41 results about "Iteration function" patented technology

System and method for automatically parallelizing iterative functionality in a data flow program. A data flow program is stored that includes a first data flow program portion, where the first data flow program portion is iterative. Program code implementing a plurality of second data flow program portions is automatically generated based on the first data flow program portion, where each of the second data flow program portions is configured to execute a respective one or more iterations. The plurality of second data flow program portions are configured to execute at least a portion of iterations concurrently during execution of the data flow program. Execution of the plurality of second data flow program portions is functionally equivalent to sequential execution of the iterations of the first data flow program portion.

The invention provides a fractal-based video compression and decompression method. The compression method is a novel video compression and encoding method which adopts the fractal iteration principle for video encoding, and comprises the following steps: encoding a starting frame through block-DCT transformation, and carrying out block motion estimation / compensation coding for a non-I frame, which comprises calculating the sum and sum of squares of the pixels of sub-blocks related to sub-block domains and parent-block domains, finding a most similar matching block in a previous frame search window through full search, and finally compressing factors of an iteration function system through Huffman encoding. The corresponding decompression method comprises the following steps: decoding the I frame through anti-DCT transformation, carrying out Huffman decoding for the non-I frame to obtain the factors of the iteration function system, and carrying out macroblock-based decoding, which comprises calculating the sum and sum of squares of the pixels of the sub-blocks relevant to the parent-block domains, and decoding the macroblocks in the current frame in sequence. The method improves the traditional method of fractal video compression, and not only greatly increases the compression ratio and the peak signal-noise ratio, but also improves the encoding speed, thereby enhancing the fractal video compression and decoding performance, and making the fractal video compression and decoding more practical.

The invention discloses a control parameter optimization method for improving the phase commutating failure restraining capacity of a direct current system. The method comprises the following steps that after initialization, through an interface technology between PSCAD and MATLAB, a first interface module and a second interface module are established, the first interface module is used for calling an optimization algorithm, a whole iteration function of the algorithm is carried out, when each-time simulation is started, through an improved particle swarm algorithm, particle positions are updated, and controller parameters are returned to the PSCAD; and the second interface module is used for receiving the particle positions returned by the PSCAD and optimizing target functions after each-time simulation is over, and the global optimum particle positions and the individual optimum positions are updated. Through the interface technology between the PSCAD and the MATLAB, combining calling of the PSCAD and the MATLAB is achieved, the improved particle swarm algorithm is used for searching for the optimum control parameters, a multi-time operation function in the PSCAD is used, calling of multi-time software can be achieved, and accordingly optimization searching of control parameters is achieved.

The invention discloses a method for calculating a working temperature of a component IGBT of a photovoltaic inverter. Based on an energy conservation law, a mathematic relational formula of an external environment temperature, input power and IGBT loss and temperature rise is established through a loop iteration method by adopting a heat transfer principle and junction temperature and heat resistance principles; and calculation is performed by utilizing a loop iteration function in EXCEL to obtain shell temperatures and junction temperatures of the high-power component of the inverter under different environment temperatures and different input power. The errors between the predicted working temperatures of the IGBT under the different environment temperatures and the different input power, and the actually measured temperatures are within + / -8%; the method is used for predicting the working temperatures of the high-power component of the photovoltaic inverter under high and low-temperature environments and high and low-load ratios; and a theoretical basis is provided for assessing the reliability of the inverter.

The invention provides a key expanding method for a block cipher algorithm, which is used for generating each round of subkeys in the block cipher algorithm. The method provided by the invention mainly comprises the following step of calculating an iteration function for generating a subkey, wherein the iteration function is formed by a nonlinear operation, a constant operation and a cycle shift operation. The method provided by the invention can effectively and conveniently generate the subkeys, effectively reduce storage space, has higher realizing speed compared with a common commercial block cipher algorithm, can meet the requirements on real-time encryption and decryption, and simultaneously has higher safety.

The invention provides a fractal and H.264-based binocular three-dimensional video compression and decompression method, which comprises the following steps of: performing H.264-based intra-frame prediction coding on a starting frame (frame I) of a left eye, performing filtration to obtain a reconstructed frame (which can be used as a reference frame), performing block motion compensated prediction (MCP) coding on a frame P of the left eye, recording fractal parameters of blocks in each optimal matching process, substituting the fractal parameters into a compression iteration function system to obtain a predicted frame of the frame P respectively, and performing filtration to obtain a reconstructed frame; performing prediction on a starting frame of a right channel from the frame I of a base layer in a disparity compensated prediction (DCP) way; performing MCP and DCP on the frame P, and selecting a result with a minimum error as a prediction result for coding; and compressing residual frames of the frames I and P by utilizing entropy coding CALCV, and compressing the fractal parameters by utilizing signed index Columbus coding.

