72 results about "Adaptive wavelet" patented technology

When making digital data recordings using some form of computer or calculator, data is input in a variety of ways and stored on some form of electronic medium. During this process calculations and transformations are performed on the data to optimize it for storage. This invention involves designing the calculations in such a way that they include what is needed for each of many different processes, such as data compression, activity detection and object recognition. As the incoming data is subjected to these calculations and stored, information about each of the processes is extracted at the same time. Calculations for the different processes can be executed either serially on a single processor, or in parallel on multiple distributed processors. We refer to the extraction process as “synoptic decomposition”, and to the extracted information as “synoptic data”. The term “synoptic data” does not normally include the main body of original data. The synoptic data is created without any prior bias to specific interrogations that may be made, so it is unnecessary to input search criteria prior to making the recording. Nor does it depend upon the nature of the algorithms / calculations used to make the synoptic decomposition. The resulting data, comprising the (processed) original data together with the (processed) synoptic data, is then stored in a relational database. Alternatively, synoptic data of a simple form can be stored as part of the main data. After the recording is made, the synoptic data can be analyzed without the need to examine the main body of data. This analysis can be done very quickly because the bulk of the necessary calculations have already been done at the time of the original recording. Analyzing the synoptic data provides markers that can be used to access the relevant data from the main data recording if required. The nett effect of doing an analysis in this way is that a large amount of recorded digital data, that might take days or weeks to analyze by conventional means, can be analyzed in seconds or minutes. This invention also relates to a process for generating continuous parameterised families of wavelets. Many of the wavelets can be expressed exactly within 8-bit or 16-bit representations. This invention also relates to processes for using adaptive wavelets to extract information that is robust to variations in ambient conditions, and for performing data compression using locally adaptive quantisation and thresholding schemes, and for performing post recording analysis.

The invention provides an image denoising method based on a self-adaptive wavelet threshold and a two-sided filter and aims to improve the effect of a wavelet threshold denoising algorithm and better protect the edge and the detailed information of an image. The algorithm comprises the following steps of decomposing the image by adopting a discrete wavelet to obtain a plurality of sub-bands and the wavelet coefficients of the sub-bands; selecting a threshold which is self-adaptively changed along with the changes of wavelet-decomposing scales and the sub-bands, and carrying out quantitative threshold processing by adopting a soft threshold function; carrying out inverse wavelet transformation to obtain a reconstructed image; filtering the reconstructed image by adopting the two-sided filter so as to obtain the clear image. According to the image denoising method, wavelet threshold denoising is carried out by utilizing the threshold which is self-adapted to the wavelet-decomposing scales and the sub-bands, and filtering is carried out by combining the two-sided filter, so that through the designed denoising algorithm, not only can white gaussian noise be effectively removed, but also the edge and the detailed information of the image can be well reserved.

The invention brings forward a self-adaptive wavelet threshold image de-noising algorithm and device. The image de-noising algorithm comprises the following steps: a noised image is subjected to wavelet transformation operation, and wavelet coefficients of all layers can be obtained; with signal correlation considered, coefficients in an area adjacent to each coefficient are averaged in wavelet coefficients of each layer; threshold is determined based on a wavelet coefficient which is obtained via an absolute mean value estimation method, and a self-adaptive threshold method is adopted for determining thresholds suitable for all different scales; as for the wavelet coefficients and thresholds, self-adaptive threshold functions for all directions at all layers are constructed, wavelet inverse transformation and reconstruction are performed, and a de-noised image can be obtained. According to the image de-noising algorithm, the self-adaptive threshold method is adopted for determining the thresholds, an overall uniform threshold is replaced with thresholds for different scales, wavelet threshold de-noising operation is performed via use of the self-adaptive thresholds and the self-adaptive threshold functions, and detailed information of the image can be protected; the self-adaptive wavelet threshold image de-noising algorithm is better than a conventional wavelet threshold de-noising algorithm in terms of peak signal to noise ratio and visual perception.

