226 results about "Correlational analysis" patented technology

A data processing apparatus, a method and computer program product for comparing a first object associated with said data processing apparatus with a second object associated with a remote data processing apparatus, are provided. The data processing apparatus comprises an input device operable to accept a first and a second set of data associated with said first object. Said first set of data describes said first object, and said second set of data describes a third object. Additionally, the data processing apparatus includes a memory. Said memory is adapted to store said first and said second set of data. The data processing apparatus further comprises a receiver. Said receiver is adapted to obtain a first subset of a third and a fourth set of data, respectively, from said remote data processing apparatus. Said third and said fourth set of data is associated with said second object, and said third set of data describes said second object, and said fourth set of data describes a fourth object. It also includes a transmitter. Said transmitter is adapted to send a first subset of said first and said second set of data, respectively, to said remote data processing apparatus. Finally, the data processing apparatus comprises a processor. The processor is adapted to perform correlation analysis between said first subsets of said first and said fourth set of data, respectively, to obtain a first value of correlation between said first and said second object. The processor is further adapted to perform correlation analysis between said first subsets of said second and said third set of data, respectively, to obtain a second value of correlation between said first and said second object.

A steady-state visual evoked brain-computer interface method is based on the stimulation of two frequencies of left and right visual fields, and relates to a visual evoked presentation method of a visual brain-computer interface. The brain-computer interface method is characterized in that four different targets are combined through two blinking frequencies by utilizing a principle of the crossing of the left and the right visual fields; and a correlational analysis method is used to identify which target is watched by a user after brain electrical acquisition equipment records a multi-brain electrical signal in the visual brain area. The equipment and the method of the invention comprise a recording mode of the brain electrical signal, lead selection and signal preprocessing, and a feature extraction method. The visual target presentation method of the invention can be realized through a screen display of a common computer, and the size, the brightness and even the color of the target can be controlled easily. Compared with other steady-state visual evoked brain-computer interfaces, the presentation method can increase the number of the targets under a condition that the usable frequencies are limited, thereby improving the brain-computer interaction ability.

The invention provides a fault line selection method based on time domain waveform correlation analysis of a three-phase current breaking variable, and belongs to the technical field of electric system relay protection. The method includes the steps that as for a power distribution network with a feed-out line only provided with a three-phase current mutual inductor and not provided with a zero sequence current mutual inductor, when single-phase earth fault of the power distribution network occurs, a device is immediately started, and three-phase breaking currents of feeder lines on measurement positions are measured. A three-phase breaking current time domain waveform inside a time window 5 milliseconds after the fault occurs is extracted, correlation analysis of the three-phase breaking current time domain waveform is carried out to acquire an interphase correlation coefficient and a comprehensive correlation coefficient, the comprehensive correlation coefficient and the interphase correlation coefficient respectively serve as a line selection criterion characteristic quantity and a phase selection criterion characteristic quantity, and line selection and phase selection of the distribution network single-phase earth fault are achieved according to criterion principles. Theoretical analysis and a great quantity of emulation indicate that the fault line selection method is good in effect.

The invention discloses a leakage detection method and a leakage detection device based on a combination of a wavelet and EMD reconstruction. The pipeline leakage detection device takes an STM32 single-chip microcomputer as a core, and mainly comprises a sensor, an amplifier, a filter, an A / D conversion module and a transmission module. After the pipeline leakage detection device collects data, the data are transmitted to a computer through a wireless communication module; the computer uses Matlab software to carry out denoising processing on the collected data by combining the wavelet universal threshold denoising and the EMD reconstruction of a multi-azimuth neighborhood mode, positioning is carried out through correlation analysis, so that the position of a pipeline leakage point is obtained. According to the leakage detection method and the leakage detection device, noise can be effectively removed, useful signals can be extracted, and therefore the accuracy of positioning the pipeline leakage point is remarkably improved.

