88 results about "Time frequency resolution" patented technology

A system and method for monitoring a patient includes a sensor configured to acquire physiological data from a patient and a processor configured to receive the physiological data from the at least one sensor. The processor is also configured to apply a spectral estimation framework that utilizes structured time-frequency representations defined by imposing, to the physiological data, a prior distributions on a time-frequency plane that enforces spectral estimates that are smooth in time and sparse in a frequency domain. The processor is further configured to perform an iteratively re-weighted least squares algorithm to perform yield a denoised time-varying spectral decomposition of the physiological data and generate a report indicating a physiological state of the patient.

The invention provides a high-precision time-frequency analysis method, and belongs to the field of digital signal processing. The method comprises the following steps of: (1)inputting original digital signals x(t), then carrying out empirical mode decomposition (EMD)to the original digital signals x(t), decomposing the original digital signals x(t) into sets of intrinsic mode function (IMF)components, namely ci(t) is the IMF component; (2) carrying out matching pursuit (MP)decomposition to each IMF component ci(t) respectively to obtain a time-frequency spectrum of each IMF component; and (3) carrying out mutual superposition to the time-frequency spectrum of each IMF component obtained in the step (2) to obtain the time-frequency spectrum of a high time-frequency resolution of the original digital signals x(t). The high-precision time-frequency analysis method disclosed by the invention avoids inaccuracy of time-frequency atoms when MP decomposition matching is implemented for composite signals directly, and MP decomposition is implemented for each IMF component, thereby being capable of obtaining the time-frequency spectrum with higher accuracy and time-frequency resolution and richer information.

The invention relates to a power transmission line short circuit fault diagnosis method based on empirical wavelet transformation and local energy. The method is characterized by comprising the following steps of: collecting short-circuit fault voltage signals; using an empirical wavelet transformation method to process the short-circuit fault voltage signals to obtain an empirical mode; performing short-circuit fault detection on the basis of empirical wavelet transformation; and performing short-circuit fault classification feature extraction on the basis of block local energy. The method has the advantages of scientificity, rationality, high time-frequency resolution, high noise resistance, high adaptability, good application effect and the like.

In an automotive system (300), a knock detection scheme is provided for detecting knock events in an internal combustion engine that are sensed by a transducer structure, such as a non-intrusive acoustic accelerometer sensor (310) to generate sensor signal information which is processed by a signal processing structure (312, 316, 318, 326) which extracts digital signal parameters from the sensor signal information to identify a predetermined pitch frequency (330) and any short-term energy increase (328) in the digital signal information which in combination are used to provide a positive indication of engine knock behavior. When a short term Fourier transform (324) is used to extract the digital signal parameters, time frequency resolution may be improved by appropriately windowing the digital signal being transformed.

The invention discloses a rolling bearing fault intelligent diagnosis method based on improved S-transformation and deep learning. According to the method, a window width adjustment factor is added toS-transformation, so that the window width of a Gaussian window function can be changed, and therefore, the time-frequency resolution of the S-transformation can be improved, and the S-transformationcan accurately detect an impact component in vibration signals, and thus, the fault characteristics of the vibration signals of a rolling bearing can be better extracted. Improved S-transformation isperformed on the vibration signals of the fault of the bearing, so that the feature matrix of the signals can be obtained; the feature matrix is subjected to column-based expansion so as be convertedinto a feature vector, and the feature vector is inputted into a sparse autoencoder model; on the basis of the characteristics of the encoder, the deep features of the data are further extracted, sothat some important hidden information that is not recognized by humans can be mined; the extracted features are accurately classified. With the method of the invention adopted, the accuracy of the fault diagnosis of the rolling bearing can be effectively improved.

