28 results about "Diagonal matrix" patented technology

This invention discloses a GNSS interference source localization method based on ADS-B data. The method involves acquiring ADS-B data, saving data points affected by GNSS interference, and, if no NIC is available, constructing ADS-B data with a NIC. The message data is then sorted by ICAO and timestamp. A diagonal matrix W is calculated based on the ICAO, timestamp, NIC, a table showing the correspondence between NIC and received interference power, and the hyperparameter value τ. A residual vector R is calculated based on the message data and the optimization vector, and the optimized vector is then output. This invention analyzes large-scale aviation ADS-B data, sorts and classifies the original message data to determine whether flights are subject to GNSS interference, and abstracts the interference source localization problem into a least-squares problem. An optimization function is constructed, and an iterative optimization algorithm is used to calculate the possible GNSS interference source locations and transmission power, thus more efficiently locating GNSS interference sources.

The application discloses a kind of based on Koopman and CEF model electric carbon collaborative scheduling method, system, medium and equipment, belong to electric power system optimization field, the method is: according to CEF model obtains the original expression of carbon intensity constraint;According to branch outflow power distribution matrix and diagonal matrix, reconstruct the original expression, obtain the nonlinear mapping relationship expression between node carbon emission intensity and node power injection, and according to low rank kernel approximation improved Koopman operator and node power injection dataset are processed, obtain linear mapping relationship expression;According to linear mapping relationship expression, construct electric carbon collaborative scheduling model, and based on electric power system current operation data, solve scheduling model, obtain scheduling optimization strategy;According to the strategy, regulate electric power system operation, complete electric carbon collaborative scheduling.Therefore, by implementing the present application, the problem of insufficient timeliness of electric power system in combination with carbon emission flow for electric carbon collaborative scheduling in the prior art can be solved.

The application discloses an optical matrix vector multiplier based on a unitary matrix-diagonal matrix decomposition. The multiplier comprises a Mach-Zehnder interferometer array for realizing a synthetic unitary matrix Omega and a modulation unit for realizing a corresponding diagonal matrix Sigma, wherein a target weight matrix M satisfies a decomposition form of M=ΩΣ. The method fuses two unitary matrices required for mapping in a traditional singular value decomposition into a single synthetic unitary matrix Omega, so that a unitary transformation can be completed by only one interferometer array of optical hardware, and a complete matrix multiplication is realized in cooperation with the modulation unit. Through algorithm-hardware collaborative design, the number of basic optical interference units required is significantly reduced, the hardware complexity, manufacturing cost and system power consumption of the optical neural network are greatly reduced while the computing performance is maintained, and an effective solution is provided for efficient optical computing acceleration of artificial intelligence.

The application provides a multi-signal spectrum separation method, device, equipment and medium, which is used in the technical field of radio monitoring and can solve the technical problem of low spectrum separation accuracy caused by the limitation of channel number and poor adaptability of directional antenna array. The method comprises the following steps: calculating a real-time conjugate transpose signal and a time-delay conjugate transpose signal under multiple different time delays according to a real-time composite signal, and determining a corresponding covariance matrix and multiple pseudo-covariance matrices; determining a matrix product according to the conjugate transpose matrix of an initial demixing matrix, obtaining multiple initial product matrices, and judging whether the initial product matrices are all approximate diagonal matrices; iteratively processing the initial demixing matrix to obtain a target demixing matrix; determining a corresponding array manifold matrix based on the target demixing matrix, and determining the signal amplitude of each signal at the frequency point corresponding to the array manifold matrix and the mixed amplitude and phase vector corresponding to each frequency point; thus, the spectrum separation accuracy is improved.

The application discloses a spectrum graph convolution network heterogeneous graph representation learning method based on path collaborative graph enhancement, relates to the technical field of graph neural networks, and comprises the following steps: obtaining to-be-processed heterogeneous graph data, projecting node features of different types into a unified latent feature space by using type-aware linear transformation; calculating structure prior weights; constructing a path collaborative graph, learning path interaction weights by using a graph attention network; constructing a collaborative polynomial spectrum filter; performing spectrum graph convolution based on path collaborative graph enhancement, introducing a learnable positive definite diagonal matrix into the collaborative polynomial spectrum filter, performing weighting and feature transformation on the filtered node features, and obtaining final node representation. The application can learn the importance of meta-paths in a fine-grained manner and capture semantic interaction between meta-paths, solves the problems that existing spectrum heterogeneous graph convolution cannot distinguish the importance of fine-grained paths and lacks semantic collaboration, effectively improves the classification performance of heterogeneous graph nodes, and has a good application prospect.