The invention provides an earthquake early warning method based on Internet of Things multi-sensor information fusion and a neutral network technology, which comprises the concrete steps of firstly performing networking fusion on the multi-sensor information of multiple earthquake parameter monitoring points by the Internet of Things technology, wherein the information comprises soil radon alpha energy spectrum, well radon alpha energy spectrum, temperature in soil, humidity in the soil, air pressure in the soil, surface temperature, surface humidity, surface air pressure, surface wind speed, surface rain fall, well water level, well water temperature, well water turbidity, well water mercury, well water carbon dioxide, earthquake sound, earthquake tilting and the like; then obtaining abnormal information feature vectors by wavelet analysis, a correlational analysis method, a feature tree search method, a non-linear quantifying and processing method and a main component analysis method; and finally obtaining the relation between the multi-sensor information abnormality and an earthquake by the neutral network technology, thus achieving the aim of predicting the earthquake. The technical scheme of earthquake early warning method based on the Internet of Things multi-sensor information fusion and the neutral network technology is feasible and convenient to implement and can achieve the aim of more accurately predicting the earthquake.

The invention relates to an atom sparse evidence fusion-based adaptive fault line selection method of small current grounding faults. The method comprises the following steps: firstly, zero sequence current of each line is decomposed by a sparse decomposition method to obtain four optimum atoms, correlation coefficient between the four optimum atoms of each branch line and the optimum atom 1 is calculated by a correlational analysis method, and three atoms which have the highest correlation to the optimum atom 1 are selected as evidence atoms and the three atoms are sorted according to energy from large to small; secondly, energy entropy of each evidence atom is calculated, and an ascertainable fault measure value of each evidence atom is solved by an ascertainable fault measure function; thirdly, the ascertainable fault measure function is corrected to obtain a fault belief function of each evidence atom, and fault belief degree of each evidence atom is solved; and finally, each evidence fault belief degree value is fused to obtain each line fault comprehensive belief degree, a line corresponding to the maximum fault comprehensive belief degree value is selected as a fault line, and a line selection result is output.

Provided is a method for detecting frame sync and frame structure in a satellite broadcasting system, which acquires an estimated value for detecting frame structure and frame sync and overcomes distortion of correlation analysis values by summing differential correlation values for SOF positions in consideration of the variable frame length, and selecting a maximum value in a channel environment with low signal-to-noise ratio and high frequency error. SOF is a sync word indicating the start point of a frame. The method includes the steps of: acquiring SOF differential correlation value sequences; acquiring sums (di,t) of the correlation values normalized for SOF positions based on the number of symbols per frame by using the above-generated SOF differential correlation value sequences; and selecting a maximum value (dz,x) among the sums of correlation values, detecting z as a frame sync position, and detecting x as a frame structure index.

The invention relates to a method for eliminating vibration noise in shaft torsional deformation test. The method comprises the following steps of: arranging reflecting gratings on a tested shaft at equal intervals along an axial direction; arranging a group of laser fiber sensors at a position over against each reflecting grating to constitute a test channel; arranging another completely the same group of laser fiber sensors at a position which is orthogonal to the first group of laser fiber sensors to constitute another test channel; synchronously acquiring two paths of original signals through double channels for performing empirical mode decomposition (EMD) to obtain each order of intrinsic mode function (IMF) comprising different types of information; performing correlation analysis on each order of IMF for the two paths of signals; extracting an IMF reflecting a specific load function; and performing weighted averaging on the IMF to obtain tested information, i.e., shaft torsional deformation. The method has the advantages of simple structure, low cost, high anti-jamming capacity and adaptability to severe environments. Measuring accuracy is increased greatly.

