Patents
Literature
Hiro is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.
Hiro

32results about How to "Facilitates parallel computing" patented technology

DNA dynamic coding based colored image encryption method

The present invention relates to a DNA dynamic coding based colored image encryption method, comprising: obtaining a key through a colored plain-text image; by the key, calculating an initial value and a parameter of a chaotic system used in scrambling and diffusion processes; substituting the initial value and the parameter to the chaotic system for iteration to obtain W1 and W2, obtaining a scrambling matrix K from W1, obtaining an integer matrix Y and an index matrix Ind from W2, and performing DNA coding on Y according to the index matrix Ind; and performing scrambling on the plain-text image row by row by using the matrix K, then performing coding and diffusion until all rows finish DNA coding, and finally performing decoding and recomposing on a DNA matrix, to obtain a final colored ciphertext image. According to the method, a function SHA256 is used to generate the key, so that the key space is expanded; the generation of the key depends on a plain-text, so that plain-text attack can be effectively resisted; chaotic characteristics are combined with DNA dynamic coding, so that the security is further improved; and meanwhile, scrambling and diffusion operations are performed according to the row of images, so that parallel computing is facilitated and the efficiency is improved.
Owner:HENAN UNIVERSITY

Improved SM4 parallel encryption and decryption communication method based on GPU (Graphics Processing Unit)

The invention discloses an improved SM4 parallel encryption and decryption communication method based on a GPU (Graphics Processing Unit). The encryption and decryption method comprises the following steps: using an AES (Advanced Encryption Standard)-128 key expansion algorithm to generate an expanded subkey needed by an SM4 encryption algorithm; improving a round function in the SM4 encryption algorithm; and using the parallel computing power of the GPU, improving the SM4 encryption algorithm into a form of parallel encryption. By adopting the encryption and decryption communication method of the invention, the security and the speed of implementation can be effectively improved, real-time and parallel encryption can be performed on data, and the encryption speed can be improved, so that normal transmission of data flow can be guaranteed and the processing delay can be reduced.
Owner:桂林蓝港科技有限公司

Image encryption algorithm based on memoristor hyper-chaotic system, cellular automaton and DNA calculation

ActiveCN106023052AImprove securityEnhanced resistance to brute force attacksImage data processing detailsChosen-plaintext attackCellular automation
The invention relates to an image encryption algorithm based on memoristor hyper-chaotic system, cellular automaton and DNA calculation. First of all, a four-dimensional memoristor hyper-chaotic system based on a Chua circuit is employed, at the same time, a secret key is generated by enabling a plaintext image on an SHA256 function; a pseudo random sequence generated by the chaos system is taken as a basis of DNA dynamic encoding, rules of cellular dynamic evolution are determined according to a DNA matrix after diffusion, and in an encryption process, partitioning processing is performed on the image; and finally, the image is encrypted through a pseudo DNA calculation method, and diffusion operation is carried out while the image is scrambled. According to the invention, on one hand, secret key space is increased and the capability of the algorithm in fending off violent attacks is enhanced; and on the other hand, the secret key is enabled to be generated through close reliance on a plaintext, the capability of the algorithm in fending off plaintext attacks is improved, the capability of fending off exhausive attacks is higher, parallel calculation is facilitated, the operation efficiency is further improved, and the security of the algorithm is improved.
Owner:HENAN UNIVERSITY

Deep stack network-based electroencephalogram signal feature extraction and classification method

The invention discloses a deep stack network-based electroencephalogram signal feature extraction and classification method. The method comprises the steps of firstly acquiring electroencephalogram signal data by using an Emotiv electroencephalogram signal acquisition instrument; performing preprocessing of mean removal, filtering, normalization and the like on electroencephalogram signals; then performing independent pre-training on electroencephalogram signals of a single channel by using a plurality of restricted Boltzmann machines, extracting the electroencephalogram signals of the single channel, and applying parameters obtained by training to parameter initialization of a neural network; finally performing micro-adjustment on the network by adopting a batch gradient descent method, and effectively fusing electroencephalogram signal features of all channels; and performing performance testing on the network and implementing classification. According to the method, relatively high classification accuracy can be obtained.
Owner:西安慧脑智能科技有限公司