This invention proposes a 3D Gaussian splatter style transfer method based on learning orthogonal color bases. The method includes: obtaining a local mixing coefficient matrix, a global color base matrix, and a singular value diagonal matrix through a pre-trained 3D Gaussian splatter model, while simultaneously acquiring the original content image; obtaining a final global color base matrix through coarse-stage global color base optimization; refining the local mixing coefficients through a fine-stage local mixing coefficient matrix; updating the pre-trained 3D Gaussian splatter model using the final global color base matrix and the final local mixing coefficient matrix; and generating style transfer results using the stylized 3D Gaussian splatter model. This invention transforms high-dimensional color space optimization into low-parameter subspace optimization through SVD decomposition, significantly improving the optimization speed.

This invention discloses a method, system, and storage medium for weighted and rate optimization of wireless networks assisted by intelligent reflectors, relating to the field of wireless communication, and addressing the problem of the lack of such optimization methods in existing technologies. The method includes: Step 1: Under transmit power constraints, establishing a weighted and rate optimization model for wireless networks assisted by intelligent reflectors based on the joint optimization of the transmit beamforming matrix at the base station and the diagonal matrix at the intelligent reflector; Step 2: Initializing a population of cultured Remora fish using the optimization model as the fitness function; Step 3: Allowing the cultured Remora fish to roam freely; Step 4: Allowing the cultured Remora fish to feed; Step 5: If the number of iterations is less than the preset maximum number of iterations, setting ε = ε + 1 and returning to Step 3; otherwise, terminating the iteration, outputting the global optimal position of the cultured Remora fish, returning the optimal fitness value, and finally obtaining the weighted and rate optimization method for wireless networks assisted by intelligent reflectors.

This invention discloses an image illumination correction method based on a width learning system, belonging to the field of image illumination correction technology. The method includes: first, acquiring the image to be corrected; second, inputting the preprocessed image into a DenseNet201 feature extraction network based on transfer learning, and further reducing the dimensionality of the output to obtain image illumination features; further inputting the image illumination features into a BLS pre-trained using the HHO-HOA algorithm to obtain predicted RGB values; finally, restoring the image based on the RGB values ​​through diagonal matrix transformation to obtain the corrected image. This invention can eliminate the influence of ambient lighting on photographic imaging, restore the true color of objects, and thus provide accurate color information for related engineering tasks in computer vision. The model possesses excellent prediction accuracy, good convergence and stability, improving the efficiency and accuracy of illumination correction work.

The present invention is applicable to the technical field of wireless communications, and particularly relates to a microwave network connection method and system, and a related device. The method of the present invention comprises: on the basis of whether port attributes are equipotential, dividing a first sparse block diagonal matrix into four sub-matrices, so as to obtain a second sparse block diagonal matrix comprising the four sub-matrices; acquiring a transfer matrix used for representing power distribution of the equipotential ports of microwave networks to be connected; and, on the basis of a preset rule, performing computation on the transfer matrix and the second sparse block diagonal matrix, so as to obtain a scattering parameter matrix; and, by means of the scattering parameter matrix, connecting the plurality of microwave networks to be connected. Compared with the prior art, the present invention involves dividing a sparse block diagonal matrix into four smaller sub-matrices, which have lower computation time complexity, thus significantly reducing computation time and improving computational efficiency during computation, thereby improving the efficiency of microwave network connection.

The application discloses a hybrid precoding method based on model-driven deep learning, and belongs to the wireless communication field; specifically, first, the channel states of all users are collected, the virtual channels of the users are respectively preprocessed to obtain preprocessed features, and dimension transformation is performed; then, a bipartite graph for learning an analog precoding matrix is modeled; the preprocessed features are input into a GNN, the edge hidden representation of the bipartite graph is updated layer by layer, the feature-to-action mapping of the edge is learned, and the analog precoding matrix is obtained; further, an equivalent baseband channel is calculated; then, sub-bipartite graphs for learning a power allocation diagonal matrix and a beamforming matrix are respectively modeled; the features of the edges are respectively input into the GNN, the feature-to-action mapping is learned by layer-by-layer updating the edge hidden representation of the bipartite graph, the power allocation diagonal matrix and the beamforming matrix are obtained, and an equivalent baseband precoding matrix is obtained by combining; finally, an actual baseband precoding matrix is restored; the application has a faster inference speed and a high learning efficiency.