The invention relates to a fault location method of a power distribution network cable-wire combined circuit based on tested simulation, and belongs to the technical field of electrical power system relay protection. When the cable-wire combined circuit breaks down, a time difference between a second wave head and a first wave head or a time difference between a third wave head and the first wave head of a fault traveling wave is obtained through the modulus maximum value so that simulated failure distance can be obtained, and the simulated failure distance is compared with the length of a short sound feeder line so that a time difference of tested simulation location can be determined. After the interference of the short sound feeder line is removed, the simulated failure distance between a cable part and an overhead part is calculated according to the corresponding time difference. A correlation analysis is carried out on wavelet coefficients corresponding to a zero sequence current traveling wave and an actual fault zero sequence current traveling wave feature band under all simulated failure distances, and the simulated failure distance of a maximum correlation coefficient obtained in an analysis result is regarded as an actual failure distance. The fault location method is simple in principle, capable of not being interfered by a reflected wave at the tail end of the sound feeder line and good in robustness.

The invention relates to a myoelectrical denoising method based on CEEMD (complementary ensemble empirical mode decomposition) and interval thresholds. The method comprises the steps as follows: decomposing sEMG (surface electromyography) by CEEMD to obtain intrinsic mode function components; then, selecting proper intrinsic mode function components through component correlation analysis, and performing improved interval threshold processing on each selected intrinsic mode function component; finally, performing signal reconstruction by treated intrinsic mode function components and intrinsicmode function components not subjected to improved interval threshold processing to obtain denoised signals. Due to self-adaptability in signal processing, the method is suitable for analysis of nonlinear and non-stationary sEMG, adverse effects caused by mode aliasing can be reduced, besides, useful information in signals is reserved as much as possible, effects of noise are reduced, so that thesignal-to-noise ratio is increased, the recognition rate is increased, and the application of sEMG is wider.

The invention provides a hydraulic-pump fault feature signal extraction method based on local main value decomposition. Endpoint continuation processing is conducted on collected hydraulic-pump fault signals with a related image method, so that signal endpoints are extreme points; smooth estimation is conducted on extreme values and local mean values with the method of fitting an envelope line through a cubic B spline, so that fitting errors are reduced; then, LMD decomposition is conducted on the preprocessed signals to obtain PF components, correlation analysis is conducted on PFs and the original signals, the real PF components are selected, and finally Hilbert transformation is conducted on the real PF components to obtain Hilbert spectrums of the signals, wherein the Hilbert spectrums are signal features of the hydraulic-pump fault signals. The method is suitable for fault diagnosis of a hydraulic pump in the industrial field.

The invention relates to the field of permanent magnet motor health state monitoring, and discloses a permanent magnet synchronous motor health state monitoring method and system based on multi-mode canonical correlation analysis to discover abnormal conditions of a permanent magnet motor during multi-working-condition operation in time, and the method and system are convenient to use and easy toimplement. The method comprises the following steps: selecting normal operation data of the permanent magnet synchronous motor under more than two working conditions as a training data set, and inputting the training data set into an EM algorithm to calculate and obtain parameters required by modeling; acquiring online operation data of the permanent magnet synchronous motor, and establishing a multi-modal monitoring model according to the parameters and the operation data; and monitoring the health state of the permanent magnet synchronous motor in combination with the evaluation indexes obtained by the model and a preset fault probability threshold.

The invention relates to the technical field of automatic driving, in particular to a vision-laser radar fusion method and system based on deep canonical correlation analysis. The method comprises the following steps: synchronously acquiring an RGB image and point cloud data of a road surface; performing feature extraction on the RGB image to obtain RGB features; sequentially performing coordinate system conversion and rasterization processing on the point cloud data, and then performing feature extraction to obtain point cloud features; inputting the point cloud features and the RGB features into a pre-established and trained fusion model at the same time, outputting feature-enhanced fusion point cloud features, and fusing, by the fusion model, the RGB features into the point cloud features by using correlation analysis in combination with a deep neural network; and inputting the fused point cloud features into a pre-established target detection network to realize target detection. The depth canonical correlation analysis method is creatively used, and the similarity calculation matrix is utilized to fuse two different modal features of the point cloud and the image, so that the precision is improved under the condition of necessary speed compromise.

