42results about How to "Reduce approximation error" patented technology

Disclosed is an implementation method for simplifying a complicated message-passing function in a decoder for decoding block codes encoded with low-density parity-check (LDPC) codes and only using a summator and a shifter to simplify the hardware structure of the decoder, in which method the input interval of the message-passing function for binary representation of a message input is divided and the respective divided intervals are linearized to allow the calculation of the output of the message-passing function without using a memory. Based on the fact that the message-passing function is similar in structure to an exponential function, the linearized intervals are set to make the maximum value expressible in each digit of the binary representation as the boundary of the intervals.

The invention belongs to the field of array signal processing and mainly relates to robustness of a standard Capon self-adaptive beam forming algorithm based on covariance matrix reconstruction to interference signal steering vector errors. The beam forming algorithm based on subspace interference-plus-noise covariance matrix reconstruction comprises the following steps: at first, utilizing matrix received data to estimate the steering vector a(theta d), d=2,3,...,D and the power sigma<2> d(theta d) of all D-1 interference signals and meanwhile estimate the noise power sigma<2>, wherein d=2,3,...,D; then reconstructing an interference-plus-noise covariance matrix R' according to the definition R of the interference-plus-noise covariance matrix; finally, constructing a signal covariance matrix in a relatively small angle range 1; taking the dominant eigenvector of the signal covariance matrix as a desired signal steering vector to estimate a(theta 1); together with the reconstructed R', obtaining a novel beam forming weighing vector W, wherein the formulas of R, R' and W are shown in the description. The beam forming method provided by the invention overcomes the defects of the conventional beam forming algorithm, thereby having good robustness to interference signal steering vector errors.

The invention relates to a program segment smooth compression processing method suitable for a numerical control device, comprising the following steps: (1) analysis of a processing path, i.e. removing irregular programming points in a numerical control program by filtration, and deducing parts required to conduct smoothness in shape; (2) parameterization of the programming points, i.e. carrying out parameterization on each programming point by means of distances among the programming points; (3) selection of characteristic programming points, i.e. dividing the programming points into the characteristic programming points and non-characteristic programming points by means of processing shape and bending direction at the programming points; (4) calculation of tangent vector at the characteristic programming points, i.e. calculating the tangent vector at the characteristic programming points by constructing an interpolation curve; (5) compression of program segments, i.e. compressing the program segments between the adjacent programming points into a segment of a spline curve; and (6) control of mismachining tolerance, i.e. ensuring the compressed spline curve to meet requirements of machining accuracy by adjusting the shape of the curve segment. The method in the invention can prevent irregularity of a workpiece surface from being caused by line interpolation at transition of the programming segment and has the advantages of high machining accuracy and high efficiency.

The invention discloses an MIMO (Multiple Input and Multiple Output) radar single measurement vector DOA (Direction of Arrival) estimation method based on iterative weighted near-end projection. The method comprises the following steps of: vectorizing a covariance matrix of received data after dimensionality reduction; constructing a weighting matrix by using high-order power of a covariance inverse matrix after dimensionality reduction so as to perform proper weight constraint on a sparse vector; establishing a weighted near-end function optimization model to represent a non-convex and non-smooth sparse optimization problem in MIMO radar single measurement vector DOA estimation; and finally, obtaining a near-end operator through an SCAD (Smoothly Clipped Absolute Deviation Penalty) function in an iteration process, and projecting the near-end operator to a feasible set to solve the weighted function optimization model so as to obtain a sparse solution, and obtaining a real target DOA estimation value by searching the position of a spectral peak. Compared with a reweighted l1-SVD algorithm and a weighted SL0 (Smoothed l0norm) algorithm, the method can obtain the better DOA estimation performance, and the prior information of the number of the targets is not needed to be known in advance.

Systems and methods for generating check node updates in the decoding of low-density parity-check (LDPC) codes use new approximations in order to reduce the complexity of implementing a LDPC decoder, while maintaining accuracy. The new approximations approximate the standard sum-product algorithm (SPA), and can reduce the approximation error of min-sum algorithm (MSA) and have almost the same performance as sum-product algorithm (SPA) under both floating precision operation and fixed-point operation.