The invention discloses a microseismic signal noise reduction filtering method based on a VMD and a wavelet packet and belongs to the signal processing technology field. In the invention, a mode of combining the VMD and the wavelet packet is adopted; adaptivity of a VMD decomposition method and characteristics of a high mathematical theory basis, high frequency noise suppression and the like in the algorithm are used; and the wavelet packet has the characteristics of carrying out multilevel division on a frequency band, further decomposing a high frequency portion which is not subdivided in multiresolution analysis, and adaptively selecting a frequency band according to a characteristic of an analyzed signal so that a time frequency resolution is increased. In the invention, based on a condition of maintaining randomness, non-stationary performance and a burst transient characteristic of a microseismic signal, the microseismic signal is filtered. The algorithm is simple and easy and an effect is ideal. Effective noise reduction filtering can be performed on a mine noisy microseismic signal. And a good technical value and an application prospect are possessed.

The invention provides a spectrum analysis method of a seismic signal and an apparatus thereof. The method comprises the following steps: acquiring the seismic signal; carrying out Hilbert transformation operation on the seismic signal to obtain an analytic signal of the seismic signal; carrying out WVD processing on the analytic signal; according to a result of the WVD processing, determining a Gauss window function and an attenuation coefficient; according to the determined Gauss window function and the attenuation coefficient, carrying out smooth pseudo-WVD distribution time frequency analysis processing on the result of the WVD processing. Through the method and the apparatus in the invention, a high time frequency resolution of seismic signal spectrum analysis can be obtained.

The invention discloses an estimation method of a multi-component non-stationary signal instantaneous frequency based on focusing S-transform, and mainly solves the problems that the multi-component non-stationary signal instantaneous frequency with high precision cannot be realized because the energy aggregation is not high and the time frequency resolution ratio cannot be flexible to adjust and cannot realize high precision. The estimation method comprises the realizing steps: 1, inputting a to-be-analyzed non-stationary signal; 2, optimizing a control parameter of a window function in the focusing S-transform according to an aggregation criterion; 3, calculating the optimized focusing S-transform and acquiring a signal time frequency distribution energy diagram; 4, carrying out binarization processing on the time frequency distribution energy diagram of the non-stationary signal; and 5, obtaining the instantaneous frequency estimation on a signal according to a binaryzation time frequency diagram after the binarization processing. According to the method provided by the invention, the self-adaption adjustment of the control parameter of the window function can be realized by improving the S-transform window function, and the time frequency distribution energy aggregation and the instantaneous frequency estimation accuracy can be improved, therefore the method can be used for radar interception, communication countermeasure, voice identification, and medical brain wave signal analysis.

The invention discloses a seismic spectrum imaging method based on deconvolution generalized S transform, which comprises the steps that a generalized S transform spectrum is acquired by performing two-dimensional convolution on Wigner distribution of original signals and a Gaussian window, a transform spectrum is acquired by performing generalized S transform on seismic data, and time-frequency distribution of the original signal can be acquired through deconvolution when the generalized S transform spectrum and the Wigner distribution of a window function are known. The seismic spectrum imaging method combines advantages of generalized S transform and Wigner-Ville distribution, generation of a cross term of the Wigner-Ville distribution is effectively suppressed through a generalized S transform window, and the generalized S transform spectrum is enabled to acquire high time-frequency aggregation at the same time; and deconvolution generalized S transform can adaptively adjust an analysis window along with variations of a frequency component, is applicable to time-frequency analysis for unstable seismic data, and can acquire high time-frequency resolution; and the seismic spectrum imaging method is applied to detecting the oil-gas possibility of a reservoir, thereby being conducive to improving the reservoir prediction accuracy.

A multirate spectral analyzer receives a data sequence through a tapped delay line (510) onto a plurality of filter channels, each of which includes a first decimator (520), a polyphase filter (530), and a second decimator (540). Each filter channel is coupled to a transform processor (550) which modulates the frequency response of the polyphase filters to a plurality of identical bandpass filters spaced evenly at 2πk/M intervals, where k=1, 2, . . . , M−1, and M is the number of filter channels. The transform processor (550) outputs a plurality of time series representative of the frequency component for each of the corresponding filter channels. By choosing a prototype filter order to be much larger than the number of filter channels, the effects of aliasing in the output are greatly reduced. The reduction in aliasing further allows, by prudent selection of the two decimators, for fractional temporal overlap to be implemented in the spectral estimates.

