123 results about "Sparse structure" patented technology

A method for sparse channel estimation in MIMO OFDM systems with a plurality of subchannels having the same sparsity structure is presented. The inventive method comprises initializing a plurality of residual vectors and observation generating matrices modeling the channel, sending a pilot signal for each subcarrier, converting the pilot signals to tap positions, detecting an optimal tap position, updating the residual vectors by removing the one residual vector having the optimal tap position, updating the generating matrices in accordance with the optimal residual vector, calculating weighted residuals based on the updated residual vectors, and repeating the steps, except initializing, until a stopping condition is met, wherein the updated observation matrices estimate the sparse channel. In one embodiment, the observation generating matrices are omitted. In one embodiment, multiple vectors are removed during one iteration. Pilot placement and pilot allocation techniques are presented to optimize the method.

The invention relates to a method for preparing a nanofiber membrane through electrostatic spinning. High-molecular polymer is dissolved in solvent to be stirred till an even solution is obtained; the spinning solution is used for electrostatic spinning, and the nanofiber membrane of a sparse structure can be obtained. The preparing method is convenient and fast to conduct, due to the fact that the obtained nanofiber membrane is of the unique sparse structure, the filtering resistance of the nanofiber membrane is lowered greatly, the influence on the filtering efficiency of the nanofiber membrane is very low and the method has very good application prospects in the filtering field, especially in the air filtering field.

The invention discloses a convolutional neural network based ribbon edge burr defect detection method. A camera collects a ribbon picture, the edge is extracted from a ribbon, and a sample picture with a burr defect and a sample picture without the burr defect are obtained; and the collected sample pictures are detected in a classified way via the convolutional neural network with a multiscale parallel training structure, the convolutional neural network increase the depth and width of the neural network, a full connection layer is removed from a common convolutional neural network, common convolution is converted into sparse connection, and a quasi-optimal local sparse structure via dense components is used to maintain a high computing performance of the neural network. Thus, the blur defect detection method can be used to detect blur defects of the ribbon edge effectively, maintain or reduce the computational complexity of the convolutional neural network, and thus, improve the computing performance.

The invention discloses a generalized tree sparse-based weight nuclear norm magnetic resonance imaging reconstruction method. The method comprises the following steps of: firstly obtaining a test magnetic resonance imaging sampling data sample to carry out Fourier transform; constructing a sparse of a tree structure according to a sampled signal, and approaching to sparse expression of a constrained target function by utilizing a nuclear norm with a weight; optimizing the constrained target function through an augmented lagrangian multiplier method, and carrying out iterative updating on the test data through an alternating direction search algorithm until estimated recovery data is obtained; and obtaining a final recovery image through constructing tree sparse inverse transformation. According to the method, an internal structure relationship between image signals is sufficiently mined, the generalized tree sparse structure characteristics of image blocks are combined with weight nuclear norms, and the calculation process is simplified by utilizing an ADMM algorithm, so that the complexity of the algorithm is reduced, the performance of a part of spatial data reconstruction images is improved, images can be reconstructed more accurately under less scanning and measurement, fake shadows of the reconstructed images can be decreased and rapid magnetic resonance imaging is realized.

The invention discloses an underwater acoustic target radiation noise modulation spectrum reconstruction method based on the group sparse structure. The method comprises the following steps that step1, a continuous spectral component R<c> (t) and a linear spectral component R<l> (t) of underwater acoustic target radiation noise are simulated to form underwater acoustic target radiation noise R (t); step 2, the R (t) is subjected to amplitude modulation to obtain a modulating signal x (t); step 3, the x (t) is subjected to normalization, and a band-pass filter is used for obtaining noise modulation signals y<l> (t) of L frequency bands; step 4, amplitude modulation data of the y<l> (t) on each sub-band are estimated, wherein the formula (please see the specification for the formula); step5, a formula (please see the specification for the formula) are subjected to discrete sampling and are represented by a sparse frequency coefficient, wherein the formula (please see the specificationfor the formula), and through designing of prior distribution of a formula (please see the specification for the formula), a high resolution modulation spectrum generation model based on the group sparse structure is constructed; step 6, posteriori distribution of a formula (please see the specification for the formula) is derived based on the expectation maximization method; and step 7, a parameter estimation formula is used for an iterative solution of a formula (please see the specification for the formula) to estimate a high resolution sparse modulation spectrum. Correlation of modulationspectrum positions of sub-bands of underwater acoustic target radiation noise is used for achieving high resolution reconstruction of the underwater acoustic target radiation noise modulation spectrum.

The invention relates to a personal value calculation method and system based on big data. The method comprises steps of extracting person data from a lot of resumes; grading the extracted data according to a grading system; establishing a sparse structure penalty function with an organization structure prior, putting graded fields into a function mode and selecting a field; using the graded fields, establishing a regression model taking salary prediction as an object, selected fields as dependent variables and a revised expect salary as an independent variable , and calculating the coefficient of each field; and obtaining a new personal resume, and calculating the personal value corresponding to the personal resume according to the obtained coefficients. The resume in the table format in the prior art is technically processed, resume elements with real values are extracted from complicated written description and are embodied in a graphical data chart, so that all qualities of a person are clear at a glance. The invention accurately evaluate a person through standard variable calculation.