The invention provides a flow control strategy processing method and apparatus. The method comprises the steps that a BGP module of a network device receives a FlowSpec flow control strategy sent by an EBGP neighbor, and sends a redirection next hop address carried in the flow control strategy to an iteration module, the iteration module sends an iteration result change notification to the BGP module in the case of a reachability change of the next hop address or an optimal matching routing change of the next hop address, and the BGP module checks the validity of the next hop address again according to the iteration result change notification, and processes the flow control strategy according to a validity check result. According to the method, the redirection next hop address of the FlowSpec flow control strategy is quickly checked by the existing routing iteration function, so that massive unnecessary retrieving operations of the flow control strategy in the case of the routing change are reduced, and unnecessary resource occupation is reduced for the network device.

The invention discloses an intelligent resource cataloguing method for big data resource management. The intelligent resource cataloguing method comprises three parts: a data element knowledge base, data table field benchmarking and data resource intelligent classification, and is mainly embodied in calculation of an algorithm and a calculation method of the whole cataloguing, and the specific operation comprises the following steps: creating the data element knowledge base; carrying out data table field benchmarking; performing intelligent cataloguing by the system according to the knowledge base; and updating the iteration data element knowledge base. The intelligent resource cataloguing method has the advantages that the number of the data resource tables is large; only data element classification of metadata knowledge needs to be continuously perfected to realize automatic cataloguing of resources; moreover, the scheme also provides a data element automatic improvement and iteration function, and the higher the subsequent intelligent cataloguing precision is along with the increase of the cataloguing times of the data resources, and the problems that the resource cataloguing efficiency is low and the accuracy is difficult to ensure by manpower at present are solved, and the resource cataloguing working efficiency and accuracy are improved.

The present invention provides a detection and analysis method of a communication machine room device health state. The method comprises the steps of: collecting sound data and vibration data of a communication machine room device in a certain time range, and converting the collected sound data and vibration data to a digital quantity and transmitting the digital quantity to a mobile phone app; performing primary data processing of the received digital quantity by the mobile phone app, if the digital quantity exceeds a reference range value, taking the digital quantity as an abnormal data to upload to an analysis platform, or else, taking the digital quantity as normal data, wherein the reference range has a self-iteration function; performing deep analysis of the abnormal data through theanalysis platform so as to determine whether the health state of the communication machine room device is abnormal or not at present; and predicting the development trend of the communication machineroom device health state and a time point reaching the dangerous level. The detection and analysis method employs the sound data and the vibration data of the communication machine room device to determine whether the health state of the communication machine room device at present is abnormal or not and predicts the development trend of the communication machine room device health state and thetime point reaching the dangerous level.

The invention discloses an embedded rotation angle calculation IP soft core based on a PLB bus and a rotation angle calculation method. The calculation IP soft core comprises a main control module, a PLB bus timing translation module, a configuration register module, a lookup table RAM module, a FIFO module, a quadrant transformation module and multiple iteration modules; the multiple iteration modules form a multistage pipeline organization, the output is fed back to the input through the iteration modules so that a multiple iteration function can be achieved, a feedback type pipeline organization is adopted by the rotation angle calculation, namely a 'true' pipeline organization with feedback performed earlier and pipeline performed later, single precision floating-point type sine values and cosine values are transformed into single precision floating-point type angle values, and the PLB bus is connected with an on-chip processor through an on-chip bus. The method has the advantages that the method is superior to a 'pure' feedback type iterative structure and a 'false' pipeline organization with pipeline performed earlier and feedback performed later in computation speed, the method is superior to a 'pure' pipeline organization in occupied resources, and the method is superior to a traditional design in portability in embedded systems.

The invention relates to a generation method of patterns being used for printed fabric and having fractal textures, and in particular relates to a generation method of printed fabric patterns based on generalized Mandelbrot set. According to a complex variable function theory, a spinning pattern is designed by adopting methods of copying, rotating, splicing and the like through constructing a generalized Mandelbrot set-based fractal graph and local detail graphs thereof as basic elements by using computer programming, and then is directly transferred to the fabric by an electronic inkjet printing machine or subjected to color separation by adopting CAD (computer aided design) software for carrying out plate-making printing to obtain the printed fabric with the fractal textures. The invention provides an iteration function and a real part and imaginary part separation method. Through changing arrangement of parameters and color values in an iteration formula, an image amplification region, amplification factors and the like, images with different styles are obtained, and innumerable and changeful image series can be conveniently and rapidly obtained, thus rapid design of the printed pattern is realized.