The invention provides an ECG signal processing method. ECG signals can be analyzed from a plurality of time-frequency resolutions, and SB-LSTM and TD-CNN are separately utilized to mines the overallfluctuation pattern and local fluctuation pattern of signals from ECG sub-layer signals, and a Sum Rule is finally used for fusing the intermediate classification results of each ECG sub-layer to obtain more accurate and robust classification results. The ECG signal processing method avoids individual differences and intra-individual differences caused by manual classification, and can obtain thestable classification results; and in addition, the ECG signal processing method does not need any expert knowledge, does not need to manually design various features, does not need to perform featureselection processing, does not need to separately construct classifiers, is a typical end-to-end method, and has the characteristics of convenient use, high classification accuracy and the like.

The invention aims to provide a method for oil and gas detection by utilizing seismic data instantaneous frequency attribute, which comprises the following steps of: acquiring and recording seismic initial data; carrying out generalized S transformation on the data in a frequency domain to obtain the time-frequency spectrum of each sampling point thereof, and determining the frequencies respectively corresponding to 15 percent and 35 percent of the total energy in the time-frequency spectrum as the initial frequency and the final frequency for computing low-frequency-band absorption attenuation gradient; after taking a natural logarithm from the time-frequency spectrum p (Tk, f) between the initial frequency and the final frequency and fitting by utilizing a least square method to obtain the low-frequency-band absorption attenuation gradient of the sampling point; and corresponding a position with strong low-frequency-band absorption attenuation gradient to a favorable oil and gas contained area, and determining an oil and gas detection result. The invention has lower requirements for signal-to-noise ratio, strong time-frequency resolution capability, simple algorithm and high computational efficiency.

The invention discloses a method for inhibiting cross terms in the time-frequency distribution of multi-component linear frequency modulation (LFM) signals based on subspace decomposition, which belongs to the technical field of signal processing. In the method, a time-frequency distribution matrix comprising noises and the cross terms is decomposed into signal subspaces and noise subspaces by utilizing singular value decomposition (SVD)-based subspace decomposition. For the problem that the linear modulation signals occupy relatively more bandwidths to make singular values are reduced at relatively lower rates and cannot be separated from the noises effectively, an angle of inclination of the time-frequency distribution of the signals can be obtained by utilizing Wigner-Hough transform. The time-frequency distribution of the signals is rotated according to the obtained angle so as to be parallel to a time base. The method is characterized in that: the singular values are rapidly decreased to zero, and then the signal sub-spaces can be separated effectively. By the method, the cross terms and the noises in the time-frequency distribution of the multi-component LFM signals are inhibited without reducing time-frequency resolution; therefore, the method is vast in application prospect.

An embodiment of the invention discloses an anomaly detection method of network flow, comprising: performing line frequency modulation wavelet transformation on detected flow signals, obtaining a transforming spectrum; dividing the transforming spectrum into at least two frequency spectrum coverages; performing signal reconstruction on each frequency spectrum coverage, operating the reconstructed signal through deviation fraction algorithm, obtaining deviation fraction; comparing the deviation fraction with presupposed threshold, determining whether the network flow is abnormal according to the comparing result. An embodiment of the invention also discloses a device realizing the method. The embodiment of the invention decomposes according to different time frequency resolution, thereby effectively detecting continuous anomaly flow for a long time and breaking anomaly flow for a short time, and detecting high frequency and middle frequency attack flow which are not detected by the network flow anomaly detection method based on wavelet transformation, through a second time detection after signal self-adapting reconstitution, anomaly character is further determined, and detecting accuracy is advanced.

The invention discloses a traveling wave time-frequency analysis method based on variational mode decomposition and Wigner-ville distribution. The method includes the following steps that: fault traveling wave signals are detected, and Karebauer phase-mode transformation is performed on three-phase voltage traveling wave signals, so that a traveling wave aerial mode component is obtained; variational mode decomposition is performed on the traveling wave aerial mode component, so that K intrinsic mode components are generated; Wigner-Ville analysis is performed on each intrinsic mode component;and the Wigner-Ville distribution of each intrinsic mode component is linearly superimposed, so that the time-frequency domain distribution of original traveling wave aerial mode signals is obtained.With the method of the invention adopted, the interference of cross terms in Wigner-Ville distribution can be effectively suppressed; a good noise suppression effect enabling good VMD (Variational Mode Decomposition) is preserved; time-frequency resolution enabling high Wigner-Ville distribution and good time-frequency aggregation are preserved; traveling wave time-frequency domain information characteristics are truly and accurately represented; and fault traveling waves can be completely observed. The method is of important theoretical and practical significance for the practical application of fault traveling wave protection and positioning.