This disclosure relates to a computer-implemented method for converting a multi-bit message into a 3D transmission vector for transmission over a physical resource, where the bit is an integer greater than 1. The method includes: generating two bit subsets by splitting the multiple bits; and generating a block-expanded matrix that is not a block diagonal matrix using the first bit subset from the two bit subsets. The method further includes: generating a symbol vector carrying information using the second bit subset from the two bit subsets; and generating a transmission vector using the expanded matrix and the symbol vector carrying information. A communication device is provided for wirelessly transmitting a multi-bit message over a physical resource. This communication device is used to convert the message into a 3D transmission vector by performing the above-described method.

The application discloses a kind of amplitude and phase high-precision estimation method and device suitable for fast dynamic change signal, the method of the application includes the discrete signal sequence of continuous signal sampling, constructs Hankel matrix and decomposes left singular matrix U, diagonal matrix D, right singular matrix V;The estimated value of the number of frequency components p is estimated by D, combined with extracting the subspace of right singular matrix V and using least square method to obtain diagonal matrix Q;The damping coefficient of frequency component is calculated by Q and constructs system matrix A;The dynamic phasor coefficient vector S is calculated by A and discrete signal sequence using least square method;The amplitude and phase of continuous signal are calculated by S and damping coefficient.The application aims at solving the problem that the main lobe overlap of dense harmonic / interharmonic with small frequency interval in power system causes spectral leakage interference, extremely serious even can submerge part of main spectrum line, and then lead to the accuracy of spectral leakage affecting the amplitude and phase estimation of dynamic change signal.

This application provides a method, apparatus, computer device, and storage medium for human behavior recognition, belonging to the field of computer vision technology. The method includes: acquiring a sequence of behaviors to be recognized; acquiring a template behavior sequence; performing alignment calculations based on a preset diagonal matrix and an initial column vector to be recognized to obtain a target aligned behavior sequence; calculating the similarity between the template behavior sequence and the target aligned sequence to obtain a sequence similarity; and determining the behavior recognition result of the target object based on the sequence similarity. This application first converts the sequence of behaviors to be recognized into a target aligned behavior sequence aligned with the template behavior sequence using a diagonal matrix, and then determines the behavior recognition result of the target object based on the sequence similarity between the target aligned sequence and the template behavior sequence. Because this application performs an alignment operation before sequence matching, it can improve the accuracy of behavior recognition.

The present application relates to the technical field of program control, and discloses an advanced process control method for disaster area unmanned aerial vehicle light source cooperative lighting. In view of the light intensity superposition distortion problem in the light field overlapping area of multiple machines, the unmanned aerial vehicle pose change sequence and the light source optical axis pointing angle change sequence under strong wind load are obtained, and a multivariate non-minimum phase transfer function matrix is constructed with the pose change as input and the ground grid node light intensity change as output. A diagonal matrix decoupling compensator is introduced to decouple the above-mentioned predicted output vector into independent single variable channels, and the compensation angular velocity control sequence of the light source holder is solved by optimizing the target of approximating the preset light intensity value with the predicted value. The sequence is reversely superimposed with the bottom attitude angular velocity to drive the holder to execute compensation. The scheme cuts off the cross-coupling path of the pose disturbance and the adjacent light beams, eliminates the light intensity flicker and overexposure in the overlapping area, and maintains the static consistency of the light distribution on the rescue operation surface.

This invention discloses a fault-tolerant control method, program, device, and storage medium for underwater robots based on physical information residual networks and dynamic control allocation, belonging to the technical field of underwater robot control methods. First, a generalized dynamic model is constructed, incorporating a diagonal matrix of thruster health and an external lumped disturbance term. Then, a fully connected neural network is used to estimate the total system disturbance. Next, a composite control law is designed based on a non-singular terminal sliding surface, and the virtual total control force required to resist faults is calculated. Finally, the control allocation is transformed into a constrained quadratic programming optimization problem. A thruster health-aware weighting mechanism and relaxation variables are introduced into the optimization objective function, and the thrust of each thruster is redistributed through an optimization algorithm. When a fault occurs in the underwater robot's propulsion system, a graded degradation strategy is determined based on the relaxation variables in the optimization objective function. This ensures the robot's normal navigation even when some thrusters are damaged.

