61 results about "Random matrix" patented technology

The invention discloses a body fat prediction system and method based on a multi-band impedance signal, and relates to the technical field of big data analys.The method comprises the steps that impedance data of different segments under different frequency bands, the body fat amount, the lean body weight and the human body weight are collected; integrity verification is carried out on the impedance data, missing values are filled with mean values, abnormal values are removed and corrected, and meanwhile timestamp alignment is carried out; extracting features to construct an original impedance vector, and mapping the original impedance vector to a high-dimensional space through a random matrix to generate a random mapping function; constructing a time sequence regression model, taking the embedding dimension as an input layer, taking the body fat amount and the lean body mass as an output layer, and using mean square error calibration and back propagation updating; and through a mean square error weighting evaluation model, outputting a performance standard result. The system comprises a data acquisition module, a data preprocessing module, a random distribution embedding module, a time sequence model training module and a display module. The method can adapt to impedance characteristics of different crowds, is suitable for portable terminal or smart home body measurement, and can be used at high frequency in daily life.

The invention discloses a power grid harmonic wave and inter-harmonic wave detection method and device based on distributed multi-frequency measurement data, equipment and a medium. A sine wave method comprises the following steps: detecting harmonic wave and inter-harmonic wave components existing at each sine wave measurement point based on a random matrix theory; determining the frequency and amplitude of each harmonic wave and inter-harmonic wave of each sine wave measurement point by using a rotation invariant technology; clustering the frequency of each harmonic wave and inter-harmonic wave on each measurement point by using a density-based noise application space clustering method to obtain the target frequency of each frequency cluster; and determining harmonics and inter-harmonics of the power grid according to the target frequency and the amplitude. Each harmonic and inter-harmonic component existing on a measurement point is detected through a random matrix theory and a rotation invariant technology, accurate frequencies and amplitudes of the harmonic and the inter-harmonic are obtained, and more accurate frequencies of the harmonic and the inter-harmonic of a power grid are obtained by fusing data of a plurality of distributed measurement points through a clustering method. Therefore, the detection capability of harmonic waves and inter-harmonic waves of a power grid is improved.

The application discloses a bit data transmission scheduling method based on 6G space-ground integrated quantum network, a storage medium and equipment, and comprises the following steps: constructing a 6G space-ground integrated quantum communication network; searching for nodes meeting network scheduling task requirements in the quantum communication network, constructing a candidate node set, and screening out a determined best emission node; sending a quantum state to a receiving end through a quantum key distribution protocol at the best emission node, calculating a quantum bit error rate, selecting a candidate relay node from the candidate node set with the minimum quantum bit error rate as the target, and forming a quantum transmission channel; and the quantum transmission channel prepares a quantum key by using an entangled state, generates a random matrix density, generates a ciphertext through the random matrix density, and transmits the ciphertext to the receiving end. The application can reasonably schedule network resources, ensure data security through quantum communication, and the like.

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 large language model parameter fine-tuning method and device based on manifold constraint multi-track adaptation and a medium. In view of the problems of unstable training and limited expression capacity of an existing low-rank adaptation technology, a plurality of parallel low-rank tracks are constructed, and a double random matrix is introduced to constrain the information flow between the tracks, so that the stability of gradient propagation is ensured. Meanwhile, the expression capacity of the adaptation module is enhanced under a limited parameter budget by dynamically fusing the outputs of the tracks through a dynamic gating mechanism. The method can realize stable, efficient and high-performance model fine-tuning by fine-tuning a small number of parameters, and reasoning has no additional overhead, and is particularly suitable for application scenarios with limited resources and high stability requirements, and has wide practical value.

