79 results about "Wavelet algorithm" patented technology

The invention belongs to the field of train fault diagnosis, and particularly relates to an online monitoring fault diagnosis system and method of a train power system. The system mainly comprises a train power device, a signal detection unit, a signal processing unit and a fault diagnosis unit. According to the method, the related state quantity of each device of the power system is obtained through sensors; a signal conditioning board of the power system conducts digital filtering and wavelet de-noising on signals so as to complete elimination of the noise signals and extract the useful signals; a wavelet algorithm is used for completing feature extraction of the wavelet energy spectrum, and the feature quantity of each subsystem is input into a corresponding sub-network to undergo fault diagnosis; the diagnosis results of all the sub-networks are fused at a central node, and fault diagnosis is completed initially; finally, fuzzy fusion is conducted on the diagnosis results of all the sub-networks so as to complete final positioning and diagnosis of faults. The online monitoring fault diagnosis system and method of the train power system have the advantages of being rapid, efficient, high in accuracy rate, high in reliability and high in network automation degree.

The invention discloses an electrocardiosignal denoising method based on morphological filtering and lifting wavelet transformation. Firstly, three times of decomposing are conducted on electrocardiosignals f based on the lifting wavelet theory, three layers of high-frequency coefficients and three layers of low-frequency coefficients are obtained, a lifting threshold denoising method is used for processing the high-frequency coefficients, the high-frequency coefficients on the bottom layer and the low-frequency coefficients are reconstructed twice, the reconstructed low-frequency coefficients can be obtained, morphological filtering is conducted on the reconstructed low-frequency coefficients, signal reconstruction is conducted according to the processed reconstructed low-frequency coefficients and the processed high-frequency coefficients on the highest layer, and denoised electrocardiosignals f' are obtained. The electrocardiosignal denoising method has the advantages that the method is simple and easy to achieve, the morphological algorithm and the lifting wavelet transformation algorithm are organically combined, compared with a traditional wavelet algorithm, the high-frequency electrocardio-noise and the low-frequency electrocardio-noise can be removed at the same time, the quality of the denoised signals is improved, computing is easy, the occupied space is small, and the electrocardiosignal denoising method can be achieved on hardware more easily.

The invention discloses a multi-source heterogeneous flight accident track data fusion method, relates to a flight accident data processing method and belongs to the technical field of data processing. The multi-source heterogeneous flight accident track data fusion method comprises the steps of firstly performing unified time reference correction on various data sources, then recognizing and eliminating abnormal flight data by using a wavelet algorithm, then performing data supplement on sparse flight parameter data by using a least square algorithm and flight dynamics model data, and finally performing heterogeneous fusion on a variety of heterogeneous flight data based on probabilistic graphical model method so as to generate a final flight accident track. Under the unified probabilistic framework, comprehensive data fusion is performed on the flight accident track by using different configurations and different sources of flight data and flight dynamics models of the corresponding aircraft models, and the estimation accuracy and the update frequency of the flight accident track are improved.

The invention relates to image fusion method based on direction filter group, which is used to fuse different sensor images and multi-focus image in the same scene. It uses atrous wavelet algorithm to decompose the source image and acquires a group of band pass subgraph, then the band pass subgraph is decomposed with direction filter group and it is represented with multi-resolution and multiple direction. Carries on fusion operation to theses subgraphs according to some fusion rules, and acquires a group of new fused subgraphs, finally acquires the fused image after reconstruction. The invention is better than traditional multi-resolution image fusion method.

The invention relates to the field of machine vision and image processing, in particular to a method for attaching a mobile phone cover lens to a liquid crystal display .The method comprises the steps of detecting and positioning the angular point of the cover lens through Canny edge detection and Hough transformation, detecting and positioning the angular point of the liquid crystal display through wavelet algorithm, Sobel edge detection and Hough transformation, calculating the actual three-dimensional coordinates of the cover lens and the liquid crystal display through coordinate system transformation based on the stereoscopic parallax principle, and then adjusting the cover lens and the liquid crystal display according to the three-dimensional coordinate difference value of the two to coincide in the X-Y plane, so that automatic attaching of the cover lens and the liquid crystal display is achieved under visual guidance .By the adoption of the method, automatic attaching of the cover lens and the liquid crystal display is achieved, attaching speed is high, efficiency and precision are high, processes are simple, no personnel limitation is set, manpower resource is saved a lot, cost is low, and adaptability is high.

