61 results about "Functional principal component analysis" patented technology

The invention discloses a time series data cleaning method based on correlation analysis and principal component analysis, comprising the following steps: finding out hidden correlation between transformer faults and other power data by using a Pearson's correlation coefficient (PCC); reducing the dimension and noise of all relevant time series through principal component analysis (PCA); and inputting part of cleaned data as a training set into a BP neural network (BPNN) for training and learning, and taking the remaining data as a test set to verify a model. Compared with the traditional technology, the accuracy of transformer fault diagnosis is significantly improved, the accuracy of classification is improved, and the operation time is shorter for high-dimensional data.

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 discloses a vehicle driving condition establishment method combining principal component analysis with fuzzy c-mean clustering. The method comprises the steps of extracting satellite positioning data of each road traffic condition in a vehicle dynamic monitoring system of a road transport enterprise, and calculating and dividing the data into small sections of micro-strokes; and performing calculation of characteristic parameters such as an average speed, an idle time proportion and the like for each micro-stroke, obtaining a matrix of a sample quantity (row) X the characteristic parameters (column), adopting mean normalization and principal component analysis for matrix data, selecting preorder principal components which meet the conditions that the cumulative contribution rate of characteristic values of the principal components is greater than 85% and the principal components can comprehensively reflect all the characteristic parameters, performing fuzzy c-mean clustering analysis on scores of the principal components, and clustering the micro-strokes into different groups, namely, screening sub-conditions. An initial synthesis condition is smoothed by adopting a filter with a double-weighted smoothing kernel function. According to the method, existing satellite positioning data is fully utilized; compliance testing is easily carried out on a dynamometer; relatively high universality is achieved; and the research cost of vehicle driving condition establishment is reduced.

The invention discloses a brain cognitive state judgment method based on polyteny principal component analysis (PCA). The method includes the following steps of firstly, inputting sample sets, and processing input data; secondly, calculating characteristic decomposition of training sample sets, determining an optimal feature transformation transformational matrix, and projecting training samples into tensor characteristic subspace to obtain feature tensor sets of the training sets; thirdly, vectorizing lower dimension feature tensor data which are subjected to dimensionality reduction as input of linear discriminant analysis (LDA), determining an LDA optimal projection matrix, and projecting the vectorized lower dimension feature tensor data into LDA feature subspace for further extracting discriminant feature vectors of the training sets; and fourthly, classifying features, subjecting the discriminant feature vectors obtained by projection of training images and test images to feature matching, and further classifying the features . According to the brain cognitive state judgment method, PCA is utilized to directly perform dimensionality reduction and feature extraction to multi-level tensor data, the defect that structures and correlation of original image data are destroyed and redundancy and structures in the original images can not be completely maintained due to the fact that traditional PCA simply performs dimensionality reduction is overcome, and space structure information of functional magnetic resonance image (fMRI) imaging data is kept.

The invention discloses a method for predicting remaining service life of a lithium battery based on FPCA (functional principal component analysis) and Bayesian updating. The method comprises the following steps of utilizing FPCA to build a lithium battery non-parameter degrading model, predicting the remaining service life of the lithium battery based on the degrading model, and analyzing the influence of different amount of modeling data on the predicting accuracy. The method has the advantages that an empirical Bayesian method is provided for updating the lithium battery degrading model in real time; the lithium battery degrading model is corrected in real time by the Bayesian method, so as to obtain the more accurate lithium battery degrading model; the distributing of the remaining service life of the lithium battery is calculated by a parameter bootstrap method, and the confidence interval is calculated.

