130 results about "Higher-order statistics" patented technology

A method for coding and decoding seismic data acquired, based on the concept of multishooting, is disclosed. In this concept, waves generated simultaneously from several locations at the surface of the earth, near the sea surface, at the sea floor, or inside a borehole propagate in the subsurface before being recorded at sensor locations as mixtures of various signals. The coding and decoding method for seismic data described here works with both instantaneous mixtures and convolutive mixtures. Furthermore, the mixtures can be underdetemined [i.e., the number of mixtures (K) is smaller than the number of seismic sources (I) associated with a multishot] or determined [i.e., the number of mixtures is equal to or greater than the number of sources). When mixtures are determined, we can reorganize our seismic data as zero-mean random variables and use the independent component analysis (ICA) or, alternatively, the principal component analysis (PCA) to decode. We can also alternatively take advantage of the sparsity of seismic data in our decoding process. When mixtures are underdetermined and the number of mixtures is at least two, we utilize higher-order statistics to overcome the underdeterminacy. Alternatively, we can use the constraint that seismic data are sparse to overcome the underdeterminacy. When mixtures are underdetermined and limited to single mixtures, we use a priori knowledge about seismic acquisition to computationally generate additional mixtures from the actual recorded mixtures. Then we organize our data as zero-mean random variables and use ICA or PCA to decode the data. The a priori knowledge includes source encoding, seismic acquisition geometries, and reference data collected for the purpose of aiding the decoding processing.The coding and decoding processes described can be used to acquire and process real seismic data in the field or in laboratories, and to model and process synthetic data.

A method and system are disclosed to detect a broad class of signals including Advanced Television Systems Committee (ATSC) digital television (DTV) and wireless microphone signals. This signal detection method performs in Gaussian noise, employing Higher Order Statistics (HOS). Signals are processed in time and frequency domains as well as by real and imaginary components. The spectrum sensing employed also supports Denial of Service (DoS) signal classification. The method can include parameters that may be tailored to adjust the probability of detection and false alarm.

The invention discloses a method for effectively recognizing digital modulating signals in non-Gaussian noise. Non-linear transformation is performed on a received signal s(t); the generalized first-order cyclic cumulant and the generalized second-order cyclic cumulant of the received signal s(t) are calculated, and a 2FSK signal is recognized by calculating the characteristic parameters of the received signal s(t) and utilizing a minimum mean square error classifier; the generalized second-order cyclic cumulant of the received signal s(t) is calculated, and by calculating the characteristic parameters of the received signal s(t) and utilizing the minimum mean square error classifier, the number of spectral peaks of a generalized cyclic cumulant magnitude spectrum is detected so that a BPSK signal and an MSK signal can be recognized; the generalized fourth-order cyclic cumulant of the received signal s(t) is calculated, and a QPSK signal, an 8PSK signal and other signals are recognized through the calculated characteristic parameters and the minimum mean square error classifier. The method for effectively recognizing digital modulating signals in non-Gaussian noise solves the problem that signals in Alpha stable distribution noise do not have second or higher order statistics, effectively recognizes the digital modulating signals and can be used for recognizing the modulation mode of the digital modulating signals in the Alpha stable distribution noise.

The invention discloses an early warning method of stable-state index of power quality. The method is realized by support from a power quality online monitoring device and a computer. The method is characterized by comprising the following steps of: (1) selecting power quality indexes in advance, including voltage deviation, voltage total harmonic wave, short time flicking, etc, and setting earlywarning threshold values at four grades; (2) acquiring power quality data in real time, and analyzing each power quality index to carry out early warning grading; and (3) saving the values calculatedand a final early warning grade into a database. The early warning method of stable-state index of power quality establishes an early warning structure with two types, including index standard-exceeding type and index abnormal type, and four grades, and adopts time sequence mining algorithm based on high-order statistics in abnormal mining link to analyze potential abnormity of power quality datawith skewness and peakedness. Practice proves that the early warning method is clear in hierarchy, high in algorithm efficiency, strong in accuracy and good in operability. The early warning method is high-efficient and comprehensive.