The invention discloses an image detection method for the side-slipping motion of a vehicle, which is used for solving the technical problem of poor instantaneity in the existing detection method for the side-slipping motion of the vehicle. By adopting the technical scheme, the image detection method comprises the following steps of: obtaining an RGB (Red-Green-Blue) images containing side-slipping grinding crack and wheel angle information of a road vehicle by four-path CCD (Charged Coupled Device) image sensors fixed on a vehicle body, and carrying out denoising processing on the RGB images; cutting the two-path denoised front-wheel angle images according to the cutting mode of an image database; adopting a Canny operator to extract edges for the cut images; matching the edge images with edge images in the image database; matching marks for the nearest database images, i.e. the front-wheel angle value at the moment; adopting a wavelet algorithm to enhance the filtered road images ofrear-wheel side-slipping grinding cracks; adopting the Canny operator to extract the edges of the enhanced images; carrying out expansive operation on the edges; and calculating the number of pixel points in the horizontal direction to obtain side-slipping amount. By the image detection method, the difficulty of detecting real-time side-slipping amount in driving and turning of the vehicle is solved.

The invention relates to a traveling wave distance measurement device based on a passive magnetooptic glass current transformer principle. A small voltage signal reflecting primary side current amplitude of a transmission line is obtained by adopting a passive magnetooptic glass electronic current transformer, and a high-speed acquiring circuit is controlled by a CPLD (Complex Programmable Logic Device) for recording and sampling 625KHz travelling wave data and starting the sampling of 20KHz criterion data. When the transmission line fails, the travelling wave distance measurement device judges whether the transmission line fails through the sampled 20KHz data, and controls a 625KHz high-speed sampling part to completely record high-speed sampled data comprising travelling signals and sends data to a background computer through communication, and the background computer automatically computes and recognizes a failed travelling wave head through a wavelet algorithm after obtaining dual-end circuit data, and realizes failure location of the transmission line according to different moments of the wave head arriving at a dual-end bus.

The invention discloses an image fingerprint extracting method and checking method. The extracting method comprises the following steps: transforming the image into a plurality of grey square image blocks; using the wave filter to filter each image block according to the line and row by the convolution formula, it reads and arranges the grey result value of each image block of the result value array left part area; the positive value of the sequence is corresponding to 1, the negative value is corresponding to, 0, then it forms an image fingerprint of the total picture. The image fingerprint checking method dose, or exclusive or operation to the corresponding location of the two image fingerprints and ascertains weather the relative amount is identification by the, 1 value.

The invention discloses a drug detector standardization method based on the dual-tree complex wavelet algorithm. Firstly, multi-scale spectrum decomposition and reconstruction are conducted with the dual-tree complex wavelet, then each layer of reconstructed spectrum is corrected through piecewise direct standardization, predication models based on partial least squares and leave one cross validation are established, and finally all the predication models are fused and evaluated according to a calculated weight. According to the dual-tree complex wavelet and direct standardization combined method, due to the translation invariance and multi-scale property of the dual-tree complex wavelet, the defects of existing model transfer methods are overcome, the capacity of correcting drifting in the X-axis direction and Y-axis direction is excellent, and the method also has the advantages of being exquisite, precise and efficient and can be widely applied to the fields including near-infrared and Raman spectrum.

The invention provides an a<`> trous based wavelet threshold denoising method which is applied to high-throughput gene sequencing image denoising. The denoising method provided by the invention is mainly an improvement for a hard threshold based isotropic un-decimated discrete wavelet algorithm. The specific operations are that wavelet decomposition is carried out on the high-throughput gene sequencing image by using specific wavelets, and utilization of an l1 norm is proposed to calculate a global threshold in allusion to each layer of wavelet coefficient. A wavelet coefficient expression is constructed through the global threshold and each layer of wavelet coefficient, and finally, a denoised image is acquired by using a wavelet reconstruction algorithm. Compared with the prior art, the method provided by the invention has the advantage of better denoising performance. The rationality and the robustness of the denoising method are proved through comparison of a plurality of groups of experimental data.

