182 results about "Sparse matrix" patented technology

The invention provides a multi-source constraint gridding atmospheric pollutant and greenhouse gas emission estimation method, and relates to the technical field of environmental information processing, and the method comprises the steps: constructing a hierarchical grid structure of an urban core domain and a peripheral background domain based on a unified data basic set, forming a gridding baseline emission list through unified activity data and emission factors, establishing a multi-source scale coordination mechanism, fusing ground monitoring and satellite observation information, and generating an observation data set; according to the method, an SMOKE sparse matrix processing chain is embedded to construct a joint estimation model, parameter inversion and consistency constraint optimization are implemented, boundary flux coordination is carried out between a core domain and a background domain, and finally an emission list which has cross-atmospheric pollutant and greenhouse gas consistency and can be used for refined emission inversion is formed. According to the method, the problem of unstable data fusion caused by scale mismatching between the multi-source observation data and the emission model in the prior art is solved.

The invention relates to the technical field of image analysis, in particular to a metal composite material surface defect intelligent detection method based on machine vision, which comprises the following steps: acquiring an original image and converting the original image into an observation matrix; decomposing the matrix into a low-rank matrix and a sparse matrix through a robust principal component analysis algorithm; utilizing a nuclear norm constraint low-rank matrix to establish a background statistical model, and utilizing a norm constraint sparse matrix; positioning a defect area in the sparse matrix and extracting a gray level co-occurrence matrix feature vector; performing connected domain analysis on the sparse matrix, calculating geometrical morphology parameters and evaluating a stress concentration coefficient; and establishing a self-adaptive decision model to carry out fusion judgment so as to judge the physical damage. According to the method, through low-rank sparse matrix decoupling and multi-dimensional statistical geometric feature fusion analysis, accurate stripping of weak defects and quantitative evaluation of physical damage attributes under a complex background are realized.

The invention relates to the technical field of signal processing, and discloses a multichannel data acquisition system and method based on an FPGA (Field Programmable Gate Array), an acquisition module drives a global counter by using a global synchronous clock, writes acquired data into an annular buffer area with a physical address and a counter value in a linear mapping relationship, and establishes implicit time index storage. When a trigger event occurs, the system broadcasts the locked global trigger timestamp, and the acquisition module backtracks and reads historical data according to the global trigger timestamp, and packages the historical data into a sparse matrix type data packet in combination with the channel validity mask. And after receiving the data packet, the data processing module directly calculates a memory mapping address by using global time information in the data packet, and writes a data load into a corresponding position of the waveform reconstruction buffer area. According to the method, through strict binding of the physical address and the absolute time, independent time label redundancy is eliminated, high-precision synchronization of distributed multiple channels is ensured, and out-of-order automatic in-situ recombination and efficient waveform reconstruction of data of a receiving end are achieved.

The invention discloses a dynamic compression coding method, which is applied to a sparse matrix, and comprises the following steps: S1, obtaining the sparse matrix, and generating a bitmap with the same dimension as the sparse matrix, a value in the bitmap being used for indicating whether an element at a corresponding position in the sparse matrix is zero; s2, a mark sequence is generated according to the bitmap, and each mark corresponds to multiple continuous elements in the sparse matrix and is used for representing whether each element in the multiple elements is zero or not; s3, extracting non-zero elements in the sparse matrix according to the mark sequence; and S4, combining the flag sequence with the non-zero elements to form a data stream after compression coding. The invention also discloses a dynamic compression coding system, a storage medium and a device. According to the dynamic compression coding method, the index storage cost can be reduced, the data access efficiency can be improved, the storage and bandwidth occupation can be reduced by supporting parallel coding and decoding, and finally the calculation throughput can be improved.

The invention relates to a calculation device, a calculation method, and a program. This computing device (10) computes a matrix product of a sparse matrix and a weight matrix in a neural network, the sparse matrix having a structure in which a predetermined number of non-zero elements are included in each block of a predetermined size. The sparse matrix is defined by a local index representing the position of the non-zero element in each block and the value of the non-zero element, and the arithmetic device (10) performs: a process for acquiring, from the weight matrix, an element corresponding to the position of the non-zero element in the sparse matrix by referring to the local index; and multiplication and accumulation operation of non-zero elements of the sparse matrix and corresponding elements of the weight matrix is carried out.