A wireless transmitter comprises ID signal storage module for storing an ID signal that identifies a group to which a data signal to be transmitted belongs and provides, for each group, second or higher-order statistic properties unique to each group; modulator for modulating the data signal; wireless signal ID assignment module for associating the modulated signal thus modulated with the stored ID signal and for generating a signal to be transmitted, by assigning second or higher-order statistic properties that correspond with the ID signal to the modulated signal that is associated with the ID signal; and wireless signal transmission module for wirelessly transmitting the signal thus generated.

An apparatus, system, and method for detecting high impedance faults in electrical power lines using a composite high impedance fault detection system having a voter logic that samples the logical outputs from a plurality of independent high impedance detection systems and determines a high impedance fault if any two of the plurality of independent high impedance detection systems indicates a high impedance fault. Preferably, the plurality of high impedance detection systems include a wavelet based high impedance fault detection system having a first logical output, a higher order statistics based high impedance fault detection system having a second logical output, and a neural net based high impedance fault detection system having a third logical output. Preferably, each of the plurality of high impedance fault detection systems includes an independent high impedance fault detection application that independently detects a high impedance fault on the electrical power line.

Automatic detection and accurate time picking of weak events embedded in strong noise such as microseismicity induced by hydraulic fracturing is accomplished by: a noise reduction step to separate out the noise and estimate its spectrum; an events detection and confidence indicator step, in which a new statistical test is applied to detect which time windows contain coherent arrivals across components and sensors in the multicomponent array and to indicate the confidence in this detection; and a time-picking step to accurately estimate the time of onset of the arrivals detected above and measure the time delay across the array using a hybrid beamforming method incorporating the use of higher order statistics. In the context of hydraulic fracturing, this could enhance the coverage and mapping of the fractures while also enabling monitoring from the treatment well itself where there is usually much higher and spatially correlated noise.

The invention discloses an asymmetrical voice conversion method based on deep neural network feature mapping, and belongs to the technical field of voice conversion. The asymmetrical voice conversion method based on deep neural network feature mapping disclosed by the invention is specific to asymmetrical data of source voice and target voice. The method comprises the following steps: firstly, performing probability modeling by using the pre-training function of a deep network, and extracting high-order statistics features in a voice signal to provide a standby preferred space of network coefficients; secondly, performing incremental learning by using a small quantity of asymmetrical data, and correcting network weight coefficients by using an optimized transmission error to realize mapping of feature parameters. According to the asymmetrical voice conversion method, a network coefficient structure is optimized and is taken as a parameter initial value of a deep forward prediction network, network structure parameters are further transmitted reversely and optimized in the incremental learning process of a small quantity of asymmetrical data, so that mapping of the personal feature parameters of a speaker is realized.

An apparatus, system, and method for detecting high impedance faults in electrical power lines using a composite high impedance fault detection system having a voter logic that samples the logical outputs from a plurality of independent high impedance detection systems and determines a high impedance fault if any two of the plurality of independent high impedance detection systems indicates a high impedance fault. Preferably, the plurality of high impedance detection systems include a wavelet based high impedance fault detection system having a first logical output, a higher order statistics based high impedance fault detection system having a second logical output, and a neural net based high impedance fault detection system having a third logical output. Preferably, each of the plurality of high impedance fault detection systems includes an independent high impedance fault detection application that independently detects a high impedance fault on the electrical power line.

