risk real-time monitoring method, which belongs to the technical field of construction safety, and comprises the following steps: arranging multiple types of sensors to form a three-dimensional monitoring network to collect real-time data, and extracting load response time-frequency characteristics through principal component analysis dimensionality reduction and variational mode decomposition; using particle swarm optimization to invert a support stiffness parameter and a pre-stress parameter, constructing a double-layer game model to calculate a load safety margin coefficient, determining a load state sudden change and triggering matrix reconstruction, dynamically adjusting a sensor sampling frequency according to the load safety margin coefficient, and determining a load state sudden change; the inversion convergence is judged through the Jacobian matrix spectral radius, the search step size is adaptively adjusted, the Kriging interpolation and the radial basis function neural network are fused to reconstruct the three-dimensional deformation field, and the technical problem that cooperative monitoring of parameter inversion and load state change of the foundation pit support structure cannot be achieved in the construction load overload state is solved.

The application discloses a Koopman time series prediction system based on Chebyshev KAN dimensionality and a method thereof, and the system comprises: a time series data preprocessing module, which synchronously collects multi-source original time series signals of various devices, and pre-processes the multi-source original time series signals to obtain input time series; a linearization embedding module, which is used for acquiring the input time series, and mapping the input time series to a linear space based on a Chebyshev KAN encoder; a dynamic operator generation module, which is used for generating a dynamic operator according to the acquired time series data mapped to the linear space; a segmented prediction module, which is used for dividing a prediction time domain into a plurality of continuous time periods, and performing state recursion and prediction in each time period through the dynamic operator; and a training module, which trains the system in an end-to-end manner, and shares an encoder and a decoder, wherein a training process adopts a weighted combination of a reconstruction loss, a prediction loss and a spectral radius constraint loss.

The application provides a numerical simulation method and device for a hypersonic flow field and an electronic device, including establishing a three-dimensional space discrete model based on a parabolic Navier-Stokes equation, arranging flow variables at the center of a control volume surface formed by adjacent sections along a main flow direction; introducing a pressure gradient splitting mechanism in the calculation of a flow direction flux, forming a flow direction inviscid Jacobian matrix by directional processing of the pressure gradient on the current section in combination with the relationship between the conservative variables and the grid geometric derivative; constructing a corresponding transverse inviscid Jacobian matrix and obtaining its spectral radius information, calculating a transverse viscous flux based on a central difference format, and constructing an implicit correction equation, splitting the lower triangular, upper triangular and diagonal parts of the coefficient matrix, solving a block diagonal system through forward scanning and backward scanning, obtaining variable increments of the next section and updating the flow parameters, and realizing numerical simulation of the hypersonic flow field.

The invention provides a random simple complex network containment synchronization analysis method based on spectral moments, and relates to the technical field of control and information. The method comprises the following steps: firstly, deriving an analytical expression of first three-order expected spectral moments of an augmented Laplacian matrix, and representing the analytical expression as a function of high-order network local structure properties, so as to establish a quantitative corresponding relation between the expected spectral moments and the network local structure; and then, based on the obtained expected spectral moment set, proposing a piecewise linear reconstruction strategy, performing triangular approximation on feature root distribution of the augmented Laplacian matrix, further estimating a minimum feature root and a maximum feature root, and accordingly realizing pinning synchronization judgment and performance prediction of the random simple complex network. Finally, numerical simulation is carried out by taking a random simple complex network formed by a Lorenz system as an example, and the accuracy and effectiveness of the method are verified.

The invention provides an incompressible fluid elliptic partial differential equation wavelet transform convolutional neural network multi-network (WTCNN-MG) numerical solution method, and belongs to the field of incompressible fluid mechanics numerical calculation, and the method specifically comprises the steps: adaptively optimizing convolution-based smoothing, differential, limiting and continuation operations, and reducing the spectral radius of an iteration matrix; the WTCNN is integrated to carry out additional self-adaptive low-frequency smoothing error correction on the coarse grid hierarchy; and low-frequency and high-frequency characteristics are effectively extracted by using the multi-resolution characteristic of the wavelet, and rapid approximation of a low-frequency smoothing error is realized. The invention designs an integrally differentiable WTCNN-MG method, on the premise of ensuring the same precision as the MG method, the calculation efficiency is improved by nearly 90 times, and an efficient incompressible fluid elliptic partial differential equation numerical solution scheme is provided.