The invention discloses a wrist watch type multi-parameter biosensor and relates to a self-adaptive detection method for physiological parameters and a multi-parameter intelligent monitoring wrist watch by the adoption of the method. The method comprises the steps that (1) human physiological parameter signals are collected; (2) the human physiological parameter signals are subjected to self-adaptive wavelet decomposition, and feature predictors and updaters are selected point by point in the decomposition process; (3) waveform analysis is conducted, interference signals in the decomposed signals are eliminated, and human physiological parameter signals without interference are obtained. The monitoring wrist watch comprises a base body, wherein the base body is provided with a central processing unit, a physiological parameter sensor used for collecting the human physiological parameter signals and various other sensors. The central processing unit processes data by the adoption of a self-adaptive wavelet decomposition method to obtain the human physiological parameter signals without interference. The wrist watch type multi-parameter biosensor effectively solves the problem of instability caused by large contact resistance generated when a daily wearing mode is adopted. The wrist watch type multi-parameter biosensor is convenient to use and can collect various human physiological signals, movement signals, environment signals and the like in real time.

The invention relates to a self-adaption wavelet neural network abnormity detection and fault diagnosis classification system and method, which can be applied to the fields, such as economic management abnormity detection, image recognition analysis, video retrieval, audio retrieval, signal abnormity detection, safety detection, and the like. The system comprises the following seven parts: an acquisition device, a transmitter device, an A / D (Analog / Digital) conversion device, a self-adaption wavelet neural network abnormity detection and fault diagnosis classification processor, a display interaction device, an abnormity alarm device and an abnormity processing device. The abnormity detection and fault diagnosis classification object of the self-adaption wavelet neural network abnormity detection and fault diagnosis classification system is acquired from samples for which a self-adaption mechanism is automatically established by the self-adaption wavelet neural network of a system to be detected, the characteristic information of a signal can be effectively extracted through wavelet transform multi-scale analysis, and a more accurate abnormity detection and fault diagnosis locating result can be obtained. The device adopting the method has the advantages of generalization, high accuracy in the application field, capability of real-time monitoring and low cost.

The invention provides an online partial discharge detection signal recognition method of a cable. The method comprises the following steps: a partial discharge time domain waveform of a known source is acquired; a partial discharge waveform time domain waveform sample library is established; a wavelet packet is used to denoise each partial discharge time domain waveform in the sample library; a self-adaptive wavelet neural network model with a preset number of layers is constructed; according to each denoised partial discharge time domain waveform, a PSO algorithm is used to train the constructed self-adaptive wavelet neural network model; a BP algorithm is then used to train the trained self-adaptive wavelet neural network model for a second training so as to get a well-trained wavelet neural network; and the partial discharge signal of a to-be-recognized source is received, the well-trained wavelet neural network is inputted for recognition, and the source of the to-be-recognized partial discharge signal is obtained. The method of the invention is fast in recognition speed and high in recognition precision.

The invention discloses noise inhibition which is an important processing step in a micro-seismic signal processing process. Complete ensemble empirical mode decomposition CEEMD and wavelet transformation WT are widely applied to seismic noise elimination; however, the CEEMD is lack of theoretical foundations and the self-adaptability of the WT is relatively weak. Therefore, the noise eliminationeffect is poor. According to the invention, experience wavelet transformation (EWT) is combined with various threshold functions to carry out micro-seismic noise elimination for the first time. The EWT is used for establishing a self-adaptive wavelet filter group through spectrum segmentation to extract different frequency blocks of a detected signal. In the EWT, four types of spectrum segmentation methods are adopted; an experiment finds out that an adaptive algorithm can be used for separating an effective signal and noises of micro-seismic data very well; after the EWT is carried out, the signal can be divided into two components through analyzing a spectrum and energy of each module. A hard threshold function is applied to the component containing more effective signals and an improvedthreshold function is applied to the component containing less effective signals. An extraction method is compared with the CEEMD and the WT in an analogue signal and an actual signal to prove the effectiveness of a provided method.

The invention discloses a method for extracting a heartbeat signal based on radar echo strong noise background. The method includes the following steps that: S100: the radar echo is transmitted to a data preprocessing end in the form of a data frame; S200, an original sequence of the radar echo is preprocessed, the echo of a stationary target is filtered out, and an echo signal of a distance unitof a human target is obtained; S300, based on the acceleration, the random motion of the human body is determined, which is used for reducing the error of the estimation of the sign signal parametersat a later period; S400, signal separation of signs can be separated, and an adaptive wavelet scale selection algorithm is employed for realizing effective separation of respiratory signals and heartbeat signals of vital signs; and S500, time-domain peak finding and downsampling processing are carried out on a respiratory signal and a heartbeat signal, and finally the respiratory frequency and theheartbeat frequency can be obtained. The invention also discloses an extraction system for the heartbeat signals based on the radar echo strong noise background. The system fully eliminates noise interference, recognizes useful heartbeat and respiratory signals, and has the advantages of high accuracy and strong real-time performance.

