256 results about "Eigenvalues and eigenvectors" patented technology

The invention relates to a comprehensive evaluation method of smart power grid construction based on principal component cluster analysis, which is technologically characterized by comprising the steps that at the step 1, a generally approved comprehensive evaluation index system of the smart power grid construction is established or selected; at the step 2, index data is processed by standardization; at the step 3, an index data correlation coefficient matrix is established, an eigenvalue and an eigenvector of the matrix are solved, and a principal component expression is generated; at the step 4, a principal component variance contribution rate and a cumulative variance contribution rate are calculated, and quantity of principal components is determined; at the step 5, a comprehensive principal component evaluation index function is established, and a comprehensive evaluation result of a development and construction level of a smart power grid is given; and at the step 6, a principal component factor load matrix is established, and the cluster analysis is carried out to comprehensive evaluation indexes of the smart power grid. The comprehensive evaluation method of the smart power grid construction based on the principal component cluster analysis provided by the invention combines principal component analysis and the cluster analysis in order to simplify and reconstruct the evaluation index system of the smart power grid construction and provides suggestions for the smart power grid construction is laggard areas.

The invention provides a document representation method and a document analysis method including extraction of important sentences from a given document and / or determination of similarity between two documents.The inventive method detects terms that occur in the input document, segments the input document into document segments, each segment being an appropriately sized chunk and generates document segment vectors, each vector including as its element values according to occurrence frequencies of the terms occurring in the document segments. The method further calculates eigenvalues and eigenvectors of a square sum matrix in which a rank of the respective document segment vector is represented by R and selects from the eigenvectors a plural (L) of eigenvectors to be used for determining the importance. Then, the method calculates a weighted sum of the squared projections of the respective document segment vectors onto the respective selected eigenvectors and selects document segments having the significant importance based on the calculated weighted sum of the squared projections of the respective document segment vectors.

The invention relates to an estimating method of the main curvature and the main direction of point cloud data, which comprises the following steps of: preprocessing, estimating a normal transversal curvature, fitting a Weingarten matrix, computing the characteristic value and the characteristic vector of the Weingarten matrix and computing the main curvature and the main direction. In the method, the main curvature and the main direction faithful to an original object are obtained only by utilizing the scanning data and the estimated normal vector of a laser scanner. The main curvature and the main direction are computed by the method through least square linear fitting and the characteristic value and the characteristic vector of the matrix, the algorithm is simple, the computing result is accurate, and the time complexity is highly effective. The method is called as a normal transversal fitting method, and the computing result thereof has important application value in the fields of virtual reality, computer games, natural scene simulation, urban landscape design, data compression, characteristic extraction, object 3D reconstruction and the like.

The invention provides a method for acquiring steady-state equivalent wind speed and generated power in wind power station based on correlation analysis, belonging to the technical field of wind powergeneration. The method comprises the following steps: preprocessing the operational data in the wind power station; acquiring the steady-state equivalent wind speed in the wind power station by a correlation analysis method; forming a wind speed matrix, calculating correlated matrix of all unit wind speed in the wind power station, acquiring a characteristic value and a characteristic vector of the correlated matrix to obtain the equivalent wind speed finally; and acquiring the generated power in the wind power station, and the like. The method can accurately acquire the equivalent wind speedand generated power in the wind power station, and has the characteristics of high precision, simple method, convenient operation, and the like. The method has wide application range, and can be usedfor equivalent modeling of the wind power station, confirmation of the maximum penetration power of the wind power station, technique and system for predicating the generated power in the wind powerstation, reliability and economy application of the wind power station, analysis of influence of a wind power access network to the power system, and the like, and has great application value for planning and designing a wind power station access system.

A method for reconstructing seismic data signals of poor quality to improve the signal-to-noise ratio of the data for display and analysis in connection with the selection of drilling sites for recovery of hydrocarbons. The method includes providing a signal model by applying a Karhunen-Loeve transform to selected input seismic data collected for the target zone, to form a co-variance matrix from the dot products of all pairs of input data. Eigenvalues and eigenvectors for the matrix are computed, and the most significant eigenvectors are inversely transformed to provide a coherent estimate of the signal. The input data is combined with the model data based on the determination that the model data lacks continuity, wherein the good quality signal-to-noise ratio data experiences little change and discontinuous data is enhanced by a contribution of the signal estimate data. The reconstructed seismic data of the target zone can be displayed for analysis.