A handwriting authentication and verification method based on single character characterized by that, four traverse line element characteristic is extracted to better reflect the feature of the Chinese characters, then one of the following methods is used to reflect the optical diagnostic character of the difference between different writers, one is to extract the most distinctive feature through direct LDA (linear differentiation analysis) transformation, the other is to first extract the most distinctive feature through direct PCA (principal component analysis) transformation dimensionality reduction, then using LDA transformation to obtain the most distinctive diagnostic character. In accordance with the invention, an Euler's distance classifier is used for categorization authentication.íí

The invention discloses a hyperspectral image classification method based on K nearest neighbor filtering. The classification process mainly includes (1) support vector machine (SVM) classification: rough classification of a hyperspectral image using a SVM classifier to obtain an initial probability graph; (2) principal component analysis dimensionality reduction: dimensionality reduction of the hyperspectral image by way of principal component analysis to obtain a first principal component image; (3) K nearest neighbor filtering: extraction of spatial information of the hyperspectral image under the guidance of the first principal component image based on a non local K nearest neighbor filter to optimize the initial probability graph; and (4) accurate classification of the hyperspectral image according to the optimized probability graph. The greatest advantage of the method in the invention over a traditional hyperspectral classification algorithm is that the non local spatial information of the hyperspectral image can be extracted for optimized classification without solving for a complex global energy optimization problem. Thus, the classification speed is high, and the accuracy is high.

A method, system and program product are disclosed for statistical modeling an integrated circuit that provides information about partial correlations between model parameters. The invention determines a variance-covariance matrix for data to be modeled; conducts principal component analysis on the variance-covariance matrix; and creates a statistical model with an independent distribution for each principal component, allowing calculation of each individual model parameter as a weighted sum by a circuit simulator. The statistical model provides information about how well individual transistors will track one another based on layout similarity. This allows the designer to quantify and take advantage of design practices that make all transistors similar, for example, by orienting all gates in the same direction. A method, system and program product for simulating a circuit using the statistical model are also included.

The invention discloses a daily load curve dimensionality reduction clustering method based on kernel principal component analysis, and the method comprises the steps: carrying out the nonlinear mapping of a preprocessed daily load data matrix to a high-dimensional space through employing a kernel function, and obtaining a high-dimensional kernel matrix; correcting the high-dimensional kernel matrix, performing eigenvalue decomposition to obtain a corresponding eigenvalue and a unitized eigenvector, and setting extraction efficiency according to an eigenvalue descending trend to obtain a principal component and the number of dimension reduction indexes; secondly, taking the eigenvalue as a weight, performing normalization processing with the sum of the eigenvalue and the weight being 1 toobtain a weight vector, and taking the projection of the corrected Gaussian kernel matrix on the principal component as a dimensionality reduction data matrix; and finally, clustering the daily load curve by taking the dimension reduction data matrix and the weight vector as input of a weighting algorithm, and determining an optimal clustering number and a clustering result based on a Silhouette index. According to the method, the clustering quality is improved while the calculation efficiency is improved. And the clustering result is consistent with the reality, so that the method has a certain engineering value.

The invention discloses a double main element-dynamic kernel principal component analysis (DME-DKPCA) fault diagnosis method based on a chemical TE process. According to the method, a DOD algorithm isused to determine an optimal parameter order of process data, after a generated dynamic matrix projects in a kernel principal element space, the suppression ability of an R principal component method to high-dimensional noise is used to effectively determine the number of principal components of normal state data in a kernel feature space, and T2 and SPE control limits of a fault detection high-dimensional space are generated; then a CPV method is used to retains fault information to the maximum extent, the number of new principle components is determined from data to be detected, and T2 and SPE statistics are recalculated to monitor and detect a fault. According to DME-DKPCA method, the recognition rate of fault detection is effectively improved, and the method has better model accuracy than a DKPCA method.

The invention belongs to the field of autonomous navigation, guidance and control of spacecrafts, and provides a star pattern recognition method based on plane triangles. The method comprises the following steps: firstly, establishing a navigation star database, specifically, constructing characteristic plane triangles, carrying out principal component analysis on all matrixes formed by the characteristic plane triangles, calculating projection values of the characteristic plane triangles in the direction of intrinsic axis, and sequencing the projection values; then, constructing observation plane triangles; and finally, searching for characteristic triangles matched with the observation plane triangles in the navigation star database by means of a K-vector searching method, and getting rid of redundancy match by means of least square errors to determine the final matched triangles. The star pattern recognition method based on the plane triangles provided by the invention overcomes the defects that the poor noise robustness is poor, more redundancy matches exist and the identification rate is low in conventional triangle identifying methods. The retrieval of the navigation star database is accelerated, the redundancy matches are eliminated, the recognition rate is increased, and the noise robustness is good.