The invention discloses a multisource signal collaborative compressed sensing data recovery method comprising the steps as follows: 1, an aggregation node in a wireless sensor network acquires historical data of each sensor node, and computes a sparse structure information matrix by using the historical data and a first optimization equation; 2, the aggregation node receives a compressed sensing measurement result matrix that is to be processed and is transmitted by each sensor node; and 3, the aggregation node performs a data recovery operation on the compressed sensing measurement result matrix to be processed of each sensor node by using the sparse structure information matrix computed in the step 1, wherein the recovery data is a recovery result corresponding to current data to be transmitted of each sensor node after compressed sending measurement. According to the method provided by the invention, the sparse structure information matrix is trained by the compressed sensing measurement result matrix of the historical data, and the training process considers information loss during the compressed sensing measurement process, and thus the subsequent data recovery accuracy is improved.

The invention discloses an image quality evaluation method based on a sparse structure. The method is used for solving the technical problem that the evaluation effect of an existing image quality evaluation method is poor. According to the technical scheme, firstly, input reference images and input degraded images are sampled to obtain a reference image sampling matrix and a degraded image sampling matrix; then, a dictionary is obtained in a studying mode through the reference image sampling matrix, and in the process of working out a sparse solution, sparse representation is carried out on the reference image sampling matrix and the degraded image sampling matrix through the dictionary obtained in the studying mode to obtain a reference image sparse representation coefficient matrix and a degraded image sparse representation coefficient matrix; finally, the image quality is evaluated according to the change degree of the sparse coefficient structure. According to the method, by the adoption of the sparse structure in image quality evaluation, the image quality can be better evaluated. In addition, calculation is simpler, and the robustness is higher because the amplitude of a specific sparse representation coefficient is not involved.

The invention discloses a hybrid Gaussian spectrum sensing method based on block sparsity characteristics. The method comprises the following steps: 1) constructing a block sparse spectrum sensing system model influenced by hybrid Gaussian noise; 2) exploring a block sparse structure of a main user power spectrum signal according to the block sparse spectrum sensing system model; 3) reconstructingthe main user power spectrum signal by using the block sparse structure of the main user power spectrum signal, thereby judging whether the channel is occupied or not. The hybrid Gaussian spectrum sensing method disclosed by the invention is based on a virtual reference grid point scheme, and the block sparsity characteristics of the main user power spectrum signal are explored by importing a prior variable capable of controlling the block sparse structure so as to reconstruct the power spectrum information of the main user signal. In addition, in the hybrid Gaussian spectrum sensing method disclosed by the invention, the influence of the hybrid Gaussian noise on a cognitive radio system is considered, and the main user power spectrum signal can be reconstructed effectively without knowing the information of the block sparse structure of the main user power spectrum signal in advance.

The invention discloses a high-standard agroforestry construction method suitable for a plain area. The high-standard agroforestry construction method comprises forest network structural configuration and forest belt mode configuration, wherein a forest belt direction is integrally planned according to fields, forests, roads and water bodies, and an east-west rectangular grid is formed according to the directions of ditches, trenches, roads and dykes; the row number of the forest belt is 2-4, the plant line space is 2-4m*2-5m, the gap of the main forest belt is 200-350m, the forest height of trees of the main forest belt is 15+/-1m, and the gap of an auxiliary forest belt is 500-600m; a sparse structure with the sparseness degree of 0.15-0.45 is formed by using a planting manner in triangle form, mixed planting of arbors and shrubs and tending management; and the ratio of the mixed planting of arbors and shrubs is 2:1-4:1. Through reasonable forest belt mode configuration and forestry network structural configuration, a high-standard agroforestry constructed by the invention has desired sparseness or ventilation structure, and can not only reduce land occupation of the forest belts effectively but also improve the protective efficiency of the agroforestry in the plain area.

The invention relates to a three-phase distribution network topology identification method based on AMI measurement neighbor regression. The method is characterized by comprising the following steps of 1, establishing a model; 2, solving the model; and 3, inspecting suspicious line. The difference between the voltage amplitudes of adjacent nodes at adjacent moments is regarded as Gaussian random variable. A precision matrix estimation model of a Gaussian Markov random field composed of the respective random variables is established and is solved by using a neighbor regression algorithm. The three-phase distribution network topology corresponding to the Gaussian Markov random field is identified based on the sparseness of the estimated precision matrix. The conditional independence test iscarried out on the random variables of the nodes at both ends of the suspicious line to further verify whether the suspicious line really has a connection relationship. The method is simple in structure and good in practicability.