Equipment health management system and database modeling method thereof

The invention discloses an equipment health management system and a database modeling method thereof, and belongs to the technical field of management systems. The equipment health management system comprises a CBM equipment health management subsystem, an equipment health information visualization system and an expert system, and further comprises a space-grid-based database including a space grid node data packet composed of space grids and grid nodes, wherein relative space coordinates of nodes and activation time of the nodes are stored in the space grid node data packet. The system and the method provided by the invention have the advantages that firstly, an equipment management space is subjected to space grid division, so that grid data can be stored in a distributed manner, grid parallel computing is facilitated, and data processing efficiency is improved; secondly, the traditional journal recording mode based on a time line is converted into a space-based journal recording mode, so that journal recording modes are enriched; thirdly, three-dimensional visualization processing is performed on equipment health state information corresponding to the space grids.
Owner:CHINA UNIV OF MINING & TECH +1

Method and apparatus for processing datum

The invention provides a method and an apparatus for processing datum, wherein a data space is established; pointers and address information of the datum are stored in the data space; the data space is divided into space dimension of over three dimensions to enable the process operation of a computer to be above the space dimension of over three dimensions; dependency doesn't exist between processes of different space dimensions; steps of the method are that the datum to be processed are loaded into the data space according to the pointers and the address information stored in the data space; the datum to be processed are distributed to processes located in different space dimensions; and the processes of the different space dimensions are controlled and the distributed datum to be processed are processed, thus the parallel computation of the computer processes is improved.
Owner:INSPUR FINANCIAL INFORMATION TECH CO LTD

Three-dimensional vibration analysis method for composite laminated structure

The invention relates to a three-dimensional vibration analysis method for a composite laminated structure, and the method is applied to the fields of engineering mechanics and vibration engineering. The method comprises: extracting geometric, material and boundary condition parameters of the structure, and setting J non-uniformly distributed calculation planes in a thickness direction for each layer of the structure; configuring displacements of the structure on the calculation planes; performing in-plane component expansion on the displacement of the structure on each calculation plane, applying a boundary condition through a boundary feature function, and obtaining the displacement of the structure on the calculation plane; calculating in-plane directional strain of the structure on the jth calculation plane in an lth layer; calculating strain and stress of the structure in any position of the lth layer; establishing an energy equation of the structure; and calculating a feature equation of the structure. According to the method, the structure is decomposed into a plurality of calculation planes through spatial lamination plane setting; on one hand, the dimensionality of the structure is reduced, so that the calculation speed is increased and the calculation cost is reduced; and on the other hand, the structure is divided into parts, so that parallel calculation is facilitated and the calculation efficiency is improved.
Owner:HARBIN ENG UNIV

A coordinated control method of a multi-energy complementary system

The invention discloses a coordinated control method of a multi-energy complementary system, which comprises the following steps: in a day-ahead scheduling stage, determining a start-stop period of adevice with a large inertia link through a day-ahead robust optimal scheduling based on a long-time prediction information of the next day; in the intra-day scheduling stage, determining the output power of the continuously adjustable device by intra-day multi-objective quantum particle swarm optimization based on intra-day short-time prediction information; by using the cooperative optimization of day-ahead and day-within-day scheduling, finally determining the operation mode of the multi-energy complementary system, and realizing the coordinated control of the multi-energy complementary system. While guaranteeing the safe and stable operation condition of the system, the method considers the optimal utilization of resources and realizes the minimum operation cost of the system.
Owner:NR ELECTRIC CO LTD +1

Gravitational field numerical simulation method and device based on complex terrain and computer equipment

The invention relates to a gravitational field numerical simulation method and device based on complex terrains, computer equipment and a storage medium. The method comprises the following steps: constructing a three-dimensional cuboid model containing a complex terrain and an exploration target, performing mesh generation to obtain a plurality of small cuboids, and setting spatial domain density values of the small cuboids; obtaining an offset wave number according to grid parameters of grid subdivision and preset Gaussian parameters, dividing the three-dimensional cuboid model into a lower part model and an upper part model according to topographic information, performing block superposition to calculate a frequency domain gravity anomaly field of a plurality of observation surfaces, and performing two-dimensional Gaussian Fourier inversion to obtain spatial domain gravity anomaly fields of multiple observation surfaces, and obtaining a spatial domain gravity anomaly field of the complex topographic relief surface through a spline interpolation algorithm. The vertical direction is reserved in a spatial domain, mesh generation is flexible, the method and device are suitable for simulating complex terrains and complex geologic bodies, multiple offset wave numbers are convenient for parallel calculation, and the gravity anomaly calculation efficiency is improved.
Owner:CENT SOUTH UNIV