The invention discloses an NMF-based micro-seismic weak signal recognition method. The method comprises the following steps of: firstly carrying out time frequency analysis on micro-seismic signals byadoption of S transformation, and rearranging time frequency spectrums in a frequency direction; factorizing a rearranged time frequency matrix by adoption of non-negative matrix factorization (NMF)so as to obtain a frequency domain base vector and a time domain position vector; extracting characteristic parameters such as sharpness, derivative quadratic sum, information entropy and sparseness to construct characteristic spaces of the micro-seismic signals; and finally classifying the micro-seismic signals by adoption of a least squares support vector machine (LSSVM). The method is capable of strengthening low-frequency weak signals and improving the time frequency resolution ratios, and has good time domain and frequency domain localization abilities.

The invention discloses a fault diagnosis method for an AC/DC charging equipment power device based on wavelet packet analysis, and the method comprises the steps: analyzing a charging equipment powermodule, and determining a fault class; obtaining output signals of the charging equipment power module under the normal condition and various fault conditions; carrying out the high-low frequency decomposition of the output signals, carrying out the multilayer wavelet packet decomposition of the high-low frequency signals obtained through decomposition, and carrying out the reconstruction of a wavelet packet decomposition coefficient; calculating the extracted signal energy at all frequency bands, carrying out the normalized calculation of the signal energy, and finally determining a centralized frequency band range of the signal energy; carrying out the power spectrum analysis of the reconstructed wavelet packet decomposition coefficient, determining the feature frequency and power spectrum value of a frequency band signal, comparing the comparison with the fault analysis, and determining the type of a fault; and carrying out the fault type coding of the type of the fault. The methodimproves the time-frequency resolution of a signal, achieves the precise fault positioning through the spectrum features, and improves the precision of the fault diagnosis of the AC/DC charging equipment power device.

The invention discloses a seismic data time-frequency analysis method based on second-order extrusion wavelet transformation. Firstly, a second-order synchronous extrusion operator is introduced on the basis of continuous wavelet transformation, the transformation is rearranged in a time-scale domain according to wavelet transformation coefficients, accordingly energy diffusion is pressed, and analysis results higher in time-frequency resolution ratio can be obtained by adopting the method. Compared with existing synchronous extrusion transformation, the method is more suitable for analysis of seismic signals having stronger frequency modulation, thus the signals are applied to time-frequency analysis of actual seismic data, the boundary of an underground carbon and hydrogen reservoir can be more accurately defined, riverways, faults and other geological information can be more elaborately depicted, and accordingly further seismic data explanation and well position determination are promoted.

The invention discloses a method for estimating a stratum equivalent quality factor based on synchronous extrusion wavelet transform, and the method is a quality factor extraction method based on a Qcompensation principle. By using synchronous extrusion wavelet transform with extremely high time-frequency resolution and energy aggregation to transform seismic data of adjacent horizons (post-stackseismic data) or adjacent seismic records (vertical seismic sections) to a time-frequency domain first, then inverse Q filtering is carried out in a preset certain Q value range, and by calculating amean square error between a result after time-frequency domain Q compensation and a reference waveform frequency spectrum, a corresponding quality factor when the minimum mean square error is satisfied is searched as the best equivalent quality factor, thereby realizing optimal estimation of the quality factor. The method for estimating a stratum equivalent quality factor based on synchronous extrusion wavelet transform avoids the difficulty of frequency band selection for estimating a quality factor in a conventional spectral ratio method, the hypothesis that a seismic source frequency spectrum is a Gaussian spectrum in a centroid frequency method or the like does not exist, methods for quality factor estimation are expanded, and the estimation accuracy of the quality factor is improved.