The invention relates to the technical field of infrared spectrum wavelength, and concretely relates to a convex optimization theory-based novel infrared spectrum wavelength selection method. The method is a self-adaptive wavelength selection method adopting a Group Lasso technology without knowing spectrum partitioning priori knowledge. An infrared spectrum wavelength screening problem is converted into a Group Lasso sparse optimization problem, the sparse structure priori knowledge of the infrared spectrum is fully used to self-adaptively determine the partitioning size, a Shooting rapid algorithm is adopted to calculate the sparse solution, and Belsley colinearity examination of the sparse solution is carried out on the sparse solution to reject wavelength points with small contribution. The method has the advantages of small calculation capacity, few adjustable parameters and strong robustness, can effectively reduce the complexity of a model and improve the generalization performance of the model, and can be widely used in the field of infrared spectrum wavelength selection of a solid phase, a liquid phase and a gas phase.

The invention discloses a designated target three-dimensional reconstruction method and system. The method comprises the steps of obtaining a multi-view image; determining an appointed target image needing to be reconstructed; performing scene recovery on the multi-view image by adopting SFM to obtain a scene sparse structure of the multi-view image, and marking a part of 3D points in the scene sparse structure of the multi-view image to obtain a training sample set; constructing a strong classifier by adopting an AdaBoost algorithm based on the training sample set; classifying the rest 3D points in the scene sparse structure by adopting a strong classifier to obtain a sparse point cloud structure of a specified target; determining a dense point cloud structure of the scene of the multi-view image by adopting an MVS based on the sparse point cloud structure; and based on the dense point cloud structure, taking the training sample set as training data, and determining a specified targetdense point cloud structure by adopting a multi-decision tree judgment strategy. According to the method, three-dimensional reconstruction of the designated target can be achieved only through a small number of sparse points under one viewpoint, the calculated amount is small, and precision is high.

The invention discloses a movement noise detection method suitable for heart rate signals. The method aims at improving the heart rate measurement accuracy of a wearable heart rate measurement device. In the method, the wearable heart rate measurement device collects multiple pulse oximeter signals and movement acceleration signals, during the same time period, of a user; a frequency spectrum matrix is formed by the pulse oximeter signals and the movement acceleration signals, and a combined sparse spectrum reconstitution model is established by extracting the overall sparse and line sparse structure characteristics in the frequency spectrum matrix; then, a sparse frequency spectrum matrix in the combined sparse reconstitution model is calculated through the inaccurate augmented lagrangian multiplier method, and calculated signals has the advantage that the spectrum peak positions of movement acceleration signal frequency spectrums are basically the same as the spectrum peak positions of the pulse oximeter signals. By means of the method, strong movement noise in heart rate signals can be accurately detected, and the theoretical foundation is laid for effectively removing strong movement noise in heart rate signals.

The invention belongs to the technical field of compressed sensing of a multi-measurement vector model, and specifically relates to a parallel reconstruction method for a joint sparse vector based ona convolutional deep stacking network. Inspired by the successful application of a deep learning technology to a pattern classification problem, the parallel reconstruction method acquires a joint sparse structure of channel signals by using the convolutional deep stacking network, converts an atom selection problem into an atom classification problem and selects a plurality of candidate atoms ateach iteration, thereby effectively solving a problem of signal reconstruction under the multi-measurement vector model, and keeping low complexity of the algorithm.

The invention discloses a power system transient stability simulation method based on network node numbering optimization. A sparse vector technology is widely applied to power system calculation. However, existing network node numbering methods applied to sparse vectors aim to achieve the purpose that the average forward and backward substitution calculation amount of all nodes is the smallest, and the characteristic that forward and backward substitution only needs to be conducted on active nodes during transient stability simulation is not taken into consideration. According to the power system transient stability simulation method, the characteristic that the sparse structure of an independent vector of a network algebraic equation and the sparser structure of a solution vector are identical and decided during transient stability simulation is fully considered, the network nodes are divided into active nodes and passive nodes, the out-degrees of the nodes, the number of active precursor nodes and the number of the precursor nodes are fully considered, the network nodes are numbered, under the condition that the number of newly increased elements of a factor table is small, the average path length of the active nodes is made the smallest, and no requirement for the path tree lengths of the passive nodes exists. By the adoption of the power system transient stability simulation method, the calculation amount of solving a differential algebra equation set during transient stability simulation of a power system can be remarkably reduced.

The invention relates to the technical field of human bioengineering, in particular to a tissue engineering bracket with partitions, the tissue engineering bracket comprises a bracket body, wherein the bracket body is composed of a plurality of groups of parallel net strips with different directions between each group, and reinforcing edges for interweaving and connecting the net strips are arranged between the net strips. A sparse structure area and a compact structure area are regularly arranged on the bracket body, and a transition structure area is arranged between the sparse structure area and the compact structure area. According to the tissue engineering bracket with partitions, a plurality of groups of parallel net strips with different directions between each group are connected to form a model with a specific shape through a plurality of reinforcing edges which are interval distributed and are generally connected with each other to form a rhombic shape, and the model is emphatically partitioned to enable different parts of the model to have different densities and strengths; an auricle bracket is manufactured by adopting a 3D printing titanium alloy method, so that the artificial auricle bracket is closer to the structure of a real auricle; the preparation is made for improving the conventional small-ear deformity auricle reconstruction operation, rib cartilage does not need to be intercepted through thoracic operation, and the pain of patients is relieved.