The invention relates to a human body object monitoring method for remote locations. In the method, a physical sensor is combined with image processing; a video collection device starts only when the physical sensor detects an invasive object; a human body object is further determined by utilizing a human body object verification method based on a video image; video collection, transmission and storage devices are closed if no human body object exists; a video monitoring system starts working if a human body object exists, and meanwhile continues detecting human body objects; the video monitoring system is closed if neither the physical sensor nor video image processing discovers an effective human body object. In the method, only the physical sensor with small power consumption works uninterruptedly, thereby effectively reducing system energy consumption; meanwhile the double test of the physical sensor and the video image processing enhances human body detection performances, reduces redundant data, and effectively increases other correlation analysis processing efficiency of a monitored video.

The invention discloses a characteristic fusion method based on a kernel typical correlation analysis, comprising steps of using a kernel function to map an image matrix to a kennel space, extracting two groups of characteristic vectors of one mode, establishing a criterion function for describing the correlation between the two groups of the characteristic vectors, obtaining two groups of typical projection vector sets according to the criterion function, and extracting a combined typical correlation characteristic through a provided characteristic fusion strategy and applying in the classification identification. The beneficial effects of the characteristic fusion method are that: the kennel introduction technology successfully promotes the linear typical correlation analysis to the non-linear, can extract the characteristic having the stronger discriminability, avoids the decomposition on the matrix after mapping, and simplifies the data operation. The kennel typical correlation analysis algorithm delicately uses the correlation characteristic between the two groups of the vectors as effective determination information, which not only achieves the goal of information fusion, but also eliminates information redundancy between the characteristics and achieves a good result in the classification.

The invention relates to a frozen gait collecting and analyzing system and method based on multi-modal signal synchronization. The system is technically characterized by comprising a functional near-infrared spectrum brain imaging system, an electromyographic signal measuring system and a ground reaction force measuring system, wherein the electromyographic signal measuring system is provided withreal-time marking software; and the electromyographic signal measuring system is connected with the functional near-infrared spectral brain imaging system and the ground reaction force measuring system and achieves the functions of collecting and analyzing cerebral cortex hemoglobin concentration signals, surface electromyographic signals and plantar pressure signals under the control of the real-time marking software. According to the invention, a multi-modal signal synchronous collection mode is adopted, so that time consistency of data analysis is realized, a time synchronization functionbetween required equipment is realized, in each experiment, fNIRS, sEMG and vGRF signals are simultaneously recorded and used for gait analysis, correlation analysis can be performed on the multi-modal signals, and an evaluation continuity function is realized.

The invention discloses a feature layer fusion method and device based on graph embedding canonical correlation analysis. The feature layer fusion method comprises the steps that samples in all modesare mapped to a projection matrix of a space of the same classification result, and L21 norm regularization is applied to the projection matrix so that independent complementary features can be selected from multiple single-mode feature spaces at the same time; constructing a data similarity graph matrix to represent the similarity relation of sample points in the single-mode feature space; and learning a corresponding projection matrix for each mode through the regularization target function, and projecting the plurality of mode data to a projection subspace with the maximum discrimination, the maximum correlation and the minimum redundancy to realize multi-mode data fusion. According to the method, multi-modal data fusion is realized, the interference of redundant information in an original feature space is eliminated, the single-modal feature discriminability is improved, the correlation between multi-modal sample sets is enhanced, the recognition performance and stability are improved, the feature fusion effect is good, and the recognition effect is good.

The invention discloses a lightning stroke overvoltage identification method based on overvoltage waveform characteristics, and relates to the technical field of electric power. Correlation analysis of transient overvoltage waveforms of different lightning stroke overvoltage and general short-circuit fault overvoltage is carried out, inductive lightning stroke interference signals are distinguished from direct lightning stroke line fault signals according to the minimum similarity coefficient and the average similarity coefficient among the four lightning stroke overvoltage waveforms, and due to the fact that an inductive lightning stroke overvoltage wave head is generally positive, and a direct lightning stroke overvoltage wave head is generally negative, the difference between the polarities of the wave heads of the inductive thunder and the direct lightning is used as an auxiliary criterion, the reliability of the identification criterion is improved, inductive lightning stroke faults and direct lightning stroke tower fault signals are identified, lightning stroke interference of the distribution line and various lightning stroke faults are identified by using the similarity coefficient between overvoltage waveforms, and a very good identification effect is achieved; and the method is suitable for large-scale popularization and application.