Method and apparatus performing dynamic contrast-enhanced-magnetic resonance imaging on tissue to obtain a plurality of datasets of images. Principal component analysis is performed on each dataset to obtain a covariance matrix and its corresponding eigenvalues and eigenvectors and produce a common base of eigenvectors. The dominant eigenvectors that are not associated with instrumental and random noise, commonly the 2nd eigen-state and the 3rd eigenvectors, or the 1st and 2nd eigen vectors, are correlated with the physiological relevant parameters of the 3TP method to obtain a hybrid method. The fusion of the eigenvectors with the 3TP parameters is dictating a rotation of the two relevant eigenvectors to obtain new rotated eigenvectors that serve to calculate new projection coefficient maps of the rotated eigenvectors for the imaged tissue indicative of physiological relevant parameters reflecting wash-out and wash-in patterns that detect abnormal tissue and distinguishes between cancerous and benign tumors. Computer-readable medium containing program instructions for carrying out the above.

Te invention discloses a large-scale time-lag electric system characteristic value calculation method based on EIGD. The method includes the steps of establishing a time-lag electric system model, converting the formula of the time-lag electric system model into the abstract cauchy problem, converting the characteristic value of the time-lag electric system model into the characteristic value of an infinitesimal generator of the calculated and converted formula, conducting discretization on the infinitesimal generator to obtain an approximate matrix of the infinitesimal generator, obtaining the approximate matrix where displacement processing is conducted by conducting displacement processing on the approximate matrix, obtaining an inverse matrix after inverse conversion is conducted on the approximate matrix where displacement processing is conducted, converting the partial characteristic value required to be calculated into the partial characteristic value with the maximum module value, calculating the partial characteristic value, with the maximum module value, of the inverse matrix through the Arnoldi algorithm, and obtaining the characteristic value lambda of the approximate matrix. Verification is conducted through the Newton iteration method, and the accurate characteristic value and the accurate characteristic vector of a time-lag electric system are obtained through calculation.

An electronic scanning radar apparatus in accordance with an embodiment of the present invention, a frequency resolving unit resolves beat signals into beat frequencies having a predetermined frequency bandwidth and calculates complex data based on the resolved beat signals for each beat frequency. An azimuth calculating unit estimates a number of received waves based on eigenvalues of a matrix being part of a primary normal equation having complex data as elements calculated from the beat signals, creates coefficients calculated as a solution of a secondary normal equation of a signal subspace created based on eigenvalues and eigenvectors corresponding to the number of the estimated waves, and calculates a DOA of a received wave based on the created coefficients.

The invention is a DOA estimating method to realize circle-angle positioning based on TD-SCMDA standard, firstly adopting multiple signal classifying and correcting algorithms to realize the DOA estimation of intelligent antenna circle-angle positioning; aiming at TD-SCMDA antenna being uniform circle array, converting the array cell-space uniform circle array into mode-space virtual uniform line array; according to eigenvalue and eigenvector characteristic of the data covariance matrix of the virtual line array, advancing a MUSIC algorithm based on mode-space virtual uniform line array and deriving the MUSIC algorithm power chart for the virtual line array and the general implementation algorithm for the mode-space MUSIC algorithm; according to the character of MUSIC algorithm inapplicable to the DOA estimation of relative signal source, advancing the correcting MUSIC algorithm based on the mode-space virtual uniform line array; finally aiming at the problem that the array aperture is small and that the resistance to array cell error disturbance is bad, caused by MUSIC algorithm, advancing four-order cumulant MUSIC algorithm and four-order cumulant correcting MUSIC algorithm based on the mode-space virtual uniform line array.

The invention belongs to the technical field of mechanical analysis and numerical solution of a three-dimensional structure and relates to a contact analysis method in three-dimensional mechanical finite element model analysis. The method comprises the steps of firstly modeling a target electron device structure; introducing a displacement boundary condition or a stress boundary condition to establish a corresponding geometric structure model; generating a contact relation between parts according to a mutual relation between the parts; then subdividing the established geometric structure model by use of a tetrahedral mesh; generating a joint surface mesh according to a surface mesh in a simulated area contact surface; and finally establishing a finite element general intrinsic equation of a target electron device in consideration of a contact problem by use of a finite element method, solving the equation to obtain a characteristic value and a characteristic vector, and post-processing to obtain a vibration modal frequency and a vibration model. Therefore, the contact analysis in modal analysis is realized and a high-precision numerical calculation result is obtained.