The invention discloses a software defect prediction method based on kernel principal component analysis and extreme learning machine. Aiming at the problems that irrelevant module features damage defect prediction model performances in prediction of software defect quantity and the original defect data cannot be accurately represented by the original characteristics, the method disclosed by the invention comprises the following steps: projecting original data into a potential characteristic space through nonlinear mapping by adopting kernel principal component analysis, so that the mapped characteristics are capable of accurately characterizing the complex data structure and increasing the linear separability probability of data in the space; and extracting the representative characteristics of the data, and finally establishing a defect prediction model by adopting an extreme learning mechanism according to the data after characteristic extraction.

The invention relates to an online target tracking method based on increment bilateral two-dimensional principal component analysis (Bi-2DPCA) learning and sparse representation, and provides an increment Bi-2DPCA learning algorithm capable of rapidly and accurately updating a target sub-space model to reflect an appearance change of a target in a tracking process. Aiming at the problem that the target is frequently blocked and polluted by noise in the tracking process to cause that the tracking effect becomes worse, according to the method disclosed by the invention, a Bi-2DPCA-based sub-space model is embedded below a sparse representation frame, so that interferences to target positioning and target sub-space model updating caused by blocking and noise are furthest removed. Meanwhile, a novel method for calculating visual similarity is used. By adopting the method, energy distribution of Bi-2DPCA when an image is presented is considered, and is more accurate in comparison with a classical reconstruction error; tracking is achieved under a Bayesian inference framework; the target state is estimated by using a particle filtering algorithm.

The invention relates to a mechanical vibration signal feature extraction method based on combination of stochastic resonance and kernel principal component analysis. The method comprises the steps that firstly, the stochastic resonance method is applied for conducting pretreatment on rotor oscillation original signals measured by a sensor, the signal periodicity is improved, and the oscillation signal to noise ratio is improved; then, a time domain feature set is extracted for the pretreated output signals; then the kernel principal component analysis method is adopted for conducting nonlinear feature transformation for the extracted time domain feature set, and therefore the final needed feature set is obtained. The method is applied to feature extraction and failure diagnosis of simulated failure of an engine rotor, the result shows that the feature set extracted through the method is of linear independence, the number of dimensions is smaller, the separability is higher, the precision and efficiency of the failure diagnosis can be effectively improved, and application in engineering practice is facilitated.

The invention belongs to the field of image analysis processing, is applicable to the field of automatic detection and control of the surface quality of fabrics, and relates to a method for analyzing an image based on principal component analysis and a method applicable to detection of the defects of fabric. The method for analyzing the image based on the principal component analysis comprises the following steps of: firstly, expanding gray values in an original image sample into two groups of vectors according to rows and lines; secondly, performing template operation on the two groups of vectors, and respectively performing principal component analysis on the two groups of vectors which are subjected to template operation to obtain corresponding principal component matrixes; and finally, performing projection operation on a sample to be detected by using the two principal component matrixes, and calculating the similarity of the sample after projection and the sample before projection to analyze the characteristics of the image. The invention has the advantages that: non-uniform illumination can be eliminated without the conventional pre-processing step; calculation in a detection period is simple; the original fabric sample is respectively expanded according to the rows and the lines and then subjected to template operation, so that the longitude and latitude orientation characteristics of fabric texture can be fully utilized, the defects can be highlighted, and the random interference of the texture is restrained; and detection accuracy rate is improved.