The invention provides a multi-focus multi-source image fusion method. The method inclues: carrying out decomposition, fusion and superposition processing on the source image with the M focus points;decomposing the source image into a base part and details; filtering and denoising the base part by adopting an even number complex wavelet algorithm to obtain a fused base part; a pre-training modelVGG-S is adopted to perform depth feature extraction on details to obtain detail features, a multi-layer fusion strategy is adopted to reconstruct the detail features, then a gradient maximum value isselected for the detail features to obtain fusion details, and finally the obtained fusion base part and the fusion details are superposed to complete image fusion. The fusion image obtained by the method not only reserves the feature information of the image before fusion and improves the utilization rate of the effective information of the image, but also has higher definition, and is more detailed, comprehensive and superior. The method is wider in application range and can provide more image information in the aspects of daily life, medicine, military affairs and the like.

The invention discloses an on-line monitoring management system of GIS ultrahigh frequency partial discharge. An existing various GIS monitoring system is limited to the diagnosis of defects of the GIS ultrahigh frequency partial discharge, and ignores the unified management of GIS equipment. The on-line monitoring management system of the GIS ultrahigh frequency partial discharge comprises a signal collecting subsystem, a data processing analyzing subsystem and an equipment management subsystem. According to the signal collecting subsystem, a partial discharge signal of the GIS equipment is received through a UHF sensor, the partial discharge signal is converted to a voltage signal, and then the voltage signal is enlarged by a signal-amplifier, data are transmitted to a collecting controlunit to be stored the data, and the data are sent to an upper computer main control platform through an optical fiber. The data processing analyzing subsystem uses wavelet algorithm to carry out noise reduction process on the partial discharge signal, and realizes spectrogram display and partial discharge warning. The on-line monitoring management system of the GIS ultrahigh frequency partial discharge integrates fault monitoring diagnosis and maintenance management of the equipment, can intelligently analyze the operating condition of electrical equipment and possible discharge defect types,and meanwhile, can manage the equipment conveniently.

The invention relates to a fault line selection method based on a wavelet analysis. Sudden changing point transient signals are decomposed and extracted through sudden changing characteristics of a zero sequence current in a multi-resolution mode when single-phase faults appear on an outgoing line and a wavelet algorithm, five wavelet functions are selected to conduct multi-scale decomposition, if a plurality of high-frequency scale decomposition results are larger than a set value, and the number of the results is larger than a set number, single-phase grounding faults appear on the outgoing line. According to the method, fault positioning can be fast achieved, the automation degree of a power distribution network can be greatly improved, the power failure time of a user can be greatly shortened, losses caused by power failure can be reduced, the defect of a small-current grounding operation mode is remedied and good social and economic benefits are achieved.

The invention relates to a wavelet algorithm-based electrocardiogram (ECG) signal transmission method and a system, wherein the optimized real-time compression for ECG signals which is adaptable to the display size of a mobile terminal, a link real-time transmission rate and a segmentation percent root-mean-square difference (PRD) performance is realized through acquiring and processing the ECG signals, and taking the display size of the terminal, the link real-time transmission rate and the PRD performance as self-adaptive control parameters in a transmission process. Based on obtaining a high compression ratio and controllable fidelity, the method acts according to actual circumstances, so that the calculation load for a 3G terminal to reconstruct waveforms is lowered, the terminal resources are saved, the link resources are sufficiently utilized, the transmission delay is shortened, and a further guarantee is provided for realizing real-time monitoring.

The invention discloses a wavelet transform based extreme-low bit-rate video compressing and coding method. The wavelet transform based extreme-low bit-rate video compressing and coding method includes three steps: (1) subjecting a GOF (group of frame) image group to three-dimensional wavelet decomposition, selecting appropriate decomposition grades for time decomposition and space decomposition according to the size of the image group and formats including CIF (common intermediate format), QCIF (quarter common intermediate format), SIF (standard intermediate format) of videos, decomposing to obtain a coefficient matrix of wavelet decomposition; (2) subjecting the coefficient matrix to branching and blocking operation: firstly subjecting the coefficient matrix to first-grade virtual decomposition or second-grade virtual decomposition, and then blocking the coefficients after virtual decomposition; (3) realizing SPIHT (set partitioning in hierarchical tree) coding process. By the aid of the wavelet transform based extreme-low bit-rate video compressing and coding method, a conventional video wavelet 3D-SPIHT (three-dimensional set partitioning in hierarchical tree) algorithm is improved, a wavelet transformation coefficient is organized in a form of combining a virtual tree with a coefficient branching organization, a multi-threshold sub-band coding optimization method is utilized, important coefficient is coded preferentially, the purpose for displaying more effect information with few bit numbers is achieved, compression ratio can be improved, and compression efficiency is increased.