This invention discloses an intelligent monitoring method and system for the working status of a building climbing formwork based on the Internet of Things (IoT), belonging to the field of building structural health monitoring technology. The method includes extracting corresponding rows and columns from an initial correlation matrix to obtain a reconstructed covariance matrix, extracting the main diagonal elements of the inverse matrix, constructing a diagonal matrix, calculating the Hadamard product of the reconstructed covariance matrix and the diagonal matrix, and extracting the main diagonal elements of the matrix to form an effective independence vector, thereby monitoring the working status of the building climbing formwork. This invention improves the extraction capability of low-frequency vibration modes and enhances the sensitivity to minute structural changes through empirical mode decomposition (EMD) technology and weighted fusion of dominant mode signals. Furthermore, it enhances the collaborative working capability between different sensor signals through phase difference calculation based on instantaneous phase analysis and energy weighted correction.

The application belongs to the technical field of communication and perception integration, and particularly relates to a sensing beam forming method for reconfigurable intelligent reflecting surface and laminated intelligent metasurface. Θ The specific process of the method is as follows: taking the deployment position vector of the RIS, the reflection phase shift diagonal matrix of the RIS and the phase shift diagonal matrix of the first layer of the SIM as joint optimization variables, constructing a joint optimization model with multi-user and rate maximization as the target while meeting the target sensing accuracy constraint; taking the RIS deployment area constraint, the RIS unit phase shift constraint and the SIM layer super atom phase shift constraint as conditions; establishing the inverse relationship between the two-dimensional CRB and the target direction beam pattern gain, and equivalently converting the original CRB constraint into the beam pattern gain constraint to obtain a new optimization problem; for the new optimization problem, the joint optimization variables are decomposed into three optimization problems and solved to obtain the locally optimal RIS deployment position, SIM phase shift matrix and RIS phase shift matrix.

Precoding and reception methods, transmitter device and receiver device. The invention proposes a method for precoding, by a transmitter device (2), sequence(s) of symbol(s) spatially distributed on layers of Mlayer symbols using a plurality of complex precoding matrices of dimensions Px, where P denotes a number of antenna ports of the transmitter device, each precoding matrix being constructed from a combination of a complex matrix W of dimensions Px, a complex diagonal matrix D of dimensions x introducing a variety of cyclic delays and a complex rotation matrix U of dimensions x, said precoding method being characterized in that, for >1 and P>min(1,), the P columns of the complex matrix W are linear combinations of distinct vectors from a basis of P orthogonal vectors, p=0,..,P-1. Figure for the abstract: Fig. 1.

The application provides a knee osteoarthritis typing model training method and device and a typing method, and the method comprises the following steps: constructing a data set by taking information of a patient in a phenotype characteristic as a data sample; taking two-dimensional coordinates obtained by dimension reduction as initial clustering centers, and constructing a minimum spanning tree according to connection lines between the clustering centers; generating an adjacency matrix and a Laplacian matrix according to the connection relationship; calculating the probability of the data sample being distributed to the clustering center according to the distance between the two-dimensional coordinates and the clustering center, and generating a soft distribution probability matrix and a diagonal matrix; solving a pre-set target function based on the above matrices to obtain an updated matrix set; solving the target function based on the updated matrix set, and when the target function value converges, taking the minimum spanning tree topology as a knee osteoarthritis typing model. The application can make the knee osteoarthritis typing result meet the requirements of clinical interpretability, continuous phenotype characteristics and stable and reliable results at the same time.

The application discloses a kind of heat transfer equipment heat conduction's three diagonal heterogeneous many-core parallel solution method and system, and it is related to computer processing technical field.The steps include: collecting the heat transfer parameter of heat transfer equipment, and establishing heat transfer model;Heat transfer parameter is input to heat transfer model, and heat conduction equation is constructed;About heat conduction three diagonal linear equation set is obtained by difference and transformation;The data dependence of three diagonal matrix equation data is eliminated by the method of parallel elimination;Extract part to form small three diagonal matrix equation;Reduced three diagonal matrix equation is solved using chase and catch method;Each process is output three diagonal matrix equation remaining all solutions with the way of task parallel back substitution;And according to the solution result, draw temperature variation curve, obtain the temperature variation of heat conduction process.Solve the problem that three diagonal matrix equation scale is larger in the simulation process of heat transfer equipment heat conduction phenomenon simulation, resulting in the analysis process of heat conduction process is longer in time, and the result is not accurate enough.