Approaches in accordance with various embodiments provide for the processing of sparse matrices for mathematical and programmatic operations. In particular, various embodiments utilize a tiling approach that divides a sparse matrix into submatrices, many of which will include only zero-value entities. These empty tiles can be ignored, and only the tiles with non-zero entries processed, which reduces resource and time requirements for the processing. An indexing approach can be used for each entity that is a combination of the tile identifier and an offset value, which enables the values to be multiplied correctly against, for example, values of a dense matrix. The tiles can be processed in parallel and the results accumulated to generate a matrix product. The matrix product can then be passed to the next step in a process or operation, such as to a next layer in a deep neural network.

The invention discloses a real-time online superconducting accelerator low-temperature system construction method and system.The method comprises the steps that physical property calculation in a low-temperature system is simplified, and a simplified physical property function of a working medium in the low-temperature system is obtained; the key point temperature is simulated according to the combination of a one-dimensional pipeline model and a 0-dimensional tank sub-model based on an agent model technology, the one-dimensional pipeline model is used for simulating the delay of heat transfer, and the 0-dimensional tank sub-model is used for simulating heat capacity; and responding to the simplified physical property function and the key point temperature, and solving node parameters in the low-temperature system by adopting a large-scale nonlinear sparse matrix equation. According to the invention, the fast simulator which operates independently can be provided, and meanwhile, the fast simulator can be combined with a real low-temperature system to operate to provide analysis and prediction functions for the real low-temperature system.

The invention discloses an MAC unit supporting input activation and weight double-end sparsity, and relates to the technical field of neural network accelerator hardware design. Aiming at the technical defects existing in a sparse matrix acceleration scheme, the scheme is adopted, an activation value and non-zero elements of a weight matrix are stored through a special compressed data register, corresponding zero value distribution is recorded through a matched sparse bitmap register, and a joint sparse bitmap marking a double-non-zero effective calculation position is generated through dual-port joint operation; and the control and address generation subunit scans the joint sparse bitmap and outputs an address and an enable signal, the MAC calculation subunit is driven to read non-zero data to complete multiply-accumulate operation, and finally, an operation result is compressed and coded through an output end sparse encoder, so that efficient transmission of interlayer sparse data streams is realized. According to the method, input activation and weight double-end sparsity can be utilized at the same time, end-to-end sparse data flow is supported, zero value calculation can be dynamically skipped, and the energy efficiency ratio and the calculation throughput of neural network reasoning are improved.

The invention relates to a configurable decompression circuit supporting COO and Bi tmap compression algorithms, and belongs to the field of integrated circuits. According to the hybrid compression method suitable for the circuit, after sparseness analysis is carried out on weight matrixes of all layers of a neural network, COO compression based on a coordinate type sparse matrix or Bitmap compression based on bitmap masks is selected according to the sparseness characteristics of different layers, so that the storage space is optimized; comprising a control module, a first selector, a second selector, a bitmap description memory, a coordinate index memory, a numerical memory, a converter, a COO decoder and a decompression data storage module, decoding of two compression formats of COO and Bi tmap can be supported, and the operation efficiency and flexibility of a neural network processor are improved. The method effectively reduces the weight storage demand of the neural network, enhances the adaptability of the compression algorithm, and is suitable for hardware implementation of an efficient neural network model.

The invention relates to the technical field of power system automation and digital twinning, in particular to a regional power grid static safety risk early warning system and method. Comprising a multi-source heterogeneous data intelligent fusion module, a high-fidelity static security analysis module, a multi-dimensional dynamic risk assessment module, a hierarchical linkage early warning decision module and a full-link result tracing and optimizing module. Uniform access of four kinds of heterogeneous data including SCADA, WAMS, equipment online monitoring and meteorological prediction is supported, a data island of a traditional system is broken through, timestamp alignment is achieved by adopting a dynamic time warping algorithm, spatial deviation is corrected based on GIS Kriging interpolation, the problem of analysis distortion caused by time / space deviation of the traditional system is solved, and the analysis accuracy of the system is improved. A sparse matrix compression technology and an MPI parallel computing framework are adopted, so that the efficiency is improved, the real-time analysis requirement is met, and an N-1 / N-2 fault scene is dynamically generated based on an equipment health index, including N-1 verification of health equipment and N-2 verification of sub-health equipment.