The invention discloses a multi-response parameter optimization method based on principal component analysis and a neural network. The method comprises the following steps: 1) eliminating the correlation of a plurality of responses by principal component analysis; 2) taking the horizontal combination value of influence factor variable temperature and time as an input variable of the neural network, taking a corresponding MPI (Multi-response Performance Index) value as the expected output variable of the neural network, and establishing a RBF (Radial Basis Function) neural network model; and 3) utilizing the RBF neural network model obtained by training to search an optimal technological parameter. A RBF neural network prediction model of a mapping relationship between a production process influence factor and the multiple responses is established, the principal component analysis is applied to convert a multi-response index into an irrelevant index, the multi-response index is converted to a single-response index of comprehensive assessment through weighting, the response with high prediction ability is optimally improved, and the optimization of the integral effect of the plurality of responses is realized.

The invention discloses a feed pump fault detection method based on multiscale principal component analysis. The method comprises the following steps that: utilizing discrete wavelet transform to decompose collected variable data, utilizing a principal component analysis method to determine a wavelet coefficient in each scale, and selecting coefficients which are greater than a specific thresholdvalue to generate a multiscale model; and utilizing the principal component analysis method to carry out modeling on a new statistical amount, independently calculating a T2 statistical amount and a Qstatistical amount, and giving a fault alarm when one statistical amount exceeds the threshold value. By use of the method, the process data of a feed pump is substantially multiscale, and has differences in different time domains and frequency domains. A traditional statistical method based on a single scale can not accurately separate main variables which show a system operation state, and therefore, a multiscale principal component analysis method is selected to carry out feature extraction to improve detection sensitivity.

The invention discloses a non-linear dynamic process monitoring method based on the non-linear principal component analysis of a normalized variable, which comprises the following steps: acquiring a data matrix Y, pre-specifying a value of p and a system order n; the Hankel matrix of the past and future observational side values being combined according to the formula; calculating covariance and cross-variance matrices of past and future observations; singular value decomposition of H matrix; calculating a state vector and a residual vector; the state vector being projected onto the high dimensional feature space by explicit second order polynomial mapping; the first k principal components being determined by eigenvalue decomposition in principal component analysis; finally, the T2 statistic, the combined statistic Qc and their corresponding control limits being calculated. The method of the invention is used for monitoring three different types of faults in the Eastman chemical process of Tennessee, and the simulation results show that the proposed CV-NPCA method has high fault detection rate and relatively low fault false alarm rate.

The invention discloses a software defect prediction method based on principal component analysis and combined sampling. The software defect prediction method comprises the following steps: S1, dimensionality reduction and denoising are selected for software defect data through fusion characteristics; S2, performing SMOTE oversampling and hierarchical random sampling on the data subjected to dimensionality reduction in combination for sampling, the oversampling means that class samples in a data set are relatively balanced by increasing the number of few class samples, hierarchical random sampling means that classification is performed by dividing classes, and no-replay random sampling is adopted in each layer; and S3, selecting a classifier for the processed data, and optimizing classifier parameters. According to the method, the random forest classifier is selected, and the characteristics of the characteristic subset are randomly selected, so that the purpose of randomizing the treeis further achieved, the overfitting problem of the classifier is avoided, finally, the software defect prediction performance and prediction efficiency are improved, and a good theoretical and experimental basis is provided for predicting defective software in reality.

The invention discloses an SAR (Synthetic Aperture Radar) image target identification method based on multilinear principal component analysis and tensor analysis. The method comprises the following steps that: constructing a four-order tensor training sample; utilizing the multilinear principal component analysis to obtain a multilinear projection matrix; constructing a core tensor; carrying out linear discriminant analysis on the core tensor to obtain the weight vectors of one group of linear discrimination functions; and carrying out classified identification on test samples. By use of the method, the four-order tensor sample is constructed for the SAR image, the multilinear principal component analysis is adopted to extract features, image structure information is effectively kept, and a correct recognition rate of a target is improved.