The invention relates to a diagnosis method for impact type failure between a rolling bearing and a gear based on an optimal self-adaptive wavelet filter, which comprises the following steps: firstly establishing an exponentially damped sinusoidal impact type failure signal model, using a mode self-adaptive wavelet generation algorithm publicly disclosed by Hector Mesa to generate a self-adaptive wavelet filter which is matched with the signal model, then utilizing a fast FIR (finite impulse response) filtering algorithm to execute the wavelet filter to obtain the filtering result, then calculating the kurtosis value of the filtering result, and using an evolution differential algorithm to repeat the steps so as to finally obtain an optimized envelope spectrum. By adopting the method, impact failure characteristics in vibration signals can be precisely extracted, and a clearer envelope spectrum can be given out, thereby ensuring that failure symptoms can be displayed more clearly.

The invention provides a rotary machinery fault feature extraction method based on self-adaptive wavelet energy. According to the invention, firstly rotor vibration signals are collected on a machinery fault comprehensive simulation experiment table, and then for the problem that signal energy features are influenced by the rotating speed and the sampling frequency, a new self-adaptive wavelet energy feature extraction method is adopted to complete effective feature extraction of a rotor imbalance fault, a misalignment fault and a bearing pedestal loosening fault of a rotor of the rotary machinery. In addition, in order to reserve more original features of the rotary machinery vibration signals and overcome the problems of a conventional denoising method that constant deviation and discontinuity exist at the threshold and parameters are not adjustable, a novel threshold function is embedded in the method for carrying out denoising. By adopting the method, a relatively good effect is obtained on the machinery fault simulation experiment table, and different fault types can be effectively distinguished by the extracted energy features.

The invention discloses a self-adaptive wavelet threshold de-noising method based on neighborhood correlation. The method includes the following steps that (1) wavelet transformation is performed on images with noise to obtain wavelet coefficients; (2) according to the wavelet coefficients, self-adaptive threshold functions of each layer of wavelet coefficients are constructed, and wavelet threshold values of an ith decomposition layer are selected; (3) attenuation coefficients are selected by the utilization of a mid-point method, and threshold processing is performed by the adoption of the threshold functions and the wavelet threshold values; (4) wavelet inverse transformation is performed on the wavelet coefficients, corresponding to threshold function processing, of the selected attenuation coefficients to obtain restored original signal estimation values; (5) PSNR values of the original signal estimation values are worked out to obtain an optimal value, an optimal attenuation coefficient under the PSNR optimal value is obtained according to the mid-point method, and the wavelet coefficients corresponding to the threshold function processing are reconstructed, and obtained estimation values are used as final de-noising images. According to the method, the defects of hard threshold and soft threshold de-noising methods are overcome, more accurate wavelet coefficient estimation values are obtained, and the edges of the images are protected.

A hybrid filtering method called Karhunen Loeve Transform-Wavelet (KW) filtering is presented to de-noise dynamic cardiac magnetic resonance images that simultaneously takes advantage of the intrinsic spatial and temporal redundancies of real-time cardiac cine. This filtering technique combines a temporal Karhunen-Loeve transform (KLT) and spatial adaptive wavelet filtering. KW filtering has four steps. The first is applying the KLT along the temporal direction, generating a series of “eigenimages”. The second is applying Marcenko-Pastur (MP) law to identify and discard the noise-only eigenimages. The third applying a 2-D spatial wavelet filter with adaptive threshold to each eigenimage to define the wavelet filter strength for each of the eigenimages based on the noise variance and standard deviation of the signal. Lastly, the inverse KLT is applied to the filtered eigenimages to generate a new series of cine images with reduced image noise.

A method for processing an image is described. In the method, image data representing an image into a memory device. The image data is filtered to obtain a plurality of coefficients representing low and high frequency image data. An area of low frequency data corresponding to one of the coefficients is analyzed to identify a degree of edginess of the region. A threshold is obtained, the threshold varying depending on the degree of edginess. If the coefficient is less than the threshold, it is reduced to zero. Wavelet-based image compression can then be performed on the image using the reduced coefficients.