A method includes obtaining a first data matrix and a first index matrix that correspond to a first sparse matrix, and a second data matrix and a second index matrix that correspond to a second sparse matrix; determining, based on each row of elements in the first index matrix and each column of elements included in the second index matrix, target elements on which a dot product operation is to be performed in each row of elements in the first data matrix and each column of elements in the second data matrix; and performing a dot product operation on the target elements to obtain a first result matrix of performing a sparse multiplication operation on the first sparse matrix and the second sparse matrix.

The invention provides a method for detecting fusion of extracellular vesicles by adopting fluorescence resonance energy transfer, and belongs to the technical field of extracellular vesicles. A dynamic fluorescence monitoring system is established, low-power laser excitation is adopted, dual-wavelength fluorescence signals are monitored at the same time, fluorescence signal data are collected, graded gain regulation and control processing is carried out through five threshold values, and a fluorescence signal super-sparse matrix is constructed; establishing a fluorescence signal processing matrix model, converting the super-sparse matrix into a dense matrix, constructing a gain matrix for signal correction, calculating a fluorescence resonance energy transfer efficiency parameter, obtaining a standardized efficiency value through matrix operation, outputting an extracellular vesicle fusion kinetic analysis result, and establishing a quantitative evaluation system. The technical problems of poor signal stability and low detection precision in the process of fluorescence resonance energy transfer detection of extracellular vesicle fusion are solved.

The invention discloses a structured sparse matrix multiplier realized based on FPGA primitives, and belongs to the technical field of FPGA hardware acceleration and deep learning computing architecture. The multiplier is composed of a register area and a multiplication and addition area, and the core design thought is that a circuit is built in a customized mode by directly calling bottom layer physical resources based on FPGA primitives; the register area stores a dense matrix B and supports parallel reading of elements by utilizing the characteristic that LUT in an SLICEM can be configured to be double SRL16E; the multiplication and addition area refers to a partial product generation unit based on the 4-Booth algorithm and an improved GPC (4: 2) compressor structure, and efficient generation and rapid accumulation of partial products are achieved. According to the method, redundant loss caused by high-level HDL logic synthesis is avoided through precise physical resource binding of primitives, fine wiring constraint and structured sparse data characteristic adaptation.

This application proposes a high-precision detection method and device for stationary human targets based on a radar life detector. The method involves performing an inverse Fourier transform on the radar's fast time to convert the frequency axis into the range axis, iteratively updating the low-rank matrix and the sparse matrix, updating the dual variables and modifying the penalty parameters, and obtaining the separation result between the target and the background after convergence. The background information is stored in the low-rank matrix, while the signal containing life characteristic information is stored in the sparse matrix, thereby achieving high-precision detection of stationary human targets.

The invention provides a method for accurately stopping a thermal energy storage power station, and belongs to the technical field of thermal energy storage power stations. Running acoustic monitoring and particle distribution identification to output a particle fluidization degree classification result and a distribution uniformity evaluation value, establishing a particle distribution sparse matrix, and converting the particle distribution sparse matrix into a dense matrix through a multi-stage progressive clustering algorithm to realize accurate quantification of a particle distribution state; executing precise fluidization termination control to completely convert the energy storage particles from a fluidization state into a fixed bed state, performing fixed bed stability verification to ensure that no local overheating phenomenon exists, completing shutdown state recording, and establishing a shutdown historical database; the technical problems of equipment damage and energy loss caused by non-uniform temperature distribution of energy storage particles in the shutdown process of the heat energy storage power station are solved.