CT imaging method

The invention discloses a CT imaging method. The method comprises the following steps of: establishing a coordinate system which covers a detected area, source points and observation points; obtainingand setting coordinates of the source points and the observation points as parameters 1, calculating and setting a source point-observation point linear equation as a parameter 2, and arranging a measurement calculation result on each pair of source point-observation point line as a parameter 3; establishing a parameter database according to the parameters 1, the parameter 2 and the parameters 3;determining a calculation area, discretizing the calculation area into quasi-calculation image points, and obtaining a coordinate parameter 4 of each image point; respectively taking a quasi-calculation image point to set an enrolling rule and an enrolling line, calculating a distance Ri from each image point to the enrolling line and recording the total number M of the enrolling lines; accumulating weight parameters 3 which take Ri as association and act on the parameter 3 corresponding to the image point according to a distance inverse ratio rule, and carrying out averaging by the M value so as to obtain a parameter value 5 of the image point; and carrying out global imaging via the parameter value 5. The method does not need to construct and solve equation sets by using the traditionalCT method, does not need iterative calculation and is wide in application fields.
Owner:CENT SOUTH UNIV +1

Power distribution network state estimation method based on network decomposition

The invention provides a power distribution network state estimation method based on network decomposition. The method comprises the following steps that a power distribution network is decomposed; sub-network state variable initialization is carried out; loop iteration solving is carried out; constraint condition equivalent measuration weight is computed; a sub-network state variable is solved; whether a convergence condition is met is judged, if yes, iteration is completed, and power distribution network state estimation results are output; and otherwise, 1 is added to iteration times automatically, and loop iteration solving is carried out. The applicability of a state estimation algorithm is improved, a large network can be subjected to state estimation, by network decomposition, the data scale of a node admittance matrix is lowered, the storage space requirements and computing complexity are lowered, for the large network, compared with direct computing, computing speed is high, and efficiency is high.
Owner:STATE GRID CORP OF CHINA +1

Geosynchronous orbit SAR squint imaging processing device for on-satellite walking removal and 2D spatial-variant correction

The invention discloses a geosynchronous orbit SAR squint imaging processing device for on-satellite walking removal and 2D spatial-variant correction. According to the device, sequentially, on a satellite, a receiving window is adjusted to remove signal walking through the satellite, a walking-removal echo link is downloaded, a spatial-variant model is constructed according to a satellite-groundgeometric relation, and a range direction residual walking correction, range migration correction and compression, azimuth direction partial dechirp for reducing a time span, azimuth direction residual walking removal, azimuth direction nonlinear conversion and compression and final geometric correction and imaging result outputting are conducted. Compared with the prior art, the feature of quitelow repetition frequency under GEO SAR is used, and while range direction and azimuth direction coupling is reduced through on-satellite walking removal, the effectiveness of data storage on the satellite is improved; for received echoes, partial fined correction is conducted in the azimuth direction and range direction, so that the focusing precision of an algorithm is improved.
Owner:BEIHANG UNIV

Parallel computing method for solving rate theoretical equation based on exponential time difference format

The embodiment of the invention provides a parallel computing method for solving a rate theoretical equation based on an exponential time difference format. A physical microdefect simulation model isestablished based on a rate theory, and the rate theory has no space-time scale limitation, so that the advantages of the rate theory can be obviously reflected when microstructure evolution under a high damage dose condition is simulated, and then the exponential time difference format is used for solving the main equation, so that the solving result is better in accuracy and higher in precision.
Owner:CHINA INSTITUTE OF ATOMIC ENERGY +1