The present invention discloses a high-dimensional data hypergraph model construction method based on feature induction. The method comprises the following steps: discretizing t attribute values of n data records of a high-dimensional data set D, using one data record as one row of an initial matrix X, and using the discretized attribute values of the data record as a column, thereby obtaining the initial matrix X; under the non-negative condition, initializing a high-dimensional data set feature-based matrix U and a high-dimensional data set feature coefficient matrix V; performing repeated iteration on the U and V by using an iteration function to obtain an approximate solution until a value of an objective function Q(X, U, V) is reduced to a set threshold, thereby obtaining a matrix U' with a reduced scale; and considering each row of the matrix U' as a data record, defining different attribute values as nodes of a hypergraph, and constructing one hyperedge of the hypergraph with each row of the matrix U', thereby obtaining the hypergraph G. The high-dimensional data hypergraph model construction method based on feature induction is capable of performing complete cluster analysis on a high-dimensional data set, and can further improve the operation efficiency of a high-dimensional data clustering algorithm.

The invention relates to a firewall / a virtual private network integrated system and a circuit for connecting at least one local area network (LAN) to a wide area network (WAN), which comprises a firewall / a virtual private network (VPN) integrated chipset for transmitting and receiving data packets between the WAN and the LAN and analyzing access control function according to the data packets.The firewall / virtual private network (VPN) at least comprises a firewall part for ensuring the access control function between the WAN and the LAN, a VPN part for ensuring the safety function of the data between the WAN and the LAN and a secondary firewall engine.Complete data packets between the WAN and the LAN are analyzed and the release of data is indicated in a buffer area based on the analyzed results by the secondary firewall engine.The firewall part at least comprises a firewall hardware platform and a firewall software platform.At least the firewall hardware platform ensures the repetition iteration function related to the access control.The VPN part at least comprises a VPN hardware platform and a VPN software platform.At least the VPN hardware platform ensures the repetition iteration function related to the safety function.

The invention relates to a two-dimensional visualization encryption method for genetic information based on an iteration function. Individual genetic information comprises traditional STR (short tan-dem repeat) site information and numerous pieces of SNP (single nucleotide polymorphism) information. The genetic information is represented by small checks of different coordinates and different colours in a two-dimensional planer grid by mapping. The invention provides a method that the genetic information represented in the planer grid is iterated by an iterative function. In the iteration process, the original genetic information is concealed, patterns with a certain rule are obtained, and a visible encrypted individual 'genetic fingerprint' pattern is formed. The pattern is convenient to carry out difference comparison and can be used for aspects of individual biological feature identification, biological feature keys, genetic information visualization and the like.

The embodiment of the invention discloses a parameter error calibration method and device and a memory medium. The method comprises the steps of obtaining output information of at least one measurement device in an inertial measurement unit; obtaining the output information of an auxiliary correction module and the output information of a barometric altimeter, wherein the auxiliary correction module comprises a GPS measurement module and / or an odometer; and calibrating at least one parameter error of the measurement device in the inertial measurement unit according to the output information ofthe at least one measurement device in the inertial measurement unit, the output information of the auxiliary correction module and the output information of the barometric altimeter. According to the technical scheme provided by the embodiments of the invention, the output information of the at least one measurement device in the inertial measurement unit, the output information of the auxiliarycorrection module and the output information of the barometric altimeter are obtained, the optimum calibration parameter is obtained through utilization of an iteration function of a genetic algorithm, and measurement precision of the inertial measurement unit is improved.

The invention provides a method for predicting the shaking of an impeller. The method includes the steps of conducting grid division on an entity model of the impeller through a finite element analysis method, dividing the machining process of the impeller into a plurality of machining sub-stages, obtaining modulus parameters of all the machining sub-stages of the impeller through the finite element analysis method, establishing a transfer function of each machining sub-stage of the impeller, obtaining the modulus parameters of a tool system and the modulus parameters of an impeller system through an experiment modulus analysis method, establishing a transfer function of a tool-impeller system, calculating the difference between the modulus parameters of each machining sub-stage of the impeller and the modulus parameters, obtained through the experiment modulus analysis method, of the impeller system, obtaining the modulus parameter deviation vector, conducting iteration on the modulus parameter deviation vector through a modulus parameter iteration function, obtaining the final modulus parameters of the modulus parameter deviation vector within the modulus parameter tolerance range of all the machining sub-stages, and drawing the milling stability lobe graph of all the machining sub-stages of the impeller.