The invention belongs to the array signal processing field, and mainly relates to robustness of a standard Capon adaptive beamforming algorithm, based on covariance matrix reconstruction, against interfering signal incoming wave direction error; the interfering noise covariance matrix reconstruction method aiming at incoming wave direction error comprises the following steps: firstly using array reception data to estimate guide vectors (see formula, wherein d=1, 2..D ) of all D signals, to estimate D-1 interfering signal power (see formula, wherein d=2,3.. D), and to simultaneously estimate a noise power (see formula); then reconstructing the interfering noise covariance matrix (see formula) according to the definition of the interfering noise covariance matrix (see formula); combining the estimated guide vectors with the reconstructed interfering noise covariance matrix so as to obtain a novel beamforming weighting vector (see formula). The method can improve estimation precision of the interfering noise covariance matrix, can effectively reduce desired signal elements, can greatly weaken or prevent desired signal self-elimination phenomenon, thus greatly improving output SINR.

The invention provides a propagation time measurement method for direct signals based on autonomous underwater vehicles. The invention comprises the steps of: determining a set of measurement time ofdirect signal propagation point in a known task area with to be measured sound speed; placing a sending underwater autonomous vehicle and a receiving underwater autonomous vehicle at the known task area with to be measured sound speed to measure propagation time of direct sound signals through a single-input-multiple-output double direction interaction communication mode; according to the propagation time of direct sound signals, acquiring a sound speed profile by adopting a sound speed profile inversion algorithm which based on a matched field processing technology; and carrying out measurement and sound speed profile inversion on each latitude and longitude coordinate point to construct a sound speed profile set according to the sound speed profile inversion algorithm which based on thematched field processing technology. The method has the advantages that influence of boundary parameter mismatch is reduced, and an approximate error of representing sound speed distribution of the whole area is avoided.

The invention provides an image self-adaptive grid generation variational method, and relates to image approximation and piecewise polynomial approximation. The method comprises: S1, inputting an image, and setting related parameters; S2, generating original triangular mesh partition; S3, according to an energy function and related gradient and Hessian matrix information, calculating a new positions of vertexes of triangular meshes, and moving vertexes to the new positions; S4, updating connection relations of the triangular meshes whose vertexes are moved; S5, executing the step S3 to S4 for a plurality of times in a circulating manner, until iterations reach J, and triangular grids whose partition structure is very close to an original image being generated in an image region; and outputting optimal triangular grids and a corresponding approximating polynomial set. Using a piecewise polynomial fitting method combining with a Newton iteration optimization method, triangulation is distributed along image characteristic lines in a self-adaptive manner. Polynomial approximation is used to obtain good visual and numerical effect of original image approximation on triangulation. The method can be used for image approximation and image vectorization.

The invention discloses a three-dimensional grid model simplification method and system method and system based on optimized feature preservation. The three-dimensional grid model simplification method includes the steps: firstly, calculating the Gaussian curvature of each edge in a three-dimensional grid, combing the Gaussian curvature and the secondary error measure of the edge to calculate thefolding cost of the edge, and constructing edge folding operation based on feature preservation; combining an edge splitting operation sequence and an edge folding operation sequence based on featurepreservation to generate a mixed sequence, and using a mixing mechanism for simplifying a three-dimensional grid model; replacing a global search stage in whale optimization by crossover and mutationoperation of differential evolution, and constructing a whale differential evolution optimization algorithm; and finally, searching an optimal edge splitting operation and edge folding operation sequence combination mode by utilizing an optimization algorithm, so that the approximate error between the simplified three-dimensional model and the original model is minimum. According to the three-dimensional grid model simplification method, through grid simplification, the optimal three-dimensional grid simplification effect with the minimum approximate error can be obtained, meanwhile, geometrical characteristics can be well kept, and the quality of triangles in the model is improved.