Color Image Encryption Method Based on DNA Dynamic Coding

The present invention relates to a DNA dynamic coding based colored image encryption method, comprising: obtaining a key through a colored plain-text image; by the key, calculating an initial value and a parameter of a chaotic system used in scrambling and diffusion processes; substituting the initial value and the parameter to the chaotic system for iteration to obtain W1 and W2, obtaining a scrambling matrix K from W1, obtaining an integer matrix Y and an index matrix Ind from W2, and performing DNA coding on Y according to the index matrix Ind; and performing scrambling on the plain-text image row by row by using the matrix K, then performing coding and diffusion until all rows finish DNA coding, and finally performing decoding and recomposing on a DNA matrix, to obtain a final colored ciphertext image. According to the method, a function SHA256 is used to generate the key, so that the key space is expanded; the generation of the key depends on a plain-text, so that plain-text attack can be effectively resisted; chaotic characteristics are combined with DNA dynamic coding, so that the security is further improved; and meanwhile, scrambling and diffusion operations are performed according to the row of images, so that parallel computing is facilitated and the efficiency is improved.
Owner:HENAN UNIVERSITY

A three-dimensional vibration analysis method for composite laminate structures

The invention relates to a three-dimensional vibration analysis method for a composite laminated structure, and the method is applied to the fields of engineering mechanics and vibration engineering. The method comprises: extracting geometric, material and boundary condition parameters of the structure, and setting J non-uniformly distributed calculation planes in a thickness direction for each layer of the structure; configuring displacements of the structure on the calculation planes; performing in-plane component expansion on the displacement of the structure on each calculation plane, applying a boundary condition through a boundary feature function, and obtaining the displacement of the structure on the calculation plane; calculating in-plane directional strain of the structure on the jth calculation plane in an lth layer; calculating strain and stress of the structure in any position of the lth layer; establishing an energy equation of the structure; and calculating a feature equation of the structure. According to the method, the structure is decomposed into a plurality of calculation planes through spatial lamination plane setting; on one hand, the dimensionality of the structure is reduced, so that the calculation speed is increased and the calculation cost is reduced; and on the other hand, the structure is divided into parts, so that parallel calculation is facilitated and the calculation efficiency is improved.
Owner:HARBIN ENG UNIV

Behavior prediction method and device based on zoom dot product attention, equipment and medium

ActiveCN112651782APeriodically characterize complete and fullAchieve the goal of discontinuous behavior decouplingMachine learningEnergy efficient computingConcurrent computationAlgorithm
The invention relates to the field of artificial intelligence, and provides a behavior prediction method and device based on zoom dot product attention, equipment and a medium. Firstly, a long time sequence is split based on a time interval, a user layer and a behavior sequence layer are separated by a time interval layer and are not directly connected, the sequence length in the interval is greatly reduced, and parallel computing is facilitated; the cross calculation amount is reduced, the operation efficiency is improved, encoding is carried out based on a scaling dot product attention mechanism, data features are optimized, to-be-predicted data is subjected to standardization processing to conform to the input data format of the model, the model is conveniently utilized for prediction, and then automatic prediction of behaviors such as purchasing is achieved based on an artificial intelligence means; and the accuracy and the prediction efficiency are higher. The invention also relates to a blockchain technology, and the behavior prediction model can be stored in the blockchain.
Owner:CHINA PING AN LIFE INSURANCE CO LTD

A dem-based coding method for grid calculus order of grid hydrological model

The invention relates to a DEM-based grid-type hydrological model grid calculus sequence coding method, including: determining the grid size, obtaining the grid water flow direction, dividing the river channel grid and sub-basin, sub-basin coding, confluence layer coding, layer Internal sequence coding, inflow grid coding and channel coding. Through the coding method of the present invention, the calculation of flow yield and confluence of the grid-type hydrological model can be realized. The coding rules assign different identification values ​​to each sub-basin of the watershed, which is convenient for parallel calculation. During the coding process, the river grid and the non-river network are distinguished Grids, and the hydraulic connection between grids only occurs between adjacent layers, it is convenient to find problems in the calculation process through coding rules according to the topological relationship in the calculation process.
Owner:CHINA INST OF WATER RESOURCES & HYDROPOWER RES

Numerical simulation method, device and computer equipment for gravity field based on complex terrain