A distributed collaborative signal identification method based on blind estimation of higher order statistics and signal to noise ratio includes the following steps: establishing feature space; enabling collected features and estimated signal to noise ratio to be input as smart volume management (SVM) aiming at modulating signals of special types, and establishing a modulation signal recognition classifier; and the third step is that each sensor node makes blind estimation for signal to noise ratio of received signals and transmits characteristic parameters together with the signal to noise ratio to a fusion center, and the fusion center distributes weights for each sensor node by utilizing the signal to noise ratio, obtains fused characteristic parameters and thus performs modulation type identification. Compared with the prior art, as for a single sensor node, the algorithm is capable of maintaining high identification probability on the condition of low signal to noise ratio, identified signals are rich in variety, and by making blind estimation for the signal to noise ratio and performing fusion at a feature level, identification accuracy rate of target signals still can be maintained on the premise that a plurality of sensor node channels have extremely poor conditions.

A method is disclosed to detect a broad class of signals in Gaussian noise using higher order statistics. The method detects a number of different signal types. The signals may be in the base-band or the pass-band, single-carrier or multi-carrier, frequency hopping or non-hopping, broad-pulse or narrow-pulse etc. In a typical setting this method provides an error rate of 3% at a signal to noise ratio of 0 dB. This method gives the time frequency detection ratio which may be used to determine if the detected signal falls in Class Single-Carrier of Class Multi-Carrier. Additionally, this method may be used for a number of different applications such as multiple signal identification, finding the basis functions of the received signal.

The invention discloses a self-adapting motion compensation frame rate up-conversion method based on a motion classification mainly solving the problems of inaccurate motion compensation and over big motion complexity in the prior method and comprising: firstly dividing a video into a global motion sequence and a local motion sequence by a global motion estimating method based on an integral projection; obtaining a still background and an occlusion region by a motion detection to the local motion sequence through a high order statistical model; constructing a gray co-occurrence matrix with a pixel distance, quantizing a vein complexity by using two indexes of an angle second moment and an entropy; adding a Cross-Sectional Absolute Deviation (CSAD) criterion to a provided mixing motion estimation method, performing the motion estimation to a motion region of the global motion sequence and the local motion sequence according the vein complexity, avoiding the generation of a superposition and a pore zone during a motion compensation interpolation; finally performing the compensation interpolation to the motion region, and performing a time-domain average to the still region. The occlusion region is a copy of a current frame or former frame.

The invention relates to a feedforward / feedback combined type carrier wave tracking method of a cluster link, which belongs to the technical field of aerial data chains and radio navigation. The invention provides a system structural frame capable of realizing a high-dynamic signal precise tracking and measuring method of the cluster link on a digital signal processor DSP and a FPGA of a circuit board. The invention provides a high-order statistic estimation model for tracking and estimating the Doppler frequency parameter of the link, and the high-order statistic estimation model is directlyused for the iterative feedforward control of the carrier wave tracking. At the same time, a closed loop carrier wave tracking structure of a three-order PLL is maintained for realizing the precise tracking, and the frequency spreading code precise tracking is realized through relaying on carrier wave auxiliary code. The invention overcomes the defects of frequent losing lock and low precision ofthe traditional high-dynamic receiving machine under the severe dynamic condition in the traditional sense. The method disclosed by the invention can be widely used for satellite navigation receivingmachines, distance measuring systems and communication systems based on inhabiting carrier wave modulation direct sequence frequency spreading systems.

The invention particularly relates to a digital modulation signal classification method based on union features, and the method comprises the following steps: (A) obtaining an intermediate frequency real signal after performing ADC sampling for an intercepted signal by a receiving end; (B) preprocessing the intermediate frequency real signal to obtain an intermediate frequency complex signal; (C)obtaining an instantaneous feature quantity, a higher-order statistics and a spectrum feature of the intermediate frequency complex signal, and identifying the signal according to the features, if theidentified signal is a 2ASK signal, a BPSK signal, a QPSK signal, a OQPSK signal and a 8PSK signal, ending the classification; if the identified signal is a MFSK or MQAM modulation main class, executing next step; (D) performing line class internal identification for the MFSK or QAM modulation main class, identifying out a modulation order M in the current modulation main class and then ending the classification. The method combines various features to perform modulation identification, classifies the signals of which features are approximate through multi-level spectrum features, and therebyhas high identification rate and wide range of application.