The invention discloses a multi-domain coupling transient temperature field rapid calculation method based on a C-N format, and belongs to the field of plate shape detection, and the method comprises the steps: dividing a plate shape roller into a roller body domain, a sensor domain and a coupling boundary shared by the two domains, and building a plate shape roller double-domain coupling heat conduction numerical model and mapping discretization; the method comprises the following steps of: performing time discretization on a whole roll area of a strip-shaped roll by adopting a Crank-Nicolson format; based on a discrete result, respectively calculating sparse matrixes of a roller body domain, a sensor domain and a coupling boundary shared by the two domains, assembling a global sparse linear equation set, and solving to obtain temperature field distribution data of the two domains at the next moment; and mapping the solved temperature field distribution of the roller body domain on the image plane back to the physical plane through inverse transformation of conformal mapping, and forming a complete transient coupling temperature field of the full physical domain with the temperature field of the sensor domain. According to the method, rapid, stable and high-precision calculation of the overall transient temperature field of the strip-shaped roller is realized.

The invention relates to the technical field of image processing, in particular to a silicon carbide tray defect detection method and system based on image recognition, and the method comprises the steps: obtaining a gray image of a silicon carbide tray to construct an observation matrix, and initializing a low-rank matrix, a sparse matrix and a Lagrange multiplier matrix for representing error accumulation; iterative optimization is executed; and repeating the iteration steps until a termination condition is met, and taking the final sparse matrix as a defect detection result. Through the technical scheme of the invention, the problems of false detection and missing detection caused by misjudging the background texture as the defect can be reduced, and the accuracy and robustness of the defect detection result of the silicon carbide tray are improved.

The present disclosure provides a sparse matrix algorithm-based image model optimization method and device, which comprises the following steps: inputting a pre-acquired image training set into a current to-be-optimized image model for training, and obtaining a to-be-processed disturbance parameter corresponding to the current to-be-optimized image model; obtaining a current sparse mask; based on the current sparse mask and the to-be-processed disturbance parameter, calculating a current disturbance parameter; based on the current disturbance parameter, using a gradient descent random gradient descent method to obtain updated model parameters; in the case that the image model corresponding to the updated model parameters has not converged, taking the image model corresponding to the updated model parameters as the current to-be-optimized image model, and returning to the step of inputting the pre-acquired image training set into the current to-be-optimized image model for training until the current to-be-optimized image model converges, and obtaining an optimized image model. In this way, the generalization ability of the optimized image model can be further improved, and the accuracy of image recognition by the optimized image model can be improved.

The application relates to an abnormality detection method based on a hyperspectral image. The main body is based on a space-spectrum combined feature extraction method of filtering and low-rank decomposition to perform abnormality detection on the hyperspectral image. The specific method comprises the following steps: firstly, in the spatial dimension, a reduced dimension image is obtained through a data dimension reduction and eigenvalue weighted fusion method, and then an improved spatial filtering method is used to extract the spatial features of the image to obtain an initial spatial feature image. In the spectral dimension, a background reconstruction image of the approximate background is obtained by using a Tucker decomposition method on the original hyperspectral image, and a background dictionary of the image is obtained by using an improved k-means clustering method, then the background dictionary is input into a low-rank decomposition model to obtain a sparse matrix, and an initial spectral feature image is obtained, finally, the initial spectral feature image is fused with the spatial feature image to realize abnormality detection.

The embodiment of the application provides a sparse identification scheduling method, device, equipment and medium, which has the following beneficial effects: a sparse matrix is obtained; the sparse matrix is divided into multiple identification blocks in units of preset data length, when all elements in the identification block are zero, the identification block is determined as sparse; when there is at least one non-zero element in the identification block, the identification block is determined as valid; the determination result is stored in a memory space independent of the sparse matrix in units of bit to form sparse mask data; in units of a main block, the sparse matrix and the sparse mask data are loaded into a register stack, the register stack includes a head register and a data register, and is used for performing sparse matrix and weight matrix multiplication calculation. The application can efficiently judge the sparse block in a unified granularity under multiple data formats, jointly encode the sparse bitmap and the reconstruction parameter, thereby improving the scheduling efficiency of sparse activation and reducing the storage and bandwidth overhead.