Provided are a method and apparatus for multi-channel de-correlation processing for cancelling a multi-channel acoustic echo. The method includes: dividing an input multi-channel audio signal into units of frames to form multi-channel audio signals in units of frames; analyzing eigen values and eigen vectors related to the multi-channel audio signals by using the multi-channel audio signals in units of frames every time contents are modified; and separating the multi-channel audio signals in units of frames into a plurality of signal component spaces by using the analyzed eigen values and eigen vectors.

The invention provides a pavement crack detection method based on deep learning and principle component analysis. The method comprises the following steps: acquiring pavement images; performing gray-scale processing on the pavement images; then cutting each pavement image into W*V subimages each with a size of K*N pixels; manually selecting subimages comprising cracks and subimages not comprising cracks respectively as positive samples and negative samples which are taken together as a training set for training a convolutional neural network; establishing the convolutional neural network, training the convolutional neural network by use of the training set, detecting preprocessed pavement images to be analyzed by use of the trained convolutional neural network, and automatically extracting the subimages comprising the cracks; and performing crack type analysis, i.e., solving feature values and feature vectors by performing PCA calculation on distribution of the crack subimages, and determining crack types. According to the invention, automatic crack extraction is carried out on the pavement images by use of the convolutional neural network, pavement crack detection and crack type determination are realized, and a determining basis is provided for subsequent detection of the road crack types.

A fault detection method in a continuous annealing process based on a recursive kernel principal component analysis (RKPCA) is disclosed. The method includes: collecting data of the continuous annealing process including roll speed, current and tension of an entry loop (ELP); building a model using the RKPCA and updating the model, and calculating the eigenvectors {circumflex over (p)}. In the fault detection of the continuous annealing process, when the T2 statistic and SPE statistic are greater than their confidence limit, a fault is identified; on the contrary, the whole process is normal. The method mainly solves the nonlinear and time-varying problems of data, updates the model and calculates recursively the eigenvalues and eigenvectors of the training data covariance by the RKPCA. The results show that the method can not only greatly reduce false alarms, but also improve the accuracy of fault detection.

The invention discloses an intermediate-voltage distribution operation state fuzzy integrated evaluation method based on a principle component analysis method and an entropy weight method. Major information in source data can be reserved by use of the principle component analysis method, a power supply capability, power supply quality and economic indexes are taken as the source data representing an operation state of a distribution network, through solving a feature value and a feature vector of a correlation coefficient matrix, principle components of operation of the distribution network are determined, and an evaluation system with weak correlation yet large envelope information content is formed. Given fuzziness existing in index evaluation, new evaluation indexes are objectively endowed with weights by use of the entropy weight method in the system, and a fuzzy integrated evaluation system of a 10kV distribution network operation mode is finally formed. According to the invention, the fuzziness of the evaluation indexes is taken into consideration, qualitative analysis and quantitative determination are realized, and the method provided by the invention is reasonable and scientific to a certain degree.

Semiconductor manufacturing equipment includes: a controller controlling driving and processes of various parts of the semiconductor manufacturing equipment, and a sensor monitoring each physical amount in the semiconductor manufacturing equipment or a status of each chemical response amount; a database; and an arithmetic section executing: processing of reading out equipment data, calculating a correlation matrix between time points based on a plurality of pieces of signal data to be compared, calculating eigen values and eigen vectors from the correlation matrix, and calculating principal component scores by principal component analysis; processing of comparing magnitudes of the eigen values of the principal components, arranging the eigen values in descending order to display a list thereof; and processing of displaying a scatter diagram where the principal component scores of the respective signals are plotted in a feature space selecting the principal component corresponding to the eigen value having a contribution ratio.

The invention discloses an earthquake image structure guiding noise reduction method based on regularization mixed norm filtering. The earthquake image structure guiding noise reduction method includes the following steps that a gradient structure tensor is solved for an input three-dimensional earthquake image; regularization mixed norm filtering is conducted on the gradient structure tensor; a diffusion tensor is designed according to the eigenvalue and eigenvector of the filtered gradient structure tensor; continuity factors are calculated, the continuity factors at the position of a boundary fault feather edge and the like are close to zero, and the maintain performance of the structure is achieved; a sobel operator serves as a derivation operator so that divergence can be calculated. By means of the earthquake image structure guiding noise reduction method based on regularization mixed norm filtering, the textured edge information of earthquake-related data can be reserved, Gaussian noise, ultra Gaussian noise and sub Gaussian noise can be effectively suppressed, and therefore an efficient noise reduction method is achieved.