The invention discloses a method and a system for analyzing personalized information and audio data for mini-mental state examination. The method includes acquiring the personalized information of testees, allowing the testees to complete questions specified by the MMSE (mini-mental state examination), simultaneously recording speaking audio frequencies, extracting acoustic characteristics of the voice audio frequencies for pathological characteristics and representing the acoustic characteristics by the aid of high-order statistics; reducing redundancy of the characteristics by the aid of characteristic selection processes; fusing the acoustic characteristics with reduced dimensions and the personalized information of the testees with one another to obtain personalized characteristics; building pathological models of speaking conditions and mini-mental state examination cognition relations of the testees by the aid of the acquired data and analyzing the personalized information and the audio data by the aid of cross validation processes. The method and the system have the advantages that optional intrusive therapy can be omitted, physical states of the testees can be predicted only by means of analyzing relations between the acquired data and the pathological models, accordingly, the examination time and money can be saved, suffering experienced by the testees in examination procedures can be relieved, and influence of subjective judgment of doctors on results can be reduced.

The invention relates to independent component analysis global search method of realizing the high spectrum fine classification under no prior knowledge situation, the method including: reading in the high spectral data, the establishment of independent component analysis model based on the kurtosis, the center of the data, the ball of data, the iterative solution based on quantum genetic algorithm, independent component compositor, two value of the image, the feature classification. The invention method can established on the circumstance of no data background model, using the self high statistical data to achieve the feature fine unsupervised classification of the high spectral data; at the same time, avoided to plunge in the local best solution problem in the independent component analysis solution process, and compared with the traditional genetic algorithm, the invention used quantum genetic algorithm has less number of iterations, fast convergence, high search efficiency and the strong overall search capability and so on features.

A method equalizes and decodes a received signal including a sequence of symbols. Subsequences of the signal are selected, wherein the subsequences are overlapping and time shifted. For each subsequence, statistics of the channel corresponding to a pattern in the subsequence are selected, wherein the statistics include high-order statistics. A transmitted signal corresponding to the received signal is then estimated based on the statistics.

The invention relates to a seismic data processing method and a formation quality factor estimation method and device. Initial seismic data are subjected to static correction, spherical diffusion compensation, pre-stack noise attenuation and surface consistence amplitude compensation treatment, seismic wavelets are extracted from the stacked data through a high-order statistic method, and an initial quality factor is obtained through an amplitude ratio method; then interference, sorted into the main measuring line direction and the connecting line direction, of the initial quality factor is removed through a high-density property filter method, and finally the optimized quality factor is obtained; and inverse Q filter is carried out through the quality factor, and the final processed seismic data are obtained. An accurate quality factor Q value can be obtained according to the existing seismic data under lack of surface layer survey information and VSP data, inverse Q filter treatmentis carried out, seismic data amplitude information is recovered, the seismic data subjected to inverse Q filter treatment keep reality and the amplitude, and the reliable basis is provided for seismic inversion.

The invention discloses a regularized high-order statistics based hyperspectral space multi-target detection method, which comprises the four major steps of: 1, reading hyperspectral image data by a computer in an MATLAB (MATrix LABoratory) R2010a environment; 2, preprocessing the data by the computer, namely de-equalizing and whitening the data; 3, constructing a detection filter to minimally output high-order statistics of the data after a negative regularized item is added, and solving an optimal weight vector of the detection filter; and 4, setting a proper threshold to acquire a detection result image. Through the method, the defects of the prior art are overcome, the high-order statistics of the data is fully utilized, a good detection effect is achieved in spite of size of a target, and particularly, detection probability under a low false alarm rate condition can be increased. The method has high practical value and broad application prospect in the technical field of hyperspectral remote sensing image target detection.