The invention relates to a region growing compression method based on quaternary tree division. A region growing template defines the clustering properties of wavelet coefficients. The method overcomes the defects that a wavelet algorithm zero-tree has low forecasting efficiency and does not fully utilize the correlation among the wavelet coefficients of the same layer; and the algorithm fully considers the correlation between the wavelet coefficients of layers and the wavelet coefficients of the same layer and the clustering properties of the wavelet coefficients, the structure is simple, the algorithm has good compression performance, and the compression performance of the algorithm is superior to that of the classical SPIHT and EZW algorithms.

The invention relates to an early waning method for a rolling bearing of a rotating machine. A vibration signal generated when the rolling bearing is stably operated is a stable low frequency signal, a signal generated in time of failure of the bearing is a high frequency pulse signal, a vibrating signal is acquired, the sudden change high frequency pulse point position in the signal is extracted by utilizing a wavelet algorithm and is output through early warning, movement of the rotating machine can be intervened in time, and safety of the rotating machine is ensured. The rolling bearing of a high speed train is monitored in real time, and safety and stability of operation of the train are effectively improved, so that early waning method has important practical significance to ensuring life safety of passengers. Vibration monitoring is conducted on a generator set of a nuclear power station, the safety factor of nuclear electricity generation is greatly improved, and nuclear leakage and pollution produced by rotating vibration are avoided. The bearing of a rotating traction mechanism in the production process is monitored in real time, production continuity of enterprises can be improved, and economic loss is avoided.

The invention provides a diesel engine combustion resonant frequency extraction method based on empirical wavelet and Hilbert transform. The method comprises the following steps that: 1: carrying outFourier transform on a cylinder pressure signal to obtain a cylinder pressure frequency spectrum, utilizing a scale space algorithm in an empirical wavelet algorithm to identify characteristic sub-modality in the cylinder pressure frequency spectrum, determining the position of the sub-modality, and calculating the boundary frequency of the sub-modality; 2: constructing an empirical wavelet filter, utilizing the constructed wavelet filter to carry out band-pass filtering on the cylinder pressure signal, extracting a sub-modality signal corresponding to each piece of combustion fluctuation in the cylinder pressure frequency spectrum, and selecting the sub-modality with maximum oscillation energy as a combustion oscillation signal; and 3: carrying out the Hilbert transform on the combustionoscillation signal obtained in S2, drawing a Hilbert spectrum, and obtaining oscillation frequency and an oscillation amplitude corresponding to the oscillation frequency. The method has the advantages of good astringency, high adaptivity and high accuracy, and the obtained oscillation frequency can be used for the low-noise optimal design of the diesel engine and temperature and gas quality estimation in a cylinder.

The invention relates to the technical field of high voltage cable fault detection, and more particularly to a non-contact method for acquiring a transient voltage traveling wave signal. The method comprises: determining spatial electromagnetic signal distribution near a cable terminal according to the spatial electromagnetic distribution of the cable terminal and the high-frequency signal propagation characteristics; selecting a physical structure, electromagnetic parameters and installation position of an acquisition unit to install the acquisition unit; transmitting a fault signal collectedby the acquisition unit to a data processing unit through a signal cable; using the data processing unit to analyze and process electric quantity signals when collecting high voltage cable faults, obtaining the modulus maximum point by a wavelet algorithm to determine the fault start time. The non-contact method for acquiring the transient voltage traveling wave signal is based on the principle of space electromagnetic induction, can obtain the transient voltage traveling wave signal in the fault signal of the high-voltage cable fault and obtain accurate fault occurrence time without physicalcontact with the high-voltage cable, and meet high voltage cable fault online positioning requirements based on double-end traveling wave principle.