46. A measurement method of signal / noise ratio based on an array antenna mobile communication system: customer initial transmitted power, and get the customer initial transmitted power; control and valuate results by the fed-back power, and adjust the customer transmitted power, get the customer transmitted power of the frame; the transmitted power of the frame passing a radio transmission channel, antennas in a base station receive the customer signals, and convert them into baseband digital signals; estimate the customer channel, get the estimated values after the customer experiences the channel impulse response; based on the estimated values of the customer channel impulse response in each array element, estimate the covariance matrix; decompose eigenvalues, get the eigenvalues and eigenvectors of the customer, determine the average noise power and the signal power of the customer, measure the signal / noise ratio; compare the measured signal / noise ratio with the goal signal / noise ratio, get results about the controlled transmitted power, and feedback.

The invention provides a CPU+GPU heterogeneous parallel computing based natural frequency characteristic analysis method for a turbomachinery blade. The method comprises the steps of firstly, establishing a finite element model according to a three-dimensional model and material parameters of a to-be-analyzed turbomachinery blade, and performing pre-stress analysis on the blade; secondly, performing blade grid data preprocessing, simultaneously calculating a rigidity matrix and a mass matrix of each unit in a CPU and a GPU, and assembling the rigidity matrix and the mass matrix into total rigidity and mass matrixes; setting constraint conditions of the blade and a rim, wherein the setting process comprises constraint of rigid and elastic displacement of a boundary, contact coupling of a blade root and the rim or contact coupling of connecting pieces and correction of the total rigidity matrix; thirdly, extracting generalized eigenvalues and eigenvectors of the total rigidity and mass matrixes by using a CPU+GPU heterogeneous parallel algorithm; fourthly, converting the eigenvalues and eigenvectors into a frequency and a vibration mode of the blade, and performing output; and finally, judging a vibration type of a natural vibration mode, and drawing frequency curve distribution, a vibration safety graph or a Campbell chart of the blade according to the vibration type.

The present invention provides a ground PM2.5 concentration feature vector spatial filter value modeling method based on remote sensing data. The method comprises: obtaining data, selecting a model variable, carrying out data processing and matching, constructing a spatial adjacency matrix from the location of the national control point of a study area, carrying out centralization, calculating thematrix eigenvalues and eigenvectors, and extracting the appropriate eigenvectors from the vector group as the spatial influence factor of the PM2.5 concentration; and obtaining an eigenvector spatialfilter regression model of the PM2.5 concentration, interpolating the extracted eigenvectors raster images with the same spatial resolution as the AOD, and bringing the raster images into the eigenvector spatial filter regression model for raster calculation to obtain the continuous spatial distribution model of the PM2.5 concentration in the study area. According to the method provided by the present invention, for the problem that the number of ground control points is small and the ground control points are unevenly distributed, the remote sensing data with high resolution and continuous distribution is selected to perform the inversion of the ground PM2.5 concentration for study on a wide range of PM2.5 spatiotemporal features.

The invention relates to a method for calculating rotor dynamics performance of a multi-parallel-axis system. The method comprises the steps specifically as follows: step (1), selecting an analysis object and a calculation type, inputting parameters of a bearing-rotor system, finishing pre-processing operations like discretization of the rotor system; step (2), calling a bearing calculation program, calculating dynamic and static characteristic parameters of each bearing, forming a rotor dynamics equation of a parallel axis system which gives consideration to gear engagement according to the selected analysis object and calculation type; and step (3), calling a solving function to calculate a feature value and a feature vector of a dynamics differential equation, converting the feature value and the feature vector into dynamics characteristic result data through dimension so as to finish calculation and analysis of system stability, critical rotation speed, forced vibration response and vibration mode. According to the method disclosed by the invention, calculation and analysis means for dynamics performance of the multi-parallel-axis system, which can be applied to actual engineering, are developed on the basis of solid theoretical research; the calculated rotor system comprises supports of various types; therefore, combination of bearing calculation and rotor system dynamics analysis is implemented better.