There is provided a base station comprising: an estimation unit for estimating a second or higher order statistics for each user terminal in a set of user terminals to be served; an estimation unit for estimating a scheduling metric based on the estimated second or higher order statistics; and a scheduling unit for controlling scheduling based on the estimated scheduling metric.

The invention relates to a method for identifying a fault through seismic attributes. The number of adjacent seismic channels and a transverse combination mode of the seismic channels in standardized higher order statistics calculation are determined according to the signal-to-noise ratio of seismic data, a statistics time window is determined in a longitudinal time window, the time retardation of the adjacent seismic channels in the position with the maximum stratigraphic dip serves as a scan time window length, the standardized higher order statistics amount is calculated according to the number and the combination mode of the seismic channels, the statistics time window and the scan time window length, the statistics amount maximum value serves as an extraction attribute according to the transverse combination number and mode of the seismic channels, and time slicing and profile mapping are carried out to identify space distribution of the fault. The method for identifying the fault through seismic attributes has a higher suppressing function on random noise of the seismic data, and particularly suppresses discontinuous interference caused by the stratigraphic dip, the attribute noise resistance is high, and the effect of identifying the stratigraphic discontinuity caused by fracture systems is good.

The invention provides a noise estimation method of a multipath channel, comprising the following steps: carrying out channel estimation for the multipath channel to obtain a time domain response estimation value of each path channel of the multipath channel; setting a noise threshold; selecting paths of all paths of the each path channel, which have the power lower than the noise threshold, and computing a noise estimation value of the multipath channel according to the time domain response estimation values of the selected paths. The invention solves the problem that a noise estimation method utilizing signal subspace needs to carry out higher-order statistics for received signals in a correlation technique and has higher complexity, thereby achieving the effect on reducing channel noise estimation complexity.

The invention provides a method for automatically dividing stratum gyration by earthquake information, which comprises the following steps of: firstly, carrying out the time frequency analysis on earthquake data on the basis of a high-order statistic theory to obtain the high-resolution time frequency analysis result of an earthquake channel; then, extracting a time frequency curve and an extremepoint thereof from a two-dimensional time frequency analysis result by a mathematic means; drawing a regular triangle for representing positive gyration or an inverted triangle for representing negative gyration on two-two extreme points; and finally, completing the automatic division of the stratum gyration by a computer. In the method for automatically dividing the stratum gyration by the earthquake information, which is provided by the invention, the automatic division of gyration can be carried out on different earthquake data; the efficiency and the objectivity of gyration division are improved; and the required gyration information can be rapidly provided for isochronal sedimentation interface identification and explanation.

The invention provides a brain map and brain image registration method based on a high-order statistic deformable model. The brain map and brain image registration method based on the high-order statistic deformable model comprises a first step of selecting a three-dimensional brain map I, selecting N three-dimensional images Mi as training samples, using an affine registration method for registering Mi to the three-dimensional brain map I to obtain MA i , and then using a non-rigid body registration method for registering the images MA i to the three-dimensional brain map I to obtain a series of deformation field vectors fi, a second step of forming four-order tensors Ai by the deformation field vectors fi, solving the mean value A(_) of the four-order tensors Ai, setting Alpha 1= Ai-A(_), and using a lower dimension four-order kernel tensor and four basis matrixes for expressing an estimated value Alpha(^)1 of the Alpha 1, conducting minimization to obtain the optimal solution of the basis matrixes, and expressing the tensors Ai formed by the deformation field vectors through a low dimension four-order kernel tensor and the optimal solution, a third step of conducting deformation on the image needed to be registered based on the obtained Ai and obtaining SD and a fourth step of using the non-rigid registration method for registering the image SD to the brain map I.