The invention belongs to the field of magnetic resonance sounding signal noise filtering, and particularly relates to a magnetic resonance sounding signal de-noising method based on backtracking generalized OMP, and the method comprises the steps: employing a magnetic resonance sounding water detector to carry out the sampling of a sparse signal at a frequency lower than the Nyquist frequency, and obtaining a noisy MRS signal; converting the noisy MRS signal into a two-dimensional matrix signal; dictionary learning is carried out on the two-dimensional matrix signals through a KSVD algorithm, and an updated dictionary is obtained after iteration; calculating a recovery matrix and measurement data by using an independently distributed Gaussian random matrix; performing iterative operation on the measurement data by using a backtracking generalized orthogonal matching pursuit algorithm, selecting an atomic sequence of maximum projection of the measurement data and a recovery matrix module, and obtaining a sparse coefficient matrix of the measurement data under the recovery matrix; and calculating a recovery signal according to the dictionary and the sparse coefficient matrix of the measurement data under the recovery matrix.

The invention provides a port image analysis method, system and device based on a random matrix, and a medium, and relates to the technical field of port inspection image processing, and the method comprises the steps: obtaining a port inspection image; creating two random matrixes; performing interpolation operation on the two random matrixes by using a bilinear interpolation function; constructing a chaotic system, and adding disturbance parameters to enhance nonlinearity in combination with an interpolation function and a random matrix to obtain a Jacobian matrix; calculating a maximum index value; if the maximum index value is greater than 0, determining that the system is in a chaotic state; drawing a curve of indexes changing along with time, and analyzing sensitive parameters of the system; row and column replacement and bitwise XOR processing are carried out on pixels by using the stable segment of the chaos sequence; performing decryption processing; and outputting the comparison diagram. According to the method, the dynamic key is generated in combination with the random matrix and the sensitive parameters by adopting an encryption mode based on the chaotic sequence, and the image pixels are subjected to multi-stage encryption processing by utilizing the transient state of the chaotic system, so that the security of an encryption algorithm is improved.

The invention discloses an atmospheric laser communication turbulence compensation method based on a novel Kolmogorov-Zakharov model, and belongs to the technical field of atmospheric laser communication. The objective of the invention is to solve the technical problem that the transmission quality and stability of atmosphere laser communication are affected due to reduction of compensation precision in a strong turbulence scene in the prior art. The method comprises the following steps: acquiring phase distribution and light intensity fluctuation of a laser signal in real time to obtain an atmospheric refractive index structure constant, mapping a model parameter and an atmospheric laser communication physical quantity, and updating a model based on the atmospheric refractive index structure constant; based on the model and a Helmholtz equation of atmospheric laser transmission, an atmospheric laser channel KZ turbulence random matrix model is obtained and solved, and a prediction result is obtained; and then turbulence dynamic compensation is carried out based on a prediction result. The method is mainly used for atmosphere laser communication turbulence compensation.

This invention provides a method, medium, and system for acoustic early warning of thermal runaway in lithium-ion batteries based on feature reconstruction, belonging to the field of lithium-ion battery technology. The invention constructs a positive sample set by collecting safety valve opening sounds through multi-condition thermal runaway experiments, and expands the samples using data augmentation. Multi-resolution Mel spectra are extracted from the audio signals. These Mel spectra are then input into a complex-domain phase-aware separation model for joint estimation of complex-domain amplitude masking and phase residuals. Physical prior corrections are applied to the reconstruction results using a sound source separation algorithm based on wave equation time-frequency inverse scattering and a low-rank sparse time-frequency matrix decomposition algorithm based on random matrix theory. The three corrected signals are weighted and fused to obtain a corrected Mel spectra, which are finally input into a temporal convolutional network to classify and identify the safety valve opening sounds and output a thermal runaway early warning signal. This invention solves the technical problem of insufficient accuracy in thermal runaway early warning caused by acoustic feature reconstruction distortion in complex noise environments.

The embodiment of the application relates to an algorithm and a device of a Bayesian classifier of a non-uniform background, which are applied to an underwater active sonar system, the algorithm comprising: obtaining underwater data to be measured and auxiliary data through the active sonar system; the number K of the auxiliary data is greater than 0; modeling classification of the unknown covariance matrix structure into a binary hypothesis testing problem; hypotheses of the binary hypothesis testing problem comprise H i Wherein i=0, 1, H0 is a case that the unknown covariance matrix is a complex conjugate symmetric matrix; H1 is a case that the unknown covariance matrix is a real symmetric matrix; a Bayesian model is set, the Bayesian model comprising a complex inverse Wishart random matrix and a real inverse Wishart random matrix; a classifier for distinguishing the two hypotheses is obtained by using a minimum error probability criterion.