The invention discloses a reactor neutron noise spectrum calculation method based on an improved wavelet algorithm. Wavelet node dispersion is performed on a reactor neutron noise channel signal, frequency aliasing during a node acquisition process is eliminated, then the signal of a wavelet disperse node is rebuilt to acquire a reactor noise channel sub-frequency band signal, and at least spectrum calculation is sequentially performed on each sub-frequency band signal of a reactor noise channel to acquire frequency band spectrum information. According to the reactor neutron noise spectrum calculation method based on the improved wavelet algorithm provided by the invention, fine division of each frequency band and local feature extraction of a reactor neutron noise spectrum can be better reflected, analysis capacity and analysis accuracy are greatly improved, an input condition is provided for operation monitoring and fault diagnosis of a reactor, and a significant improvement significance is provided for safe operation of the reactor and building of an expert system.

The invention discloses a method for analyzing level signals. The method comprises the following steps: firstly, performing wavelet analysis on the level signals to find out signal mutation points; secondly, performing discretization on a sampled value of the signal mutation points by using a Fourier algorithm to calculate an effective value of the sampled value; and finally, comparing the effective value with a preset fault protection threshold value, if the effective value does not exceed the threshold value, determining that the signal mutation points are not fault points, and returning to a wavelet algorithm to continuously detect the mutation points, and if the effective value exceeds the threshold value, determining that the signal mutation points are fault points, and outputting corresponding signals. The invention also provides a method for triggering a release on the basis of the method for analyzing the level signals. According to the methods disclosed by the invention, the wavelet analysis and Fourier transformation are combined together to detect the fault points of the signals, so that the methods have the advantages of high reaction speed to faults and high sensitivity and real-time performance, thereby greatly improving the releasing performance of the release.

The invention discloses a wavelet algorithm based vehicle tracking recognition method. The method comprises the steps of S1, acquiring a real-time image; S2, updating the background of the image, and detecting an image difference portion by using a method of background subtraction; and S3, processing the image by using a wavelet algorithm, firstly performing binarization on the image, then performing expansion erosion treatment, then building an attitude optimization function by adopting the wavelet algorithm, performing location tracking so as to precisely capturing vehicle information location, and accurately tracking and recognizing a vehicle. According to the invention, a function of background updating is utilized, detection is performed on the foreground with great differences, and a background updating system is enabled not to be disturbed by small noises; feature recognition is performed, further identification is performed after the vehicle color is found out, and after a special vehicle is locked, the location of an intersection camera is utilized to act as the time so as to lock the vehicle; and various vehicles can be tracked at the same time, and the method can be applied to multi-intersection detection and real-time detection.

The invention discloses a power distribution network fault line selection method based on an optimal FIR (finite impulse response) filter and hierarchical clustering. The power distribution network fault line selection method includes design of the optimal FIR filter, a hierarchical clustering algorithm and a fault line selection method based on the optimal FIR filter and hierarchical clustering; single-phase earth fault line selection in a small current grounding system is achieved according to the hierarchical clustering algorithm by the designed optimal FIR filter. The power distribution network fault line selection method has the advantages that the FIR filter is used for signal processing, data sampling frequency only needs to reach above 6kHz and is lower than sampling frequency required by a wavelet analysis based fault line selection method, and system resource is saved; the clustering algorithm is adopted, fault feeder lines are determined according to pedigree clustering evaluation indexes, and reliability of line selection results is guaranteed; faulty lines can be detected effectively under various fault conditions by the power distribution network fault line selection method, so that the power distribution network fault line selection method has higher reliability than a wavelet algorithm.

The invention relates to a partial discharge pulse extraction method based on noise reduction clustering, which is based on a wavelet algorithm and comprises the following steps: collecting an original power generation signal; determining a wavelet decomposition coefficient of each level; extracting the wavelet decomposition coefficient of each level; clustering the extracted wavelet decomposition coefficient pulse segments of each level to obtain sub-sequences of the aliasing pulse sequence, the sub-sequences corresponding to independent signals of different discharge sources or different propagation paths. Based on the diversity and complexity of the current power equipment partial discharge detection environment interference, the signal-to-noise ratio improvement and multi-discharge clustering calculation of the interference signal are realized in real time in the detection process, and the anti-interference performance of the partial discharge detection device is practically improved.