The invention relates to a two Swarm satellite magnetic field data earthquake precursor anomaly extraction method based on principal component analysis. The method comprises the following steps: reading magnetic field Y component data of a satellite Swarm A and a satellite Swarm C, and removing incorrect data according to flag bits Flags_B of the data; selecting a study time range and a study areaaccording to an earthquake; selecting orbital data obtained when local times of the satellite Swarm A and the satellite Swarm C are night times; removing effects of the main geomagnetic field throughan IGRF model; projecting the original data on a group of new spatial orthogonal bases through the principal component analysis on the magnetic field Y component data of the satellite Swarm A and thesatellite Swarm C, so that principal components, of which the variances are arranged in a descending order, are obtained; finding out the components with high geomagnetic activity relevance through comparison between the principle components and geomagnetic indices, removing the components with the high geomagnetic activity relevance, and only analyzing the remaining principal components; and carrying out anomaly extraction on the remaining principal components through a skewness and kurtosis coefficient defined by skewness and kurtosis, so that earthquake precursor anomalies can be extracted. The method provided by the invention has the advantages that geomagnetic activity interference can be removed, so that the earthquake precursor anomalies can be accurately extracted.

The invention discloses an ultraviolet spectrum water quality abnormality detecting algorithm based on multi-scale sliding window principal component analysis. The method comprises the following steps: 1) solving a wavelet transformation dimension L, a window length N, a Cusum control limit H1 of each scale and a wavelet reconstruction data Cusum control limit H' according to historical data; 2) acquiring on-line spectral data, and waiting for the window N filled with the data; 3) performing baseline correction and standardized preprocessing on the spectral data; 4) performing MSPCA on the spectral data, and selecting a principal component number according to a threshold value method; 5) performing each scale abnormality detection based on a Cusum control chart; 6) abnormal wavelet scale combination and reconstruction, and performing PCA calculation on the reconstruction data; 7) performing reconstruction data abnormality detection based on the Cusum control chart, and generating a water quality report. According to the method, the traditional principal component analysis ultraviolet water quality abnormality detection method is improved, so that the method can dynamically adapt tothe water quality background fluctuation, multi-scale water quality abnormality detection can be performed, and the detection accuracy of dynamic change water quality detection is increased.

The invention discloses an image noise level estimation method based on principal component analysis. The method comprises: step 1, dividing a raw image into a plurality of image blocks with the same size; step2, turning the selected image blocks into an image data matrix, and calculating a covariance matrix of the image data matrix; step 3, conducting eigen decomposition of the covariance matrix, setting an obtained minimal eigenvalue as an estimate of the noise level, and using the estimate of the noise level to calculate a threshold; step 4, calculating a gradient matrix corresponding to each image block and the trace of a covariance matrix corresponding to the gradient matrix; and step 5, picking the image block with the trace not greater than the threshold, and repeating the step 2 to 4 until the estimates of the noise level of two continuous calculations are the same or the iteration is repeated for a preset number of times. The method uses principal component analysis to find a smooth block in the image more accurately. The method is precise and robust and is applicable to large-scale noise levels and various scenes.

The invention discloses a neural network prediction method based on principal component analysis, and the method comprises the following steps: 1, collecting process data and quality data, and carrying out the processing of the data through the principal component analysis; and step 2, establishing a neural network model by using the data obtained in the step 1, and performing prediction. According to the method, firstly, process variables and quality variables generated in the chemical process are collected, a principal component analysis method is used for preprocessing data, the data dimension is reduced, redundancy is avoided, the processed data are input into a prediction model of the radial basis function neural network, corresponding parameters are solved and optimized, and the model prediction accuracy reaches a preset value. Different from a traditional prediction method, the method combines a principal component analysis method and a radial basis function neural network model, reduces the complexity of modeling, and improves the precision of the model.