To provide a method and system for improving throughput for loading an input matrix into a matrix multiplication unit by increasing a rate at which values from the input matrix are loaded into an array of cells in the matrix multiplication unit.SOLUTION: The method includes receiving a plurality of non-zero input values from a first input matrix, receiving, for each non-zero input value in the plurality of input values, index metadata indicating which of the M * N cells in the array the non-zero input value is to be loaded into, and transmitting the non-zero input value and the index metadata to the M * N cells, and at a particular cell of the M * N cells in the array: The method includes receiving a particular non-zero input value and corresponding indexing metadata, and determining from the corresponding indexing metadata for the particular non-zero input value whether to store the particular non-zero input value in the cell or shift the particular non-zero input value to another cell.SELECTED DRAWING: Figure 4

The application provides a network model compression method and device, electronic equipment and storage medium, wherein the network model compression method comprises the following steps: obtaining initial weight parameters and initial bias parameters of a network model to be compressed; determining a target inference model based on the initial weight parameters and the initial bias parameters, the target inference model being used to infer the state of different gates of the network model to be compressed at different time points; performing pruning on the target inference model, and determining target weight parameters and target bias parameters of a pruned model; and determining a target network model based on the target weight parameters and the target bias parameters. The method provided by the application not only avoids the problems of increased model parameter quantity and calculation quantity, and increased storage difficulty caused by the introduction of a sparse matrix, but also achieves the purposes of reducing calculation quantity and improving inference speed.

The application provides a communication and calculation cooperative optimization method and device for a heterogeneous intelligent computing network, wherein the method comprises the following steps: generating a task access matrix according to a global sparse matrix; determining a node in which a data block resides and a processing unit in the node according to the task access matrix and node topology information; determining a communication transmission mode of the data block according to a sparse density and column dimension of the data block; and performing multi-flow concurrency to complete a calculation task of a current neural network layer. The application alleviates the parallel expansion bottleneck of a large model training of a computing network, and effectively slows down the growth trend of communication cost.

The application belongs to the technical field of computer architecture design, and discloses a unified accelerator design method and accelerator for diversified graph computing applications, which comprises the following steps: by comparing and analyzing the field-specific execution model of classical graph computing applications, observing the common data flow abstraction of different graph computing applications, establishing a graph computing execution model based on a sparse matrix, and abstracting the common part of the graph computing algorithm; based on the sparse matrix execution model, by analyzing the common hardware features of classical graph computing applications such as graph traversal, graph mining and graph learning, a field general chip architecture for graph computing is constructed, and general and efficient support for complex real-world scenarios is realized; by proposing a graph computing instruction set for a "scalar-vector-graph" heterogeneous core, the programmability of the graph computing chip is improved.

This invention relates to a method and system for reducing single-node memory usage based on distributed reading of ultra-large-scale sparse matrices. The ultra-large-scale sparse matrix is ​​stored in Rutherford-Boeing format, including: (1) converting the Rutherford-Boeing format to row compression for convenient matrix operations and parallel processing, and storing the data in row compression format; (2) each node calculates the row pointer, column index, and non-zero element value of its respective block matrix row compression; (3) each node writes the calculated row pointer, column index, and non-zero element value to its respective file in row compression format; (4) when the application needs the ultra-large-scale sparse matrix, the node reads the data from its respective file, and the process ends. This invention achieves fast reading speeds when multiple nodes read in parallel, eliminates the need for data distribution between nodes, avoids communication congestion, and improves overall performance.

The invention relates to the technical field of signal processing, in particular to a random noise suppression method and device for ground penetrating radar data, and the method comprises the steps: carrying out the self-similar block matching processing of original ground penetrating radar data, so as to obtain a self-similar block matrix; performing multiple synchronous extrusion transformation on the self-similar block matrix to obtain a multiple synchronous extrusion transformation result three-dimensional matrix; applying low-rank constraint to the multi-synchronous extrusion transformation result three-dimensional matrix to establish a low-rank matrix and a sparse matrix; and superposing the low-rank matrix to obtain a denoising result of the time domain, and performing effective signal reconstruction on the original ground penetrating radar data according to the denoising result of the time domain to generate the ground penetrating radar data after random noise suppression. Therefore, the problems that an existing ground penetrating radar data random noise suppression method is insufficient in adaptability, the signal structure is damaged, the non-stationary processing capacity is weak, and the low-energy target extraction effect is poor are solved.