The invention discloses an island type micro-grid energy storage optimized configuration method using wavelet packet decomposition and considering frequency response. Adaptive wavelet packet decomposition serves as a frequency division method of imbalance power, characteristics of an energy-storage battery, a super capacitor and system frequency offset consumption capability are taken into consideration comprehensively, the imbalance power is divided into low-frequency, intermediate-frequency and high-frequency components, the components are stabilized, a mixed energy-storage annular cost value model is established, and an optimal energy-storage configuration result is determined on the basis of the lowest cost. According to the invention, the island micro-grid cost model is established bytaking the cost of an energy-storage system into consideration, configuration of the energy-storage capacity is closely related to the wind abandoning cost and the auxiliary frequency modulation income, and further, a more comprehensive and accurate energy-storing cost model can be established.

The invention discloses a method for effectively encoding and decoding a JPEG2000 image based on a vision potential attention target area. The method comprises the following steps of: before the image is encoded, dividing the image into a plurality of Tile blocks based on the potential target area of different attentions and pre-assigning corresponding target code rates according to vision potential attention target area extraction technology; and then carrying out code stream organization and packaging to realize image compression through quantization, adaptive wavelet transformation, and a bit plane encoding and arithmetic encoder code word reset mechanism on the basis of potential attention target area quality layer code rate control. When network band width is insufficient, the image reconstruction quality of a more-concerned target area can be exchanged at the expense of part of code rates of the target area with smaller potential attention; simultaneously, retractable high-efficiency image encoding and decoding technology is combined with a qualified content based on the potential attention target area.

The invention provides an L1 adaptive ship power positioning double-loop control system based on a neural network. The control system comprises a marine ship, an outer ring position tracking loop controller, an L1 adaptive inner loop controller based on a neural network and a path generator. The outer ring position tracking loop controller can track a reference path of the system to obtain a virtual speed instruction; the L1 adaptive inner loop controller based on a neural network comprises an adaptive wavelet neural network approaching device, a state predictor, a parameter adaptive law and an L1 control law; the L1 control law comprises a filter; and the L1 adaptive inner loop controller based on a neural network solves the problem of uncertainty in the system, and approaches a coupling term in the system using a recursive wavelet neural network. The L1 controller introduces a low-pass filter into the feedback loop during feedback, thereby weakening the influence of unknown high-frequency noise in control signals on high-frequency noise. The control system can be applied to control of other ship power positioning systems.

A method is presented for visually enhancing at least one geological feature in 3D seismic survey data, comprising the steps of:(a) selecting at least one first seismic trace from a 3D seismic survey data set;(b) subdividing said at least one first seismic trace into a plurality of identified characteristic segments;(c) generating at least one first analytical model function for each of said plurality of identified characteristic segments, utilizing at least one adapted wavelet from an existing dictionary;(d) determining a matching characteristic between said at least one first seismic trace and said at least one first analytical model function;(e) optimising said at least one first analytical model function with respect to said matching characteristic;(f) repeating steps (d) and (e) until a predetermined condition is met, and(g) generating a model dataset from said optimised at least one first analytical model function for at least part of said at least one first seismic trace for visual representation.

The present invention discloses an adaptive wavelet threshold function image noise suppression method, and belongs to the gray image noise processing technology field. The method of the present invention comprises the following realization steps of 1 importing and optionally selecting a gray image containing the Gauss noise; 2 carrying out the wavelet decomposition on a noise image, and selecting an appropriate wavelet and determining the decomposition level, and then carrying out the wavelet transform to obtain a set of wavelet coefficients; 3 carrying out the threshold quantification processing on the wavelet high frequency coefficients after the wavelet decomposition to obtain a set of wavelet coefficients after the threshold processing; 4 reconstructing the wavelet coefficients after the threshold processing, thereby obtaining a noise suppression image. The method of the present invention solves the disadvantage that a hard threshold function is not continuous at a threshold, at the same time, enables the disadvantage that an error of a soft threshold function estimation coefficient is increased, to be improved correspondingly, and enables the image peak signal to noise ratio to be improved effectively, an image mean square error to be reduced and an image restoration effect to be improved better.

The invention relates to an adaptive wavelet neural network denoising modeling method based on forward linear prediction, comprising the following steps: (1): using wavelet transform to perform multi-scale decomposition on the zero-drift signal of the fiber optic gyro at the current moment; (2) : Perform single-branch reconstruction on the decomposed approximate signal and detail signal to obtain the reconstructed approximate signal and detail signal; (3): Use the FLP method for the approximate signal and detail signal in step (2) layer by layer. Denoising; (4): Use the layer-by-layer denoised signal obtained in step (3) as the input of the neural network, and use the fiber optic gyro signal at the next moment as the output to train the network, and the training is completed. Denoising and modeling of fiber optic gyroscope zero-drift signal. The method of the invention integrates denoising and modeling, effectively improves the modeling and compensation accuracy of the zero-drift signal of the fiber optic gyro, and is easy to implement.