The invention discloses a photovoltaic scene generation method and system based on random matrix theory enhancement, and the method comprises the steps: constructing a training sample set based on historical data, and constructing a generative adversarial network framework comprising a generator and a discriminator; on the basis of a random matrix theory, time sequence data of a generated photovoltaic power time sequence scene and a real photovoltaic power time sequence scene are constructed into an augmented matrix, spectral distribution of the augmented matrix is calculated, the spectral distribution is converted into a correlation measurement index, and the correlation measurement index serves as an additional loss item to be introduced into a loss function of a discriminator; adopting an alternate training strategy to perform joint training on the generator and the discriminator into which the correlation loss item is introduced, and optimizing a generative adversarial network framework; and inputting the random noise vector to the generator, and generating a photovoltaic power time sequence scene. Spectral distribution analysis is carried out on a generated scene and a real scene through a random matrix theory, a correlation index is used as a discriminator loss item, and the problems of insufficient interpretability, lack of correlation evaluation and the like of the generative adversarial network in photovoltaic scene generation are solved.

The invention provides a method, a device, equipment and a medium for eliminating errors in lattice public key encryption. The method comprises a key generation stage, an encryption stage and an auxiliary encryption stage. The secret key generation stage comprises the steps of generating a random matrix, generating a first secret vector and a first error vector according to uniform distribution sampling, and calculating a target vector according to the first secret vector and the first error vector so as to obtain a public key and a private key; the encryption stage comprises the steps of generating a second secret vector, a second error vector and a secret ring element according to uniform distribution sampling, and generating a ciphertext based on the second secret vector, the second error vector, the secret ring element and a message; the auxiliary encryption stage comprises executing a PSI-CA protocol based on the first secret vector, the first error vector, the second secret vector and the second error vector. According to the method, the uncertainty of error terms in an LPR scheme can be effectively reduced, and then an encryption scheme with smaller design parameters is supported.

Certain aspects of the present disclosure provide techniques and apparatus for efficiently adapting a machine learning model from a base task to a downstream task based on a frozen matrix. An example method generally includes receiving an input for processing through a layer of a neural network. An output of a layer of the neural network is generated based on a first product based on a first trainable scaling vector, a first frozen matrix, a second trainable scaling vector, a second frozen matrix, and the received input.

The application relates to a wireless channel modeling method, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring a small-scale channel matrix sample; generating a first distribution matrix; the first distribution matrix is an independent and identically distributed complex normal random matrix; generating a first constant matrix, a second constant matrix and a second distribution matrix according to the small-scale channel matrix sample; the first constant matrix is used for reflecting the energy coupling ability of a line-of-sight component, the second constant matrix is used for reflecting the energy coupling ability of a scattering component, and each element of the second distribution matrix is a power of a mutually independent generalized gamma distribution random variable; and acquiring a target channel matrix according to the first constant matrix, the second constant matrix, the first distribution matrix, the second distribution matrix and the small-scale channel matrix sample. The method can realize equivalence of GBSM and CBSM statistical characteristics which are universally applicable to any application scenario.

The application provides an interview method and device based on an AI primary and backup expert model, and relates to the technical field of artificial intelligence. In the application, when determining the expert set corresponding to each interview feature information of a candidate object, a primary expert is determined based on affinity score sorting and a preset cumulative confidence threshold. When the number of primary experts is less than a preset minimum number of experts, a backup expert is determined through double random matrix transformation of the Sinkhorn algorithm. In this way, the cumulative confidence threshold drives dynamic expert activation, adaptive computing resource allocation is achieved, the computing efficiency is improved under the condition of ensuring the accuracy of the interview result output by the model, unnecessary computing overhead is avoided, and when the number of activated experts is insufficient, the Sinkhorn algorithm is used for double random matrix transformation to select experts with the most global balance as backup experts, so that the load distribution is balanced and experts are not idle.