The invention belongs to the technical field of geophysical logging, and a rock type identification method based on the logging curve wavelet Mallet algorithm. The rock type identification method comprises the following steps: step 1, preprocessing an original logging curve acquired in the field, step 2, analyzing logging response characteristics of the logging curve preprocessed in the step 1 todifferent lithology, and determining the logging curve sensitive to the reflection of the different lithology; 3, carrying out normalization processing on the sensitive logging curve L determined in the step 2; 4, performing Mallet decomposition of different wavelet functions and different scales on the normalized logging curve L1 in the step 3, and determining an optimal wavelet function and a decomposition scale for identifying lithology; and 5, performing lithology identification on the drill hole with the unknown lithology type according to the optimal wavelet function and the optimal decomposition scale determined in the step 4, and determining the lithology type of each interval in the drill hole.

A method for compressing / decompressing photo of certificate holder includes analyzing data of image containing head photo to support compressing / decompressing image data of interesting region, applying conformal wavelet algorithm for enabling to store color or black / white image with high compression ratio and enabling to carry out high quality of decompression on it. In the method, it is easy to store head photo data of IC cardholder into card with form of IC card data.

A wind speed prediction method is disclosed. The wind speed data sequence for prediction can be determined, the wind speed data sequence can be decomposed and reconstructed by a wavelet algorithm, andeach sub-sequence can be predicted by the neural network optimized by a cuckoo algorithm which is trained beforehand, so that a plurality of sub-prediction results can be obtained, and finally the final prediction results can be determined according to each sub-prediction results. Therefore, according to the characteristic that wavelet algorithm can reduce the fluctuation of wind speed data, thewavelet algorithm is used to reduce the high nonlinearity of wind speed data sequence, and the cuckoo algorithm is used to optimize neural network so as to effectively avoid the problem that neural network is trapped into local optimization, and finally improve the prediction accuracy. The invention also provides a wind speed prediction device and apparatus and a computer-readable storage medium,the function of which corresponds to the method.

The invention relates to a single-phase grounding fault line selection method based on the relativities of the wavelet characteristic quantities after a fault. The method includes the following steps:a circuit measurement terminal acquiring the data of 8 paths of signals in the course of two sampling cycles after a fault, wherein the signals include three-phase and zero-sequence currents and voltages; performing wavelet transformation on the data in the course of two sampling cycles to obtain one group of waveform characteristic quantities; analyzing the voltages and the currents by using different wavelet algorithms, obtaining a total of 6 groups of waveform characteristic quantities, analyzing the relativity of each group of characteristic quantities, and comparing the relativity with asetting value; and if the number of the circuits is greater than a set number and the result does not conform to the setting value, determining that an outlet has a single-phase grounding fault. According to the invention, the method has the following beneficial advantages: the method herein makes determinations by using multiple wavelet analysis algorithms, avails the high sensitivity of the wavelet algorithms to discontinuities, can avoid wrong determination which might be caused due to the use of only one wavelet algorithm, and can accurately determine that the outlet has a single-phase grounding fault by analyzing the relativities of the characteristic quantities.

The invention provides a method for increasing wavelet packet decomposition speed based on Mallat algorithm, relates to the technical field of signal treatment, and solves the problems that in the prior art, in the process of extracting characteristics, wavelet decomposition is adopted for only decomposing low frequency signals, and information about the low frequency signals can lose. The methodcomprises four steps of selecting wavelet decomposition types and decomposition layer numbers; calculating signal frequency; through a backcasting reckoning method, obtaining all decomposed signals; and reconstructing wavelets so as to obtain useful target frequency signals. The Daubechies wavelets algorithm is adopted, the wavelets and the decomposition layer number thereof are selected, a Shannon sampling theory is utilized, so that the frequency range of each fragment after wavelet packet decomposition can be further obtained. An algorithm for combining semi-wavelet packets and an improvement convolution process is adopted, so that the defect that wavelet decomposition can only realize decomposition of the low frequency signals, is overcome, and redundant decomposition of wavelet packets is also solved; and therefore, the signal decomposition precision is improved, the complexity of the algorithm is reduced, and the effect of accelerating wavelet decomposition is achieved.