The invention relates to an adaptive wide-area damping controller. The adaptive wide-area damping controller comprises an adaptive delay compensator and an N input and output GrHDP unit; the adaptivedelay compensator is used for obtaining N first wide-area measurement signals, performing adaptive delay compensation and outputting N second wide-area measurement signals; the N input and output GrHDP unit is used for receiving the N second wide-area measurement signals and obtaining N first coordinated control signals based on an adaptive dynamic programming algorithm; and the adaptive delay compensator is also used for receiving the N first coordinated control signals and obtaining N second coordinated control signals by calculation based on a GrHDP model algorithm. According to the adaptive wide-area damping controller disclosed by the invention, multiple input and multiple output are carried out by adoption of the GrHDP unit; multi-target coordinated adaptive control on multiple low-frequency oscillation modes of a power system can be simultaneously carried out; the adaptive capability of the power system is improved; furthermore, both the first wide-area measurement signals and the first coordinated control signals are subjected to delay compensation; the control influence of delay onto the power system is greatly avoided; and the damping improvement of the controller onto the power system is improved.

The invention discloses a thermal power plant time sequence variable abnormity monitoring method and system. The method comprises the steps of collecting time sequence data of thermal power plant analog signal monitoring variables; taking the whole time sequence data as an initial optimal PLR segment, calculating a decision coefficient of the optimal PLR segment, and taking the whole time sequencein the historical sample data as a first-layer original data segment; segmenting the original data segment to form two new data segments, and respectively calculating decision coefficients of the twonew data segments; judging whether the segments are allowed or not, and if yes, continuing to segment the segments; and judging whether the segmentation is finished or not to obtain the optimal PLR segmentation number. According to the method, the optimal segment number of the PLR is determined through a coefficient determination method, and the problems that an L method highly depends on a fitting error and left and right points in a time sequence are solved.

Implementations provide an efficient system for calculating inner products between high-dimensionality vectors. An example method includes clustering database items represented as vectors, selecting acluster center for each cluster, and storing the cluster center as an entry in a first layer codebook. The method also includes, for each database item, calculating a residual based on the cluster center for the cluster the database item is assigned to and projecting the residual into subspaces. The method also includes determining, for each of the subspaces, an entry in a second layer codebook for the subspace, and storing the entry in the first layer codebook and the respective entry in the second layer codebook for each of the subspaces as a quantized vector for the database item. The entry can be used to categorize an item represented by a query vector or to provide database items responsive to a query vector.

The invention discloses a random multiple extreme value response surface method for reliability analysis of a flexible component. The method comprises the following specific steps: comprehensively considering the combined action of the nonlinear material attribute and the mechanical load of the flexible component, and finding out the maximum output response point of the blade through static analysis; selecting an input random variable, extracting an input random variable sample by applying a Latin hypercube sampling (LHS) technology, and solving a finite element basic equation for each sampleto obtain a corresponding output response; constructing a random multiple extreme value response surface function (SMERSF), completing reliability analysis of the flexible component; and carrying outvalidity verification on the SMERSM. The reliability analysis method for the multi-failure-mode structure is rapid and efficient.

A three-dimensional spot sample data reduction method based on Gaussian sphere comprises the following steps: 1) defining L2,1 shape measurement on three-dimensional sampling point data as Euclidean distance between sampling point normal vector field Gaussian mapping images, and carrying out regular triangulation for the Gaussian sphere; 2) taking discrete surface in a large scale as input data comprising position of sampling point and normal information, and determining the neighborhood of each sampling point according to the normal deviation of the sampling point; 3) initially clustering through index diffusion technique and stack data structure; 4) separating the irregular cluster for generating regular cluster, and absorbing the isolated sampling point of the normal vector projection on the top point of the Gaussian triangle into the adjacent cluster with the minimum deviation normal vector. The invention simplifies the calculation, reduces the calculation time, and reduces the approximate error.