The application provides a distributed security outsourcing computing processing method based on sparse matrix transformation, and relates to the technical field of computers, and the method comprises the following steps: acquiring a first matrix and a second matrix; randomly generating a key set comprising a binary permutation matrix, a sparse triangular matrix and a random vector; performing primary encryption on the first matrix and the second matrix based on the binary permutation matrix; performing secondary encryption on the primary encrypted matrix based on the sparse triangular matrix; distributing the secondary encrypted first secondary encrypted matrix to each edge end in an integrated manner, distributing the second secondary encrypted matrix to each edge end in an even manner after dividing the second secondary encrypted matrix into column vectors according to columns, and enabling the edge end to perform a subtask calculation operation by using the received matrix and return the result to the user end; receiving the calculation results of the edge ends, and aggregating the results; verifying the reliability of the aggregated results by using the random vector, and decrypting the final calculation result. The scheme can improve the security and timeliness of outsourcing computing.

The invention discloses precession target high-resolution imaging based on compressed sensing, and belongs to the technical field of radar imaging. The method comprises the following steps: firstly, obtaining an expression of a radar echo signal according to echo characteristics of a precession target; discretizing an imaging interval on a plane in the coordinate system into a two-dimensional grid to obtain echo signals of all grid points, and constructing a corresponding sparse matrix; parameters required in the construction process of the sparse matrix are solved according to the precession target motion model; and finally, a compressed sensing algorithm is introduced into imaging of the precession target, sparse reconstruction is realized by adopting an OMP algorithm, and a clear ISAR image is obtained. According to the method, on the basis of analyzing the echo characteristics and the spatial state of the precession target in the middle section of the trajectory, the spatial two-dimensional joint sparse base is constructed, the two-dimensional joint imaging of the precession target based on the CS is realized, and the compressed sensing two-dimensional joint imaging method of the precession target established based on the spatial state of the target adopts longer CPI imaging, has higher resolution, and can be used for realizing the two-dimensional joint imaging of the precession target. And the spatial distribution of the target scattering points can be clearly displayed.

The invention relates to the technical field of optical three-dimensional reconstruction, in particular to a three-dimensional fingerprint microstructure reconstruction method and system based on illumination self-calibration and sparse integration, and the method comprises the following steps: constructing a multi-light-source imaging system, and arranging a standard spherical reflector in an imaging field of view to perform illumination direction self-calibration; performing multi-angle illumination acquisition on the surface of the finger to obtain a finger image sequence under multi-frame illumination variation; a reconstruction area of the finger image sequence is limited; carrying out vectorization solution on the pixel brightness change of the reconstruction region under the multi-illumination condition, and obtaining a global continuous normal gradient field based on the surface normal vector and reflectivity of each pixel; the normal gradient field is converted into a depth constraint equation, sparse matrix optimization and least square integral solving are adopted, and global consistent reconstruction of the depth map is achieved. According to the method, high-fidelity three-dimensional reconstruction of the finger surface microstructure is realized through an integrated technical framework of physical calibration, photometric solution and sparse integration.

The invention discloses a window cleaning robot stain identification method and system based on image contrast. The method comprises the steps that S1, a glass surface image is acquired and preprocessed; s2, calculating a reflection intensity index according to the pixel gray scale and the gradient magnitude, and constructing a dynamic weight matrix; s3, constructing the preprocessed image into an observation matrix, and setting a low-rank matrix and sparse matrixes X1 and X2; s4, performing singular value decomposition and thresholding iteration updating on the low-rank matrix; s5, performing total variation constraint update on the sparse matrix X1, and performing L1 norm constraint update on the sparse matrix X2; s6, constructing a contrast diagram according to the difference value of the observation matrix and the low-rank matrix, and marking the positions of the stains with a sparse matrix X2 to form a stain contrast response diagram; and S7, performing threshold segmentation and connected domain marking on the response diagram to obtain a blot area mask. According to the invention, high-precision stain identification of the window cleaning robot under a strong reflection condition is realized.