The invention discloses an integrated circuit interconnection line stray capacitance extraction method based on a random matrix, and belongs to the technical field of integrated circuits, and the method comprises the steps: constructing a conductor system geometric model; constructing an internal and external sampling point sequence; constructing a random matrix to compress a matrix to be solved; constructing final to-be-solved left and right end matrixes in batches; solving a compressed matrix equation; a result obtained by solving is brought back to the previous matrix construction step to construct a new matrix to be solved; repeatedly executing the step of constructing and solving until the obtained solution meets a convergence condition; and generating a final capacitance matrix. According to the method, the coupling parasitic capacitance of the interconnection line can be efficiently and accurately extracted.

This invention discloses a solution to the joint path and destination planning problem based on a distributed generalized Nash equilibrium algorithm. First, the joint path and destination planning problem is modeled, transforming it into a non-cooperative game model. This model includes the objective function of each electric vehicle, global coupling constraints, and local constraints. Second, pseudo-gradients are used to transform the game model into a VI problem, introducing edge-based consistency constraints and a heterogeneous step size mechanism. Based on fixed-point iteration and proximal gradient operator theory, a distributed solution algorithm under complete information is proposed. Then, a global estimate of the plans of other users is introduced, proposing a distributed solution algorithm under partial information. This invention avoids the construction of double random matrices through edge-based consistency constraints, and can maintain solution accuracy while meeting low computational requirements when more users participate in the game model.

The application belongs to the technical field of image compressive sensing reconstruction, and discloses an image compressive sensing reconstruction method and system based on optimization algorithm expansion. In the sampling stage, a convolution sampling method is used instead of a traditional random matrix sampling. In the reconstruction stage, a generalized iterative threshold shrinkage algorithm is expanded into a deep network, and a jump information connection structure is designed in a reconstruction submodule R. Residual modules are used to connect the feature information before and after the modules, so that the inherent information loss in the deep expansion network is avoided. Furthermore, a double-scale denoising module is designed at the back end of the reconstruction submodule R, and different scale features are combined to denoise the image. The application not only realizes the application of the algorithm expansion method in the image compressive sensing, but also improves the reconstruction effect by using the jump connection structure and the double-scale denoising module. The application has higher accuracy and better robustness.

The invention discloses an unmanned aerial vehicle group target tracking method and system based on a random matrix filter, and the method comprises the steps: setting a first threshold related to a distance, a pitch angle and an azimuth angle, comparing all plots with a reference plot through the first threshold, rejecting split plots, traversing the plots which are not rejected, and carrying out the target tracking of the unmanned aerial vehicle group. Setting a second threshold representing the distance of the group target relative to the radar, comparing all the plots with the reference plot by using the second threshold, obtaining a neighborhood set, identifying and marking each plot in the neighborhood set, and obtaining a group plot; judging whether the group trace point is associated with an existing group; initializing a random matrix filter; predicting a group centroid state and a covariance matrix thereof, and a group size and a degree of freedom thereof; detecting the change of the group scale, adaptively adjusting the innovation covariance matrix of the random matrix filter, updating the group centroid state estimation and the covariance matrix thereof, and updating the group size estimation and the degree of freedom thereof; the method has the advantage of high tracking precision.

The invention provides a method for predicting the residual life of a bearing. The method mainly comprises the steps of feature data collection, bearing degradation point recognition, feature matrix cutting and residual life prediction. Firstly, temperature and vibration signals of a bearing are collected through sensor equipment, and a feature matrix is constructed; secondly, a random matrix model is introduced to carry out bearing degradation point identification on the feature matrix, and the bearing degradation starting moment is marked; then, according to the bearing degradation point cutting feature matrix, effective degradation feature data of the bearing are reserved, and useless healthy operation data are abandoned; and finally, training the BILSTM neural network by using the bearing degradation characteristic matrix to obtain a bearing residual life prediction model. According to the method, the degradation point of the bearing can be automatically identified, the residual life of the bearing can be predicted without feature data preprocessing, and the method has relatively high prediction precision and relatively strong robustness.