The present disclosure provides a data multi-party joint processing method, system, server and terminal device. The method comprises: generating an encryption key; sending the encryption key to a data owner, the data owner encrypting its own original data using the encryption key to obtain encrypted data; receiving the encrypted data sent by the data owner and performing aggregation processing on the encrypted data to obtain a relationship matrix; obtaining a first intermediate matrix and a first perturbation matrix according to the relationship matrix, the data owner obtaining a second perturbation matrix according to the first perturbation matrix and its own original data; receiving the second perturbation matrix sent by the data owner and obtaining a target diagonal matrix according to the second perturbation matrix, and sending the target diagonal matrix to each data owner, so that the data owner obtains a distribution verification value of its own original data according to the target diagonal matrix and the first intermediate matrix. Thus, cross-institutional sharing and fusion of data are realized, and joint modeling of a mixed linear model is completed under the premise of ensuring data privacy and security.

The application belongs to the technical field of new power system analysis, and discloses a multi-following network-structure network type converter system stability analysis method and an analysis system eq and Q ml ; joint diagonalization processing is performed on the matrices Q eq and Q ml , corresponding diagonal matrices Lambda eq and Lambda ml are obtained, n sets of characteristic parameters (lambda i , mu i ) are extracted, and it is judged whether (lambda i , mu i X RPC0 ) are all located above a critical stability curve of the following network type converter; if yes, the system is stable, otherwise, the system is unstable. Based on the above method, the stability criterion of the multi-following network-structure network type converter system can be simplified, the influence of the structure network type converter on the stability of the following network type equipment under different control modes can be determined, and thus the reactive power control strategy of the structure network type converter can be optimized, and the support capability of the structure network type converter on the power grid can be enhanced.

This invention discloses a fault-tolerant control method, program, device, and storage medium for underwater robots based on physical information residual networks and dynamic control allocation, belonging to the technical field of underwater robot control methods. First, a generalized dynamic model is constructed, incorporating a diagonal matrix of thruster health and an external lumped disturbance term. Then, a fully connected neural network is used to estimate the total system disturbance. Next, a composite control law is designed based on a non-singular terminal sliding surface, and the virtual total control force required to resist faults is calculated. Finally, the control allocation is transformed into a constrained quadratic programming optimization problem. A thruster health-aware weighting mechanism and relaxation variables are introduced into the optimization objective function, and the thrust of each thruster is redistributed through an optimization algorithm. When a fault occurs in the underwater robot's propulsion system, a graded degradation strategy is determined based on the relaxation variables in the optimization objective function. This ensures the robot's normal navigation even when some thrusters are damaged.

This invention belongs to the technical field of microbial detection and analysis instruments and systems. Specifically, it is a dynamic process control method for the fermentation process of *Prunus pedatus* probiotics. It employs an in-tank temperature change model to predict the in-tank temperature field distribution during a predicted period, obtaining a predicted in-tank temperature field. The predicted in-tank temperature field is then subjected to quantitative deviation analysis to determine whether adjustments to the fermenter temperature control parameters are necessary. If dynamic parameter adjustment is triggered, a dynamic adjustment phase is initiated. The predicted in-tank temperature field at the end of the lag period is taken as the current state, and the standard temperature field as the target state. A multivariable PID controller is constructed, using an incremental PID algorithm combined with a diagonal matrix decoupling method to iteratively calculate the adjustment values ​​of each control parameter and adjust the fermenter temperature control parameters accordingly. This invention achieves precise, stable, and efficient control of the fermentation temperature, improving the quality consistency and production efficiency of *Prunus pedatus* probiotic fermentation products.

The present application relates to the technical field of bionic tendon, especially to a PID-based bionic tendon control method, a test method and a test platform, and discloses a PID-based bionic tendon control method, which comprises the following steps: S1: adopting a Lagrange dynamics modeling method to establish a dynamics model for a planar two-joint collaborative robot, S2: determining an error function according to an uncertainty boundary and a diagonal matrix, and obtaining a control model according to the error function, S3: obtaining angle data of a robot joint and air pressure data of a pneumatic muscle, and inputting the data into the control model to obtain control parameters. The present application determines an error function through an uncertainty boundary and a diagonal matrix, and obtains a control model according to the error function. The control model obtained in this way can obtain stable control parameters, can satisfy the strategy generation of single-arm or double-arm control, and has good effectiveness and robustness.