The invention discloses a linearity and nonlinearity integrated seismic wavelet extracting method based on high-order statistics, which is characterized in that a linearity and nonlinearity combined method is utilized to solve seismic wavelet parameters after fitting objective function is established on an ARMA (auto regressive moving average) model. The method provided by the invention comprises the following steps of: firstly adopting a cumulant matrix equation method to initially estimate a wavelet model; then applying the obtained wavelet pre-estimated value in determination of an initial parameter searching space by a fitting optimization algorithm; on the basis, utilizing a cumulant fitting error to adjust a threshold value in the matrix equation method and search an accurate optimizing interval and further obtain the accurate order and parameter value of the model through the cumulant fitting method. The wavelet extracting method provided by the invention has high noise immunity, can extract the seismic wavelets with high accuracy in shorter time and has good application value in practical seismic data processing.

The invention relates to a high resolution seismic wavelet extracting method based on high-order statistics and an ARMA (autoregressive moving average) model, which belongs to the field of seismic signal processing. The high resolution seismic wavelet extracting method provided by the invention is characterized in that: under the precondition of performing ARMA parameter simplified modeling on seismic wavelets, SVD (singular value decomposition) based on autocorrelation function is adopted to determine an order of AR part, an MA order determining method is provided for integrating information content criterion function in a high-order cumulant MA order determining method, and the MA order determining accuracy rate in a seismic wavelet ARMA model is improved; an SV-TLS (singular value decomposition - total least squares estimation) and a cumulant method are respectively adopted to estimate wavelet parameters; and under the precondition of ensuring wavelet precision, the model order is decreased as far as possible to improve the operation efficiency and to finally realize seismic wavelet extraction in high efficiency and high precision. Through the data simulation verification and the practical seismic data processing demonstration, the method provided by the invention is proved to effectively improve the estimated precision and extracting efficiency for the seismic wavelets and to have obvious effect even under short-time seismic data and strong noise pollution.

The invention discloses a transformer partial discharge denoising method based on a synchronous compression wavelet transform domain. The method comprises the following steps that: carrying out wavelet transform on an analog local signal to obtain a wavelet coefficient matrix; carrying out high-order statistics analysis on the wavelet transform coefficient matrix, utilizing a kurtosis threshold value criterion to carry out noise preliminary restraining on the wavelet transform coefficient matrix to obtain a corrected wavelet transform coefficient matrix; utilizing the corrected wavelet transform coefficient matrix to obtain a synchronous compression wavelet transform coefficient matrix, and adopting a generalized cross validation algorithm to obtain the noise threshold value level of the synchronous compression wavelet transform coefficient matrix; utilizing the noise threshold value level, and adopting a shear threshold value algorithm to carry out residual noise restraining on the synchronous compression wavelet transform coefficient matrix to obtain the synchronous compression wavelet transform coefficient matrix with small noise interference; and carrying out synchronous compression wavelet inverse transform on the synchronous compression wavelet transform coefficient matrix with small noise interference to obtain a pure partial discharge one-dimensional time domain signal. The method has the advantages of high denoising accuracy and short operation time, and is suitable for occasions of transformer partial discharge denoising.

The invention discloses an optimal material classification and recognition method based on hyperspectral data information maximization. The optimal material classification and recognition method comprises the following steps: (1) selecting training data from acquired hyperspectral data; (2) successively carrying out zero-mean, energy-keeping dimensionality reduction and unit normalizing pretreatment on the training data; (3) estimating a dimension matrix of line dimensionality reduction according to pretreatment data; (4) carrying out information maximizing row dimensionality reduction characteristic matrix calculation according to line-by-line dimensionality reduction dimension arrays; (5) carrying out classifier training according to a row-by-row dimensionality reduction characteristic matrix; (6) selecting an optimal characteristic matrix and an optimal classifier according to a training result; and (7) carrying out material classification and recognition on to-be-classified hyperspectral data according to the optimal characteristic matrix and the optimal classifier. The method provided by the invention has the advantages that the reduction of the hyperspectral data can be performed from a high-order statistics angle, and thus the high classification efficiency is achieved; a new classifier is easy to expand and add, so that the classifier with the excellent property is convenient to generate, and the material classification and recognition are well performed.