The application discloses a fast calculation method and device for scattered electromagnetic field based on random matrix approximation, and belongs to the field of fast modeling of scattered field. The method comprises the following steps: precalculating an incident field vector and a reciprocal incident field vector; generating a Gaussian random matrix matched with the column number of a to-be-determined scattered response matrix; performing linear superposition on the incident field vector to obtain a synthetic incident field vector, and constructing a random projection matrix according to the scattered field vector of an equivalent current vector at a receiving point; performing linear superposition on the reciprocal incident field vector to obtain a reciprocal synthetic incident field vector, and constructing a projection coefficient matrix according to the response of a reciprocal equivalent current vector at a transmitting source; and calculating a complete scattered response matrix based on the product of an orthogonal basis matrix and the projection coefficient matrix. In the application, a large-scale geophysical electromagnetic complex scene can be adapted, higher calculation precision and stable convergence are achieved, and the reconstruction of the complete scattered response matrix can be completed at a lower relative error level.

The application discloses a kind of integrated circuit interconnection line parasitic capacitance extraction methods based on random matrix, belong to integrated circuit technical field, comprising: constructing conductor system geometric model;Constructing inside and outside sampling point sequence;Random matrix is constructed to the matrix to be solved urgently compressed;Batch construction final left and right end matrix to be solved;Solving compressed matrix equation;The result obtained by solving is fed back into the previous matrix construction step to construct new to-be-solved matrix;Repeatedly execute construction solving step until the solution obtained meets the convergence condition;Generate final capacitance matrix.The application can realize the efficient and accurate extraction of interconnection line coupling parasitic capacitance.

This invention discloses a method, apparatus, terminal, and medium for distributed bi-level optimization on time-varying directed graphs, comprising: acquiring a time-varying directed graph sequence; solving a bi-level optimization problem for a multi-agent system based on the time-varying directed graph sequence; updating the decision variables of each agent using a row random matrix and calculating the local gradient of the decision variables of each agent using a penalty function; updating the gradient tracking variables of each agent using a column random matrix to eliminate gradient estimation bias in the time-varying directed graph sequence; and outputting the bi-level optimization results for the multi-agent system. This invention solves the problem of bi-level optimization failure under dynamic topology, eliminates the risk of numerical instability, overcomes the second-order computational bottleneck, and reduces resource overhead.

The present disclosure provides a secure three-party multiplication method and system for privacy computing, involving the technical field of privacy computing. The method includes that an auxiliary compute node generates three groups of random matrix pairs randomly and transmits the random matrix pairs to three parties, and the three parties compute respective sum matrixes based on a sum of respective random matrixes and private matrixes, respectively Â, Ĉ and {circumflex over (B)}. A second party generates a matrix set according to a sum matrix, a first party obtains Ta based on the matrix set and its own secret matrix, the second party obtains Tb based on its own random secret matrix and Ta, and the third party generates its own random secret matrix based on Tb and the matrix set, and obtains a product matrix according to three random secret matrixes. The present disclosure can improve reliability of result accuracy.

The invention discloses a method for performing extended target tracking by using a random matrix and partial normal distribution, which comprises the following steps of: establishing an evolution model and a partial normal distribution measurement model, and performing prior prediction on a prior target motion state, an extended form and a measurement deflection variable; carrying out posteriori estimation on a target motion state, an expansion form and a measurement skew variable, and obtaining a skew constraint vector estimation value; re-modeling based on an IT-IMM framework: obtaining an optimal solution of a prior probability density function of a target motion state, an extension form and a measurement skew variable through a weighted KLA algorithm; the posterior probability density of the corresponding mode is obtained through a variational Bayesian algorithm, and mode probability updating is carried out; and through weighted KLA approximation, final estimation of the motion state and the expansion form of the target is obtained. According to the method, accurate tracking of the target motion state and accurate estimation of the expansion form can be realized.