The present application relates to a method, device, computer equipment and storage medium for numerical simulation of gravity field based on complex terrain. The method includes: constructing a three-dimensional cuboid model containing complex terrain and exploration targets, performing grid division to obtain a plurality of small cuboids, and setting the spatial domain density value of the small cuboids; Gaussian parameters are set to obtain the offset wavenumber. According to the terrain information, the three-dimensional cuboid model is divided into the lower part model and the upper part model, and the frequency domain gravity anomaly field of multiple observation surfaces is superimposed to calculate the two-dimensional Gaussian Fourier transform. The inverse transformation is used to obtain the gravity anomaly field in the space domain of multiple observation surfaces, and the gravity anomaly field in the space domain of the complex terrain undulation surface is obtained through the spline interpolation algorithm. The invention retains the vertical direction in the space domain, has flexible grid division, and is suitable for simulating complex terrain and complex geological bodies. Multiple offset wave numbers are convenient for parallel calculation, and the calculation efficiency of gravity anomaly is improved.
Owner:CENT SOUTH UNIV

Image Encryption Method Based on Memristor Hyperchaotic System, Cellular Automata and DNA Computing

ActiveCN106023052BImprove securityEnhanced ability to resist exhaustive attacksImage data processing detailsChosen-plaintext attackCellular automation
The invention relates to an image encryption algorithm based on memoristor hyper-chaotic system, cellular automaton and DNA calculation. First of all, a four-dimensional memoristor hyper-chaotic system based on a Chua circuit is employed, at the same time, a secret key is generated by enabling a plaintext image on an SHA256 function; a pseudo random sequence generated by the chaos system is taken as a basis of DNA dynamic encoding, rules of cellular dynamic evolution are determined according to a DNA matrix after diffusion, and in an encryption process, partitioning processing is performed on the image; and finally, the image is encrypted through a pseudo DNA calculation method, and diffusion operation is carried out while the image is scrambled. According to the invention, on one hand, secret key space is increased and the capability of the algorithm in fending off violent attacks is enhanced; and on the other hand, the secret key is enabled to be generated through close reliance on a plaintext, the capability of the algorithm in fending off plaintext attacks is improved, the capability of fending off exhausive attacks is higher, parallel calculation is facilitated, the operation efficiency is further improved, and the security of the algorithm is improved.
Owner:HENAN UNIVERSITY

Image reconstruction method and device, terminal and storage medium

The embodiment of the invention discloses an image reconstruction method, device and equipment and a storage medium, and the method comprises the steps: obtaining original projection data obtained based on an inclined scanning mode, and determining an initial value image corresponding to the original projection data; performing iterative reconstruction on the original projection data based on an iterative reconstruction algorithm and the initial value image to obtain an iterative reconstruction image; carrying out interpolation on the iterative reconstruction image to obtain a target reconstruction image under an inclined coordinate system, wherein the X coordinates and the Y coordinates of the edge pixel points on the same side of each layer of sectional image arranged in the Z-axis direction corresponding to the initial value image in the inclined coordinate system are the same, the oblique coordinate system takes the direction parallel to the rotating shaft of the rack as a Z-axis direction, takes two orthogonal directions perpendicular to the rotating shaft as an X-axis and a Y-axis, the X-axis is perpendicular to the working surface of the scanning bed, the Y-axis is parallelto the working surface of the scanning bed, and the origin of coordinates is a rotating center. According to the technical scheme of the embodiment of the invention, the iterative reconstruction efficiency can be improved.
Owner:SHANGHAI UNITED IMAGING HEALTHCARE

Multi-target scattering echo parallel computing method and device, equipment and storage medium

The invention provides a multi-target scattering echo parallel computing method and device, equipment and a storage medium, the method is applied to a VxWorks system carried on a hardware platform, and the method comprises the following steps: determining the number of CPU cores and the number of to-be-processed targets in the hardware platform; creating more than two computing tasks corresponding to the CPU cores according to the number of the CPU cores in the hardware platform; determining a target having affinity with each CPU core according to the number of the CPU cores in the hardware platform and the number of the targets; and on each CPU core with the target with the affinity, calling a corresponding calculation task to calculate the target with the affinity with the CPU core. According to the scheme, different CPU cores can call respective corresponding calculation tasks to complete the calculation process in parallel, and compared with serial calculation, the calculation efficiency of multi-target scattering echo calculation on an embedded real-time operating system is greatly improved.
Owner:BEIJING INST OF ENVIRONMENTAL FEATURES