The invention discloses a multimedia data matching information processing method. The method comprises the following steps of: obtaining a sample; obtaining global information characteristics and local information characteristics of the first sample and the second sample, wherein the first sample is an image sample; wherein the second sample is a video sample; inputting the global information characteristics and the local information characteristics into an initial neural network model; and training the initial neural network model through a preset iteration function to obtain a target neuralnetwork model, the target neural network model being capable of determining the matching degree between the image and the video according to the input global information characteristics and local information characteristics of the first target object and the second target object. According to the multimedia data matching information processing method provided by the invention, the accuracy of matching calculation can be improved when the matching calculation of the image and the video is carried out based on the global information characteristics and the local information characteristics of the image and the video.

The invention discloses a seismic signal processing method and corresponding processing system based on an improved FastICA algorithm. Through maximizing a negative entropy object function, a non-Gaussian maximum value required by observing signal linear projection is searched by using a fixed point iteration theory so as to fulfill an aim of separating signals; an iteration function is correctedby introducing correction factors improved by utilizing a gradient method, and the FastICA algorithm is optimized. According to the improved FastICA algorithm, the sensibility to an initial weight value is reduced, and the iteration times are reduced, so that the convergence speed of the algorithm is improved.

In a material cutting optimization method, a material cutting plate and a shape to be cut out are received. Then, a point cloud is generated according to the shape, and a maximal surrounding box of the material cutting plate is calculated. After that, a triangular mesh surface of the point cloud is constructed and a total area of triangles in the triangular mesh surface is computed. A maximum number of the shapes that can be contained in the material cutting plate is calculated and the shapes are arranged in the material cutting plate according to the total area of the triangles, using an iteration function.

The invention discloses a method for obtaining a control chart of a product. The method comprises the following steps: obtaining an iteration function value corresponding to a target product; when the number of iterated operations does not reach the preset number of iterated operations, adjusting a preset constraint expectation weight lifting matrix criterion according to the number of iterated operations to obtain an adjusted constraint expectation weight lifting matrix criterion; using the adjusted constraint expectation weight lifting matrix criterion and a preset constraint condition to process the iteration function value to obtain an updating point function value; and performing iterative operation by using the updated point function value until the updated number of iterated operations reaches a preset number of iterated operations, and obtaining a control chart of the target product according to the iterative function value of the last iterative operation. The invention further discloses a product control chart obtaining device, terminal equipment and a storage medium. By using the method provided by the invention, the iteration function value of the last iteration operation is improved, and the technical effect of improving the accuracy of the control chart of the product is realized.

The invention relates to a rendering method and device for a fractal pattern. The color rendering method and device for a fractal pattern combine an iteration function system theory with a hidden Markov model to see a color of the fractal pattern as an observable outer representation which is determined by a series of unobservable states. A Markov sequence is utilized to control a color change of the fractal pattern and the color control of the fractal pattern is realized. The color rendering method is simple, overcomes a limitation of a traditional iteration function, establishes a more flexible and practical random model, performs perturbation on a traditional probability matrix to realize pattern color arbitrary regulation and control through designing a coloring rule during a pattern drawing process and can fast generate a big quantity of fractal artistic patterns rich in color changing.

The embodiment of the invention discloses a test case generation method, an automatic test method and a computer program product. The method is used for generating a corresponding test case for an iteration function to be applied to an online environment in a function iteration process of a target application program, and comprises the following steps: obtaining an atomic instruction set, and providing a target page comprising an information submission area and an instruction display area; obtaining a test step submitted by a test user in a natural language description mode through the information submission area, calling an AI model, and generating a test instruction corresponding to the test step by the AI model according to the content described by the natural language and the atomic instruction set; obtaining a test instruction generated by the AI model, and displaying the test instruction through an instruction display area so as to verify the integrity and the execution sequence of the test instruction corresponding to a plurality of test steps split by the target iteration function; and performing instruction combination on the plurality of test instructions to obtain a corresponding test case so as to execute the test case to perform automatic test on the target iteration function.

The present invention provides a method for predicting impeller flutter. The method is as follows: using the finite element analysis method to mesh the solid model of the impeller, dividing the impeller processing process into several processing sub-stages, and using the finite element analysis method to solve the problem. The modal parameters of each machining sub-stage of the impeller are obtained, and the transfer function of each machining sub-stage of the impeller is established. The experimental modal analysis method is used to obtain the modal parameters of the tool system and the modal parameters of the impeller system, and the transfer function of the tool-impeller system is established. ; Calculate the difference between the modal parameters of each processing sub-stage of the impeller and the modal parameters of the impeller system obtained by the experimental modal analysis method, obtain the modal parameter deviation vector, and use the modal parameter iteration function to calculate the modal parameter deviation vector Through iteration, the final modal parameters of each processing sub-stage with the modal parameter deviation vector within the modal parameter tolerance range are obtained, and the milling stability lobe diagram of each processing sub-stage of the impeller is drawn.