The invention discloses a plant covering information extracting method based on perfect polarization SAR images. The obtained perfect polarization SAR images are preprocessed, so that standard SAR images are obtained, then coherence matrix computing is carried out, the eigenvalue and eigenvector thereof are obtained, accordingly, Cloude-pottier decomposition is carried out, therefore, polarization entropy and scattering angles are obtained, and then through H / alpha classification and Wishart H / alpha classification in sequence, scattering features of plant covering and covering information are identified and extracted. Through an H / alpha classification algorithm and a Wishart H / alpha classification algorithm, the plant covering information can be identified and extracted accurately, and the recognition error rate is greatly reduced.

The present invention discloses a low frequency oscillation identification method for accessing of wind power to a power grid in the technical field of wind power integration monitoring. The low frequency oscillation identification method comprises the steps that the system-wide linearized state matrix of a power system without a winder turbine generator is corrected, and a corrected system-wide linearized state matrix is obtained; eigenvalues and eigenvectors of the corrected system-wide linearized state matrix are figured out; the frequency, attenuation damping ratio and electromechanical circuit correlation ratio of the eigenvalues are worked out according to the eigenvalues and eigenvectors; and the eigenvalues are analyzed according to the frequency, attenuation damping ratio and the electromechanical circuit correlation ratio of the eigenvalues, and the low frequency oscillation mode of the accessing of the wind power to the power grid is determined. The low frequency oscillation identification method for the accessing of the wind power to the power grid solves the problems that an existing low frequency oscillation identification method does not take large-scale wind power accessing into consideration and cannot identity the system-leaded oscillation mode automatically.

Method of determining an islanding solution for a power system comprising representing synchronising coefficients between generators and couplings between load buses and generators of the power system as a dynamic graph, calculating eigenvalues and eigenvectors of a graph laplacian describing the synchronising coefficients, identifying reference generators, calculating eigenvectors describing dynamic coupling between load buses and reference generators, using the synchronising coefficient and dynamic coupling eigenvectors to determine proximities between reference generators and other buses, assigning other buses to reference generators according to the proximities, selecting a threshold, identifying buses in weak-areas by using the threshold to identify buses having a proximity to their reference generator sufficiently similar to their proximity to another reference generator, determining the islanding solution by selecting connections to disconnect to split the power system into two or more islands based on an analysis of power flows only on connections to, or between, buses in weak-areas.

The invention relates to a penicillin fermentation process failure monitoring method based on recursive kernel principal component analysis (RKPCA), which belongs to the technical field of failure monitoring and diagnosis. The method comprises the following steps: acquiring the ventilation rate, stirrer power, substrate feed rate, substrate feed temperature, generated heat quantity, concentrationof dissolved oxygen, pH value and concentration of carbon dioxide; and establishing an initial monitoring model by using the first N numbered standardized samples, updating the model by a RKPCA method, and computing the characteristic vectors to detect and diagnose the failure in the process of continuous annealing, wherein when the T2 statistics and SPE statistics exceed the respective control limit, judging that a failure exists, and otherwise, judging that the whole process is normal. The method mainly solves the problems of data nonlinearity and time variability; and the RKPCA method is used for updating the model by carrying out recursive computation on the characteristic values and characteristic vectors of the training data covariance. The result indicates that the method can greatly reduce the false alarm rate and enhance the failure detection accuracy.

The present invention relates to a drilling strain data anomaly extraction method based on principal component analysis. The method comprises: performing strain conversion of drilling strain data sequence of the same station, and performing preprocessing of the converted data; constructing a matrix through adoption of the drilling strain data after preprocessing; performing principal component analysis of the everyday matrix to obtain the feature value and the feature vector of each matrix; and corresponding the obtained feature value and the calculated feature vector angle and an earthquake event to obtain an abnormal detection result. The drilling strain data anomaly extraction method based on principal component analysis effectively employs the method of the principal component analysis to perform analysis of the drilling strain data, and performs extraction of possible earthquake precursor anomaly according to the correlation of each measurement item of drilling strain. The drilling strain data anomaly extraction method based on principal component analysis employs the feature values in the principal component analysis and the feature vector angle to represent the weak change of an earth crust so as to realize the accurate extraction of the drilling strain data anomaly in the condition having strong background interference.