Systems and methods for generating check node updates in the decoding of low-density parity-check (LDPC) codes use new approximations in order to reduce the complexity of implementing a LDPC decoder, while maintaining accuracy. The new approximations approximate the standard float-point sum-product algorithm (SPA), and can reduce the approximation error of min-sum algorithm (MSA) and have almost the same performance under 5 bits fix-point realization as the float-point sum-product algorithm (SPA).

The invention provides a method for inhibiting the velocity fluctuation of an electric vehicle driving system. The method comprises the following steps: obtaining a rotor position by adopting an encoder, and transmitting the rotor position to a velocity control ring; and adopting a parallel mode combining a prediction PI joint controller, a repetitive controller and a nonlinear adaptive feedback observer by the velocity control ring, adopting the parallel mode of the prediction PI joint controller and the repetitive controller during electric start, substituting two continuous sampling values into a prediction domain of the prediction PI joint controller by the encoder, switching to the parallel mode of the repetitive controller and the nonlinear adaptive feedback observer when an electric vehicle enters a stable operation mode, and then transmitting an output signal of the velocity control ring. According to the method provided by the invention, the prediction PI joint control, the nonlinear adaptive feedback observation control, the repetitive control and the vector control are combined together to improve the stability, the accuracy and the dynamic response ability of the electric vehicle driving system and inhibit the velocity fluctuation in an electric vehicle running process.

A technique for correcting the voice spectral deformations introduced by a communication network. Prior to the operation of equalization of the voice signal of a speaker, the constitution of classes of speakers is communicated, with one voice reference per class. Then, for a given speaker, the classification of this speaker is communicated, that is to say his allocation to a class from predefined classification criteria in order to make a voice reference which is closest to his own correspond to him. Then, for that given speaker, communicating the equalization of the digitized signal of the voice of the speaker carried out with, as a reference spectrum, the voice reference of the class to which the speaker has been allocated. This technique applies to the correction of the timbre of the voice in switched telephone networks, in ISDN networks and in mobile networks.

The invention discloses a geometric spline-based linear projectile model modeling solving method, which comprises the following steps of: expressing a boundary of a linear projectile problem as a geometric spline of a linear combination of a control vertex and a primary function with geometric significance by using parameters, and finishing modeling; taking a geometric spline primary function withthe same boundary representation model as a shape function of displacement and surface force in a boundary integral equation of the linear projectile problem, obtaining a linear equation set throughthe discrete boundary integral equation, building a stiffness matrix and solving to obtain the displacement and the surface force of a boundary control vertex; according to the method, the physical domain boundary of the 2D line projectile problem is modeled through the geometric spline, an accurate conversion method represented by a common model is provided, and the approximation error of the common boundary model in the modeling stage is avoided; a geometric spline basis function is adopted in a boundary integral equation to serve as a shape function for describing boundary physical quantity, and the model conversion time is saved.

The invention belongs to the field of gear manufacture, and relates to a multi-tool approach method for processing a small-size addendum circle angle by a worm grinding wheel. The problem that the worm grinding wheel can not be trimmed since a radius of curvature is too small when the small-size addendum circle angle is processed. Through the translation of the X shaft and the Y shaft of the wormgrinding wheel, multiple sections of curves are processed, a circle angle curve is enveloped by the multiple sections of curves, a point vector pairing method is put forward, the offset of the X shaftand the Y shaft of the grinding wheel is solved, the principle errors of the approach method are analyzed, and an optimal approach circular arc size and approach tool number can be established according to error upper limits required during design.

The invention discloses a medical PET image denoising method based on frequency domain direction smoothing Shearlet. A new medical PET image Gaussian noise model is provided, then frequency domain multi-scale decomposition and multi-smooth direction decomposition are carried out, new unified threshold processing is carried out on the decomposed high-frequency direction smooth Shearlet coefficient,and then a denoised PET image is generated through inverse Shearlet transformation. Compared with a traditional NSST method (non-subsampled Shearlet transform), the method has the advantages that thedenoising effect is better, the speed is higher, and the method can be better applied to the field of medical PET image denoising.