The invention relates to a method for detecting the mechanical property of a self-piercing riveting joint. A system adopted for the method comprises a riveting test piece, a material mechanical property testing device, an electron microscope, an image collection card, an image analyzing and processing system and the like. The double-riveting-joint test piece is designed. Image sequences produced before and after deformation occurs in an internal lock area in a riveting and fastening machine are obtained through an image observing system composed of the electron microscope. Before a load is exerted on the test piece, template images are selected inside a plane area formed by an angular section cutting plane of a rivet and outside the plane area respectively. Single images in the image sequences are analyzed and calculated in the deformation process through a correlational analysis method, target images corresponding to the template images are obtained through matching and by using the largest correlation coefficient, and the slide characteristic quantity of an internal lock of the riveting and fastening machine for the self-piercing riveting joint is obtained by calculating the relative slide amount between the template images and the target images. Therefore, the quality level of the self-piercing riveting joint can be comprehensively evaluated, and the scientific basis is provided for relevant research, design, crafting and the like of the self-piercing riveting joint.

The invention discloses an evaluation method and system of water quality based on fish electrocardio indexes. The method comprises the following steps: receiving fish body electrocardio signals beforeand after fish contact with a solution which contains specific organic matter, and respectively extracting multiple kinds of electrocardio indexes; analyzing the relativity between environmental stress and each extracted electrocardio index according to a Pearson correlational analysis method, and screening the electrocardio indexes for evaluating the specific organic matter in the water quality;analyzing and evaluating environmental stress and linear regression models between various electrocardio indexes according to Akaike information criteria and the residual sum of squares so as to obtain the electrocardio indexes used for finally evaluating the organic matter in the water quality. Receive the fish body electrocardio signals before and after the fish contact with to-be-determined water, and respectively extracting an electrocardio index which is corresponding to a certain organic matter in determined evaluation water quality; contrast the extracted electrocardio indexes of the fish body electrocardio signals before and after the fish contact with the to-be-determined water so as to obtain electrocardio index changes for evaluating the corresponding organic matter in the to-be-determined water quality.

The invention discloses a method for searching a human face corresponding to an unknown skull from a three-dimensional human face model library based on canonical correlation analysis. The method comprises the following steps: respectively processing skulls and human faces in a training sample set to obtain normalized training sample data; respectively performing subspace analysis on the overall skull, the overall human facial skin, and corresponding regions of forehead, eyes, nose and mouth to obtain ten corresponding subspaces; establishing correlation analysis models of the subspace characteristics of the skull and the human face by utilizing the canonical correlation analysis, wherein the correlation analysis models comprise the correlation analysis models of the whole body and the four subspaces; and realizing human face identification of the unknown skull by model fusion. By the method, the identification rate can reach 100 percent; and the method has an important application value in the fields of criminal investigation, disaster investigation and the like.

The invention discloses a reliability evaluation method and system for a distributed state sensor of power distribution main equipment, and the method comprises the steps: autonomously extracting thespace-time distribution features of data through a space-time diagram convolution neural network SASTGCN model based on a self-attention mechanism, building a nonlinear relation between the data of the distributed state sensor and an evaluation value. Problems of nonlinearity, uncertainty and the like in reliability evaluation can be well solved. The limitations that uncertainty factors and subjective factors of people in a previous reliability evaluation method greatly influence an evaluation result, the calculation complexity is high, correlation analysis and sensitivity analysis are not accurate enough and the like are overcome. By using the method, the spatial-temporal characteristics in the sensor data can be accurately captured, the complex mapping between the sensor data and the evaluation value can be well simulated, the nonlinear relationship between the evaluation value and the sensor data can be established, the evaluation precision is high, and the practical value in engineering is high.