One or more test features are extracted from a time domain signal. The test features are represented by discrete data. The discrete data is represented for each of the one or more test features by a corresponding one or more fitting functions, which are defined in terms of finite number of continuous basis functions and a corresponding finite number of expansion coefficients. Each fitting function is compressed through Functional Principal Component Analysis (FPCA) to generate corresponding sets of principal components. Each principal component for a given test feature is uncorrelated to each other principal component for the given test feature. A distance between a set of principal components for the given test feature and a set of principal components for one or more training features with the processing system is calculated. The test feature is classified according to the distance calculated with the processing system.

The invention discloses a multispectral imaging method and device based on a demosaicing algorithm and principal component analysis. The method comprises: integrating a preset number of channels of wide-spectrum multi-spectral filter arrays in front of a detector, and obtaining collected data, wherein the collected data are mosaic images corresponding to the wide-spectrum multi-spectral filter arrays MSFA; processing the acquired data by using a demosaicing algorithm, and performing reconstructing to obtain a complete airspace full-resolution image corresponding to a preset number of channels;and performing parallel reconstruction decoupling on the spectrum of each spatial point through a principal component analysis theory by using the spectrum modulation characteristics of the MSFA andthe spectrum modulation data of each spatial point to obtain a multispectral image. According to the method, multispectral imaging can be realized by utilizing a good reconstruction result of a demosaicing algorithm when the number of channels is small and a principal component analysis dimension reduction effect.

The invention discloses a medium and long term forecasting method based on mutual information and principal component analysis screening factors, and relates to the field of hydrological forecasting.The method comprises the following steps of: firstly, analyzing a forecasting factor which greatly affects the predicted runoff based on a mutual information-based factor screening method; extractingmain components of the factors by using a main component analysis method; the optimization of forecasting factors of the forecasting model is realized; according to the method, the linear and nonlinear relationship between two variables is fully considered, the information of the overlapped part is removed, the number of original variables is reduced, and the characteristics of a research object are typically shown, so that the obtained medium and long term runoff forecasting model is better in forecasting effect and higher in model stability compared with a traditional method.

The invention belongs to the field of generative adversarial network improvement. The invention discloses a semi-supervised classification method for improving a generative adversarial network based on principal component analysis. A principal component analysis (PCA) method is used for carrying out compression dimension reduction on an original picture, vectors obtained through dimension reduction are used for replacing original random noise vectors to serve as input of a generator, the quality of the picture generated by the generative adversarial network is improved, and meanwhile the accuracy of semi-supervised classification is also improved. According to the principal component analysis method, part of features of original data can be reserved while the data dimension is reduced, a random noise input generator is replaced with the data subjected to dimension reduction, data with higher quality (closer to a real picture) can be generated more quickly, the number of network iterations is reduced, and the classification accuracy is improved.

The invention discloses a method for repairing the missing points of transformer operation data based on functional principal component analysis and wavelet transform, which comprises the following steps: analyzing the collected operation data points by using FPCA method, and fitting the operation data function xi (t) on the whole time series; obtaining The residual function Epsilon (t) by the difference between the original data point and the data point obtained by FPCA. The residual function Epsilon (t) is denoised by wavelet transform, and Epsilon '(t) is obtained. The estimation function (shown in the description) substitutes time t0 at the missing point (shown in the description) to obtain (showin the description) as the repair value at the missing point. The invention adopts the modeof combining the FPCA and the wavelet transform, can grasp the characteristics of the data set as a whole, and can improve the local fitting degree. Compared with the traditional fitting method, theprediction value of the invention is more reliable.

The invention discloses an evaluation index system construction method based on distance classification of images and principal component analysis. The method comprises the steps of selecting initial evaluation indexes; performing index classification based on distance classification of images; performing index screening based on principal component analysis; and obtaining a simplified index system. Elements with high correlation are divided into a class by using a classification method of combining correlation coefficients with the shortest path in the graph theory, namely an image distance classification method, and index representative elements are selected from each class for principal component analysis, so that the information processing workload is reduced. Retention indexes in the principal component analysis and indexes belonging to the same class as the retention indexes are selected as important indexes for constructing an index system, so that missing of important indexes is avoided.