A star flash encryption transmission method, apparatus and device, and a storage medium are applied to the technical field of data security, and the method applied to a first device comprises the steps of initializing a population of evolutionary computation; performing iterative optimization of a preset number of iterations on the population, and obtaining a first password set of a nonlinear iteration function value corresponding to each individual in the population of each iteration and an exponential first population fitness value corresponding to the population; according to a first password set corresponding to the first maximum population fitness value and a preset encryption function, encrypting the to-be-encrypted plaintext to obtain a ciphertext; sending the initialized population and the first maximum population fitness value to at least one second device for security verification; and if response information which is fed back by the target second equipment and indicates that the verification is passed is received, transmitting the ciphertext to the target second equipment. Data encryption is carried out by adopting evolutionary computation, so that the security and the quantum attack resistance of the encryption process are enhanced; and meanwhile, the adaptability of the encryption process is improved.

The invention discloses a control parameter optimization method for improving the phase commutating failure restraining capacity of a direct current system. The method comprises the following steps that after initialization, through an interface technology between PSCAD and MATLAB, a first interface module and a second interface module are established, the first interface module is used for calling an optimization algorithm, a whole iteration function of the algorithm is carried out, when each-time simulation is started, through an improved particle swarm algorithm, particle positions are updated, and controller parameters are returned to the PSCAD; and the second interface module is used for receiving the particle positions returned by the PSCAD and optimizing target functions after each-time simulation is over, and the global optimum particle positions and the individual optimum positions are updated. Through the interface technology between the PSCAD and the MATLAB, combining calling of the PSCAD and the MATLAB is achieved, the improved particle swarm algorithm is used for searching for the optimum control parameters, a multi-time operation function in the PSCAD is used, calling of multi-time software can be achieved, and accordingly optimization searching of control parameters is achieved.

This invention provides a method and system for adaptive iteration count of a decoder, comprising: activating an adaptive iteration function; initializing the decoder and setting the decoder's iteration count to a maximum iteration count; calculating the average MCS value and average signal-to-noise ratio of the physical downlink shared channel within a sliding window; adjusting the decoder's iteration count as the initial iteration count based on the average MCS value and average signal-to-noise ratio; decoding a cell block and recording the actual decoding count used when decoding is successful; calculating the average of the actual decoding count used when the current cell block is successfully decoded and the actual decoding count used when previous cell blocks are successfully decoded within the sliding window; adjusting the current iteration count of the decoder based on the average decoding count, and decoding the next cell block. By dynamically adjusting the maximum iteration count of the LDPC decoder, unnecessary decoding iterations are reduced, UE power consumption is lowered, and a balance between performance and power consumption is achieved.

The invention discloses a radome parameterization and strength rapid analysis simulation system, which comprises a model input module, a material physical property parameter module, a grid division module, a load loading module, a boundary condition setting module, a solving module, an optimization module and a post-processing module. Above eight modules meet corresponding function requirements ofthe modules through independent plug-ins. Therefore, the invention provides simulation software which is designed on the basis of secondary development of commercial software ABAQUS and can realize arapid design iteration function of the structural strength of the radome, and through parametric modeling, the design efficiency of technicians can be greatly improved while the design reliability isensured, and an optimal scheme is selected from multiple design schemes.

The invention discloses a parallel-beam CT sparse angle reconstruction method based on non-standard fast Fourier transformation and alternating direction method, which overcomes the problem that the image reconstruction method still has defects in the prior art. The steps of the invention are as follows: perform one-dimensional FFT transformation on the projection data collected by parallel beam CT to obtain the corresponding projection frequency domain data in polar coordinates; based on the projection frequency domain based on polar coordinates, use NUFFT technology to realize image space Frequency domain transformation to avoid the loss of precision caused by frequency domain interpolation, and establish a TV minimization image reconstruction model; in the process of restoring the image to be reconstructed from the projected frequency domain, a frequency domain optimization model based on TV minimization is designed, using alternate The direction method is used to iteratively solve the TV minimization model through the augmented Lagrangian function method and the alternating direction multiplier method. The invention combines the NUFFT technology and the advanced alternating direction idea in the optimization strategy, can avoid interpolation in the frequency domain, requires less computing and storage resources, and has good convergence performance.