A Lane Edge Aggregation Method Based on Trajectory Direction

The invention relates to a lane sideline aggregation method based on track direction, comprising: taking the track line segment between two adjacent track points as the center, generating a track line segment buffer according to a set width; combining the buffer areas of each track line segment Generate a buffer list, judge whether to keep or discard the track point according to whether the track point is in the buffer list, and generate a reference track according to each reserved track point; extend the track point in the reference track to both sides to generate a scan line segment, for the same scan line segment The intersection points with the original lane edge are clustered to generate each cluster point, and each cluster point is classified and fitted into a line to obtain the aggregation result of the lane edge. Extracting the trajectories of characteristic directions from a large number of trajectories can avoid a large number of original trajectories from participating in the calculation and improve the calculation efficiency; the intersection of the trajectory scan line and the lane edge is introduced as a new clustering element, so that the final generated lane shape is especially miscellaneous and The lane shape of the curve is in line with the actual situation.
Owner:WUHAN ZHONGHAITING DATA TECH CO LTD

Disturbance gravitational field application-oriented fine calculation method and system

The invention relates to a disturbance gravitational field application-oriented fine calculation method, which comprises the following steps of: recovering a spherical integrable function f (theta, lambda) according to discrete disturbance gravitational observation data, calculating disturbance potential coefficients by using a formula, and substituting the disturbance potential coefficients intodifferent disturbance gravitational field element calculation formulas to obtain required gravitational field elements. The invention also provides a disturbance gravitational field application-oriented fine computing system. The method is high in stability, the significant digits can be accurate to the last six digits of a decimal point, meanwhile, the expression can calculate all values at a time according to different times m, GPU parallel computing is facilitated, the computing speed is further increased, and a computing tool is provided for modeling application of the ultra-high-order disturbance gravitational field.
Owner:BEIJING INST OF ASTRONAUTICAL SYST ENG +1

A fine calculation method and system for the application of disturbance gravitational field

The invention relates to a disturbance gravitational field application-oriented fine calculation method, which comprises the following steps of: recovering a spherical integrable function f (theta, lambda) according to discrete disturbance gravitational observation data, calculating disturbance potential coefficients by using a formula, and substituting the disturbance potential coefficients intodifferent disturbance gravitational field element calculation formulas to obtain required gravitational field elements. The invention also provides a disturbance gravitational field application-oriented fine computing system. The method is high in stability, the significant digits can be accurate to the last six digits of a decimal point, meanwhile, the expression can calculate all values at a time according to different times m, GPU parallel computing is facilitated, the computing speed is further increased, and a computing tool is provided for modeling application of the ultra-high-order disturbance gravitational field.
Owner:BEIJING INST OF ASTRONAUTICAL SYST ENG +1

A ct imaging method

The invention discloses a CT imaging method. The method comprises the following steps of: establishing a coordinate system which covers a detected area, source points and observation points; obtainingand setting coordinates of the source points and the observation points as parameters 1, calculating and setting a source point-observation point linear equation as a parameter 2, and arranging a measurement calculation result on each pair of source point-observation point line as a parameter 3; establishing a parameter database according to the parameters 1, the parameter 2 and the parameters 3;determining a calculation area, discretizing the calculation area into quasi-calculation image points, and obtaining a coordinate parameter 4 of each image point; respectively taking a quasi-calculation image point to set an enrolling rule and an enrolling line, calculating a distance Ri from each image point to the enrolling line and recording the total number M of the enrolling lines; accumulating weight parameters 3 which take Ri as association and act on the parameter 3 corresponding to the image point according to a distance inverse ratio rule, and carrying out averaging by the M value so as to obtain a parameter value 5 of the image point; and carrying out global imaging via the parameter value 5. The method does not need to construct and solve equation sets by using the traditionalCT method, does not need iterative calculation and is wide in application fields.
Owner:CENT SOUTH UNIV +1
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products