The invention relates to a de-noising method for ultrasonic testing of a high-voltage bushing lead of a transformer, comprising the following steps: performing wavelet decomposition on an ultrasonic echo signal reflected by the high-voltage bushing lead of the transformer, and performing signal analysis from a detail component and an approximate component; introducing an adaptive threshold methodbased on SURE unbiased estimation, adopting a gradient descent method by the adaptive wavelet threshold estimation, taking the gradient function deltalambda as an objective function, corresponding theminimum value of the gradient function deltalambda to the optimal threshold, introducing the Sigmoid function as a threshold function at the same time, and combining the above variables to obtain a new gradient function deltalambda; combining the grey wolf optimization algorithm with the adaptive threshold method, and optimizing the objective function by using the gray wolf algorithm; de-noisingeach detail wavelet component by using the gray wolf optimization adaptive optimal threshold, and performing wavelet inverse transform signal reconstruction on the de-noised detail component and approximate component, thereby obtaining a de-noised ultrasonic echo signal. Compared with the prior art, the invention has the advantages of improving the accuracy of ultrasonic detection and the like.

The invention relates to a multi-modal medical image fusion method based on an adaptive wavelet basis. The method comprises: S1, wavelet transform is carried out on medical images with different modes by using an adaptive wavelet basis filter, the images are decomposed into high-frequency, low-frequency and high-and-low-frequency-combined components respectively; S2, superposition of the high-frequency, low-frequency and high-and-low-frequency-combined components obtained by decomposition of any two different modal medical images is carried out to obtain high-frequency, low-frequency and high-and-low-frequency-combined components with images fused; and S3, discrete wavelet inverse transform is carried out on the high-frequency, low-frequency and high-and-low-frequency-combined components with images fused to obtain a fused image with the original size. Because of application of the wavelet transform technology to the image processing field, the detail sensitivity to low-frequency components by wavelet coefficients can be improved by the adaptive wavelet basis, so that the high-frequency and low-frequency components obtained by decomposition become accurate and thus the fusion effect can be improved.

The invention provides a self-adaption wavelet threshold solving method, and belongs to the field of seismic exploration data processing and other digital signal processing. The method includes the steps of firstly, inputting a noisy signal gather fi(t); secondly, solving a self-adaption weighted stack (please see the symbol in the specification); thirdly, solving a self-adaption variance sigma[i, new] (t) in an iteration mode point by point, judging whether the difference between the sigma[i, new] (t) and sigma[i-1, new] (t) is smaller than xi or not, if yes, executing the fourth step, and if not, replacing the existing variance with a new variance sigma [i, new] (t) and then executing the second step again; fourthly, substituting the self-adaption variance into a threshold solving function to obtain a threshold lambda; fifthly, conducting wavelet domain denoising on W[j, k] through the threshold lambda; sixthly, conducting wavelet reconstruction on the denoised symbol (please see the symbol in the specification) to obtain seismic data generated after wavelet reconstruction; seventhly, outputting the seismic data obtained in the sixth step and generated after wavelet reconstruction. Through a theoretical model and an actual seismic data test, the method is remarkable in denoising effect and is highly purposeful.

The invention provides a smoothing method and device for an intrinsic mode function decomposed on the basis of empirical mode decomposition. The method comprises the following steps that: carrying outband-pass filtering on an IMF (Intrinsic Mode Function) component signal obtained by decomposition on the basis of an EMD (Empirical Mode Decomposition) method to obtain a reference signal; regulating a wavelet packet filtering parameter, and carrying out adaptive wavelet packet filtering on the IMF component signal to obtain the filtered IMF component signal; and calculating the correlation coefficient of the reference signal and the filtered IMF component signal, and taking the filtered IMF component signal obtained by the wavelet packet filtering parameter corresponding to a maximum correlation coefficient as a smoothing processing result for the IMF component signal. By use of the method, noise retained in the IMF component signal can be cleared, and the discontinuous deformation, which is caused by a high-frequency intermittent signal, of the low-frequency IMF component signal is repaired.