The application relates to the technical field of medical artificial intelligence, and discloses a brain disease prediction method based on a double-encoder and a diffusion model, which comprises the following steps: pre-processing a functional magnetic resonance image and constructing a brain function network; through a diffusion model, semantic-preserving data enhancement is realized, wherein a double random matrix and a cosine scheduling strategy are adopted in a noise adding process, and a GraphTransformer neural network containing global topological features is utilized in a noise removing process; a double encoder is adopted to extract spatial features of the brain network and time dynamic features of BOLD signals; a triple contrast learning mechanism is designed to optimize cross-dimension feature interaction; and finally, the method is migrated to a downstream classification task to realize disease prediction. Through diffusion enhancement, small sample overfitting is alleviated, and diagnosis reliability is improved; time-space features are fused to assist multi-dimensional pathological analysis; the model cross-site adaptability is enhanced, and multi-center heterogeneous data collaborative analysis is supported. The application is suitable for the auxiliary diagnosis of brain diseases such as autism.

The application relates to a random matrix-based dry-type air-core reactor turn-to-turn short-circuit fault identification method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring a real-time pulse response signal of a dry-type air-core reactor; the real-time pulse response signal comprises current electrical characteristic change information of the dry-type air-core reactor; a random matrix is constructed based on the real-time pulse response signal; eigenvalue analysis is performed on the random matrix to obtain an operating state characteristic parameter of the dry-type air-core reactor; the operating state characteristic parameter represents the operating state of the dry-type air-core reactor; and the operating state characteristic parameter of the dry-type air-core reactor is input into a pre-trained fault identification model to obtain a fault identification result of the dry-type air-core reactor. The method can accurately and efficiently detect a turn-to-turn short-circuit fault of the dry-type air-core reactor.

This invention discloses an accurate zero-knowledge proof method for linear relations with hidden products, which transforms linear relations with hidden products on incomplete splitting rings into linear relations on the same surface; transforms the accompanying additional conditions into component-wise product relations on the same surface; and transforms the secret evidence on the same surface into element x on the same surface using an inverse NTT transformation. i The process involves selecting a uniformly random matrix and several vectors; generating a first-stage commitment and sampling vectors from a discrete Gaussian distribution to determine the public vector; selecting first challenge information; generating new commitments, blinding elements, and redundant terms using the first challenge information; randomly sampling second challenge information c from the challenge space C; determining the vector used for blinding using c; verifying whether the vector satisfies preset conditions and checking whether a series of equations hold true. If all equations hold true, the proof is accepted; otherwise, it is rejected.

The invention provides an interview method and device based on an AI main and standby expert model, and relates to the technical field of artificial intelligence. When an expert set corresponding to each piece of interview feature information of a candidate object is determined, a main expert is determined based on the sequence of affinity scores and a preset cumulative confidence threshold; and when the number of the main experts is smaller than the preset minimum number of the experts, determining the standby experts through double random matrix conversion of the Sinkhorn algorithm. In this way, dynamic expert activation is driven by accumulating the confidence threshold value, self-adaptive computing resource allocation is achieved, and under the condition that the accuracy of an interview result output by the model is guaranteed, the computing efficiency is improved, and unnecessary computing overhead is avoided; and when the number of the activated experts is insufficient, the expert with the best global balance is selected as the standby expert through double random matrix conversion of the Sinkhorn algorithm, so that the load distribution balance is ensured, and the experts are prevented from being idle.

This invention discloses a robust target tracking method using a dual-entropy fusion multi-random matrix maneuvering extension. The method employs an interactive multi-model fusion architecture. In the fusion step, a cost function is established using the target motion state estimate of the previous time-instance sub-model as the independent variable, based on the correlation entropy criterion. Maximizing this cost function yields the fused target state estimate. Using relative entropy as the criterion, the sum of information gains of each sub-model regarding the target motion state covariance and target morphological distribution parameters is minimized. Then, a correlation entropy cost function is established using the filtered target motion state estimate of the current time-instance sub-model as the independent variable. Maximizing this cost function achieves target motion state estimation fusion. Simultaneously, the total information gain of each sub-model regarding the target motion state covariance and target morphological distribution parameters is minimized again to obtain a weighted estimate, ultimately yielding a robust joint estimate of the target motion state and morphology at the current time.