The invention discloses an opportunistic group intelligent air quality monitoring and evaluating method based on mobile equipment. A group sensing system is constructed by using an opportunistic network and comprises a plurality of intelligent devices in the same area. The system is an intermittent connection system; each intelligent device with a gas sensor collects gas data for air quality monitoring, the data are sent to other intelligent devices in the same area in an opportunity multi-hop mode through a mobile repeater, and the air quality in the area is calculated through a differentiation prediction model; the related analysis of the per-hour concentration of PM2.5 and SO2, NO2, O3 is carried out and Spearman correlation analysis is carried out on each group of gas data by using k-Sstatistical test of each group of data to obtain a relationship between the concentration of PM2.5 and the concentrations of SO2, NO2, O3 and CO, and thus gas having the greatest influence on the concentration of PM2.5 is screened out; and then an effective model is constructed according to the Spearman grade correlation between the gas parameters of an air quality monitoring station and PM2.5.

The invention discloses a heavy metal content support vector machine regression method combining bands and ratio combinations of indoor and outdoor spectrums. The method comprises the following steps:S1, collecting a soil sample and measuring a soil spectrum; S2, constructing a spectrum conversion model by using a KS-DS algorithm; S3, carrying out spectrum pretreatment and correlation analysis; and S4, constructing a support vector machine heavy metal content inversion model fusing indoor and outdoor spectrums, and carrying out detection. The method has the advantages of simple principle, high inversion precision, good detection effect and the like.

A flood inundation range dynamic simulation method integrated with active and passive microwave remote sensing information comprises the steps of extracting a time sequence of M / C signals by satellite-borne passive microwave brightness temperature data, and preliminarily determining the occurrence time of a flood event; searching a satellite-borne active microwave remote sensing SAR image corresponding to the flood event based on the occurrence time of the flood event, and performing water body / land classification based on a machine learning method; based on the SAR image water body / land classification result, the surface water area ratio FWS of M pixels is extracted, making correlation analysis on the surface water area ratio FWS and corresponding M / C signals, and creating a relation model of M / C and FWS; utilizing the M / C signal to deduce the space-time distribution of the FWS at a permanent water body pixel, and realizing the dynamic simulation of the flood inundation range at the permanent water body. Based on real-time or quasi-real-time satellite microwave remote sensing observation information, the flood dynamic inundation range can be solved, flood simulation is practicallyguided, and important data support and decision basis are provided for flood emergency rescue and prevention.

The invention discloses a multi-dimensional fault detection and recognition method, wherein the method comprises the following steps: S1, constructing a fault detection statistic, and carrying out thedetection of a fault based on chi-square detection; S2, when a fault is detected, constructing a fault set by fusing identification results of a data detection method and a correlation coefficient method; and S3, identifying and eliminating faults according to the fault set. Two typical RAIM algorithms of a data detection method and a correlation analysis method are fused to identify and processa fault. Under the condition of multi-dimensional faults, compared with independently adopted two RAIM algorithms of the data detection method and the correlation analysis method, the method has higher fault recognition accuracy and recognition efficiency, and the integrity and reliability of a satellite navigation system can be further improved.

The invention provides a concentrate grade prediction method in a zinc flotation process. The concentrate grade prediction method comprises the following steps: firstly, collecting zinc flotation process image feature data and corresponding concentrate grade data through a flotation process image collecting and processing system, and preprocessing the collected data; forming a sample space according to the preprocessed zinc flotation image feature data and concentrate grade data, then dividing five independent sub-sample spaces according to the concentrate grade, and respectively performing time difference; performing correlation analysis on the features by adopting KPCA, and extracting features with high contribution rate as key foam features; training an LSSVM based on the key feature sample, and establishing a relationship between the image features and the concentrate grade; and solving the prediction model by using an improved flora algorithm, and optimizing two parameters of a penalty factor xi and a kernel width sigma of the LSSVM. The concentrate grade prediction method can be directly implemented by programming on a computer, is low in cost, high in precision and good in timeliness, and has important significance for guiding field production.