The invention discloses a high-order cumulant based bistatic MIMO (Multiple Input Multiple Output) radar parameter estimation method. A joint estimation method of DOD (Direction of Departure), DOA (Direction of Arrival) and Doppler frequency of targets under a gauss color noise background is given out according to bistatic MIMO radar echo signal characteristics through a characteristic that high-order cumulant is insensitive to gauss color noise. The high-order cumulant based bistatic MIMO radar parameter estimation method comprises constructing two opposite angle section matrixes of cross-fourth-order cumulant of output signals of adjacent matched filters through airspace and time domain information of the bistatic MIMO radar; decomposing a singular value of the opposite angle section matrixes of the four-order cumulant, estimating the number of targets, reducing dimensions through eigenvalues and eigenvectors and constructing a new matrix; obtaining joint estimation of the Doppler frequency, the DOD and the DOA of the targets through the eigenvalues and eigenvectors of the novel matrix. According to the high-order cumulant based bistatic MIMO radar parameter estimation method which is an effective bistatic MIMO radar parameter estimation method, a target parameter estimation process has no specific requirements for the number of transmission array elements and receiving array elements.

The invention relates to a self-adaption mixed filtering method of a GPS / SINS (Global Positioning System / Strapdown Inertial Navigation System) super-compact integrated navigation system. The self-adaption mixed filtering method is characterized by comprising the following steps: step 1. establishing a GPS / SINS super-compact integrated navigation system error model; step 2. carrying out data fusion by adopting a Kalman filtering method to obtain an optimal estimation value of a system state and an estimation error equation array P; step 3. carrying out output correction on navigation positioning information output to an SINS by utilizing the obtained optimal estimation value of the system state; step 4. carrying out an observability analyzing method by adopting a characteristic value and a characteristic vector of an error variance array; solving observability information of all the states by using the estimation error equation array P obtained in the step 2; and step 5. carrying out feedback correction on parameters of a GPS and the SINS by taking the observability information of the system state in the step 4 as a feedback factor and taking a product of the feedback factor and the optimal estimation value of the system state in the step 2 as a feedback amount.

A driving risk assessment method based on principal component linear weighting includes the following steps that S1, n tested vehicles are selected, and a vehicle networking device is used for obtaining assessment factors of the tested vehicles and owners of the tested vehicles; S2, a forecasting function model is constructed, S2.1, a sample matrix is constructed; S2.2, a covariance matrix of the sample matrix is obtained, orthogonal transformation is carried out on the covariance matrix, and the eigenvalue and eigenvector of the matrix after orthogonal transformation is carried out are obtained; S2.3, an assessment function of samples is constructed; S2.4, an i principal component is expressed as a linear combination of n original indexes; driving risk levels are obtained. Objective weighting coefficients can be obtained, a principal component method emphasizes the difference principle, the obtained objective weighting coefficients are completely based on data themselves, in this way, the obtained index weighting coefficients have objectivity, and excessive human factor interference is avoided.

The invention relates to a method for reconstructing an interval transit time curve by virtue of multiple logging curves. The method mainly comprises the steps of selecting a plurality of logging curves obvious in response to reservoir characteristics by performing reservoir sensitivity and correlation analysis on the logging curves; performing discrete wavelet decomposition on the logging curves together with a sonic wave curve to form 8 decomposition layers; enabling all the layers of high-frequency decomposition results of other logging curves except the sonic curve to form matrices respectively, obtaining the corresponding characteristic values and characteristic vectors of the matrices; taking the characteristic vectors corresponding to different characteristic values as new components, which, at the moment, have orthogonality (correlation ); performing multivariate regression analysis on the high-frequency components of the corresponding layer of wavelet decomposition of the sonic curve by utilizing characteristic vectors of each layer, calculating a weighting coefficient corresponding to each vector, returning a regression significance analysis result and determining the quality of regression; next, determining the number of layers of the sonic high-frequency components to be reconstructed by virtue of the characteristic vectors according to the regression significance analysis result, selecting multiple regression results as the high-frequency components for a part having more layers than selected layers, and remaining the high-frequency decomposition results of the sonic logging curves for a part having less layers than the selected layer, and then carrying out curve reconstruction by using the low-frequency components of the sonic logging curves and the high-frequency components obtained through regression so as to obtain a final sonic reconstructed curve.