The invention discloses a self-adaptive wavelet kernel neural network tracking control method based on a KLMS. The method includes the following steps that a wavelet kernel neural network is initialized; a preset value is compared with an actual observed value output by a controlled object, an error signal is obtained and input into the wavelet kernel neural network, and a cost function is solved; the self-adaptive learning rate of a hidden layer-output layer weight is adjusted, and the hidden layer-output layer weight is updated; the self-adaptive learning rate of an input layer-hidden layer weight is adjusted, and the input layer-hidden layer weight is updated; the constriction factor of a wavelet kernel function is updated; the induction local field and the output of a hidden layer are solved; the induction local field and the output of an output layer are solved, and the output serves as control signals to be transmitted to an actuating mechanism of the controlled object. Through the self-adaptive wavelet kernel neural network tracking control method based on the KLMS, memory internal storage and computation complexity in the iteration process are reduced, and a control system is made more accurate and faster.

The invention discloses a frequency-domain self-adaptation wavelet multi-mode blind equalization method for immune artificial shoal optimization. The frequency-domain self-adaptation wavelet multi-mode blind equalization method for the immune artificial shoal optimization comprises the following steps that an immune artificial shoal is a mixed group of an artificial shoal and an immune system antibody shoal, position vectors of the artificial shoal and antibody vectors of an immune system of a set of immune artificial shoal are initialized in a random mode, the position vectors and the antibody vectors serve as decision variables of an immune artificial shoal method, input signals of an orthogonal wavelet converter serve as input signals of the immune artificial shoal method, a fitness function of the immune artificial shoal is determined by a cost function of the frequency-domain self-adaptation wavelet multi-mode blind equalization method, and initialization weight vectors of the frequency-domain self-adaptation wavelet multi-mode blind equalization method are determined by an immune artificial shoal optimization method. The frequency-domain self-adaptation wavelet multi-mode blind equalization method for the immune artificial shoal optimization is high in rate of convergence, low in steady-state error, low in complexity and good in practicability when the frequency-domain self-adaptation wavelet multi-mode blind equalization method for the immune artificial shoal optimization is used for processing high-order QAM signals.

The invention relates to a fabric defect classification method based on an improved extreme learning machine. The method comprises the steps that a defect image of a training sample is preprocessed; adaptive wavelets are constructed for decomposition to detect fabric defects, and feature extraction is conducted through a multi-feature fusion method to obtain defect features; the defect features are classified, and in the classifying process, an online ELM algorithm is introduced, and online ELM pruning is conducted on hidden nodes through a sensitivity analysis method. By the adoption of the method, the deficiency of the overall processing mode of bulk data through an ELM is overcome, dependency of algorithm performance on the hidden nodes is reduced, and pruning of the hidden nodes is conducted based on sensitivity analysis.

The invention provides a seismic data de-noising method adopting an adaptive wavelet threshold function. The method addresses the problem that the quality of seismic data in a low signal-to-noise ratio region of oil and gas seismic exploration cannot meet the requirements of treatment and interpretation. By establishing a new threshold function overcoming the defects of hard and soft thresholdsand using wavelet transform as a core idea, the de-noising goal of seismic data is gained and a new threshold function is determined. Compared with the current similar technologies, the method has theadvantages that the de-noising result is closer to original useful signals and the quality of the seismic data is greatly improved.

The embodiment of the invention discloses an image de-noising method and image de-noising equipment. According to the technical scheme of the embodiment, the image texture direction is judged by image interpolation, wavelet transform is performed along the judged direction, and the direction adaptive wavelet transform can transform the texture information of an image to a low-frequency area and protect the textures of the image from being lost in threshold value de-noising, so that the texture information of other directions except the vertical direction and the horizontal direction is processed, and compared with traditional wavelet threshold value de-noising, the visual quality and the peak signal-to-noise ratio (PSNR) value of the de-noised image with rich textures are improved obviously.

The invention discloses a short-term power load prediction method based on an EWT and an LSSVM model. The method can solve the technical problems of low calculation efficiency and high calculation cost of an existing method. The method comprises the following steps: S100, decomposing an original power load sequence by using EWT to obtain IMF components under different frequencies; S200, establishing a load prediction model of each IMF component sequence by using an LSSVM; and S300, adding the prediction results of the load prediction models to obtain a total prediction result. According to themethod, through empirical wavelet transform, the problem of modal aliasing existing in EMD can be solved. Meanwhile, fewer components are obtained through decomposition, and the calculation scale isreduced. The method is a new method of establishing adaptive wavelets that improves prediction accuracy by extracting AM-FM components with compact support Fourier spectra, using EWT to decompose different modalities equivalent to segmenting the Fourier spectra and applying some filtering corresponding to each detected support.