48 results about "Wavelet modulus maxima" patented technology

The invention discloses a fault positioning method for a transmission line and belongs to the technical field of power systems. The fault positioning method comprises the following steps: arranging a Rogowski coil-based broadband core-through traveling wave detection device on the high-voltage transmission line; acquiring a current traveling wave signal of a three-phase line; performing singular value detection by applying the phase mode conversion and wavelet mode maximum principle; accurately identifying the traveling wave line mode wave head, and accurately determining the wave head arrival time by using a GPS high-precision clock; and measuring the wave speed of real-time traveling wave mode on line according to different wave head time difference information to realize accurate fault positioning. The fault positioning method has the advantages of effectively avoiding two main factors influencing the traveling wave fault positioning precision, along with high positioning precision and accurate and reliable fault positioning result.

The invention relates to an electromyographic signal gait recognition method for optimizing a support vector machine based on a genetic algorithm. According to the electromyographic signal gait recognition method, the penalty parameter and the kernel function parameter of the support vector machine are optimized with the genetic algorithm, the performance of the support vector machine is accordingly optimized, and the efficiency and the accuracy of the support vector machine for recognizing lower limb movement gaits based on electromyographic signals are improved. The electromyographic signal gait recognition method includes the steps of firstly, carrying out de-noising processing on the collected lower limb electromyographic signals with a wavelet modulus maximum de-noising method; secondly, extracting the time domain characteristics of the de-noised electromyographic signals to form characteristic samples; thirdly, optimizing parameters of the support vector machine with the genetic algorithm to obtain a set of optimal parameters with the minimum errors, and constructing a classifier through the parameters; finally, inputting a characteristic sample set into the optimized classifier for gait recognition. The electromyographic signal gait recognition method is easy to operate, rapid in calculation and high in recognition rate, and has the application value and the broad prospects in the human body lower limb gait recognition field.

The invention relates to an electromyographic signal gait recognition method based on particle swarm optimization and a support vector machine. A particle swarm optimization algorithm is utilized to optimize a penalty parameter and a kernel function parameter of the support vector machine so that the performance of the support vector machine can be optimized, and effective recognition and classification are achieved. Firstly, wavelet modulus maximum denoising is carried out on collected lower limb electromyographic signals; secondly, time domain feature extraction is conducted on the electromyographic signals after denoising is carried out to obtain feature samples; thirdly, parameter optimization is carried out on the support vector machine by means of the particle swarm optimization algorithm to obtain a set of optimal parameters with minimal errors, and a classifier is constructed; at last, a feature sample set of the electromyographic signals is input to the classifier, and then classification and recognition are conducted on gait states. According to the method, both accuracy and adaptivity of classification are taken into consideration, the computational process is simple and efficient, and the method has broad application prospects in the field of lower limb motion state recognition.

The invention provides a zero-mode and line-mode time difference radiation net fault location method achieved only through voltage without relying on two-terminal synchronization, and belongs to the technical field of electric power system relay protection. Traveling wave fault distance measurement devices are arranged on the two sides of a feeder line, and fault distance measurement is conducted through information on the two sides. After grounding faults happen to the feeder line of a power distribution network, line-mode traveling components propagated between wires and zero-mode traveling wave components propagated between the wires and the ground are generated due to the sudden change of the voltage at the fault point. Due to the fact that the line-mode propagation velocity and the zero-mode propagation velocity are different, the arrival moment of the line-mode initial traveling wave and the arrival moment of the zero-mode initial traveling wave detected by a measurement terminal are different. Moment calibration is conduced through the wavelet modulus maximum under the fifth dimension according to the initial line-mode voltage traveling wave data and the initial zero-mode voltage traveling wave data detected by the measurement device, and the fault distance is calculated according to the ground fault single-terminal traveling wave fault location calculation formula for the modulus transmission time difference. Fault location is conducted by integrating single-terminal modulus propagation time difference distance measurement information on the two sides.

An improved self-adaptive sparse sampling fault classification method belongs to the technical field of fault diagnosis. A traditional sparse classification method is improved. Firstly, a wavelet module maximum value and a kurtosis method are used for carrying out feature enhancement processing on signals, and on the premise that signal sparsity is guaranteed, a unit matrix is adopted to replace aredundant dictionary. Secondly dimension reduction is carried out on data by adopting a Gaussian random measurement matrix, thereby reducing redundant information in the signal, and reserving effective and small amount of data. Then, a sparse coefficient is solved by adopting a sparsity adaptive matching pursuit (SAMP) algorithm, and the compressed signal is reconstructed; and finally, a cross correlation coefficient is adopted as a judgment basis of the category of the fault, so that an improved adaptive sparse sampling fault classification method is provided. Experimental verification proves that redundant information in signals is effectively reduced, the influence of time shift deviation on fault type judgment is avoided, meanwhile, the operation complexity is reduced, and the calculation speed and the reconstruction precision are improved.

The invention discloses a high-resolution synthetic aperture radar (SAR) satellite image speckle de-noising method. Speckle noise statistic characteristics are analyzed, and based on wavelet modulus maximum and Lyapunov exponent theories, useful signals and speckle noise in SAR images are identified by using spatial related processing technology, so when the marginal information is not lost, the speckle noise is greatly attenuated; meanwhile, the 'drift' phenomenon caused by wavelet multi-resolution analysis is corrected, so the de-noising capacity is enhanced and the processing speed is improved; and the method can identify the useful signals and the speckle noise in the SAR images, greatly attenuate the speckle noise when the marginal information is not lost, meanwhile correct the 'drift' phenomenon caused by the wavelet multi-resolution analysis, enhance the de-noising capacity and improve the processing speed.

The invention provides a radiation net fault location method by means of a zero mode and aerial mode time difference independent of double-end synchronization and with matching of the magnitude of voltages and the magnitude of currents, and belongs to the technical field of power system relay protection. The two sides of a feeder are respectively provided with a traveling wave fault location device, and fault location is carried out according to information on the two sides. After a ground fault of the feeder of a power distribution net occurs, due to sudden changes of the voltage at a fault point, an aerial mode traveling wave component spreading between wires and a zero mode traveling wave component spreading between the wires and the ground are generated. Due to the fact that the spreading speeds of an aerial mode and a zero mode are different, the arrival moment of an aerial mode traveling wave detected by the measurement end and the arrival moment of a zero mode traveling wave detected by the measurement end are different. According to original aerial mode voltage and current traveling wave data and original zero mode voltage and current traveling wave data detected by the measurement end, wave arrival moment calibration is carried out by means of a wavelet modulus maximum value under a fifth dimension, and then the fault position is calculated according to a calculation formula of a ground fault single-end traveling wave fault location method of modulus transmission time differences. Fault location is carried out according to measurement information of the single-end modulus transmission time differences at two sides.

The invention discloses a diagnosis method for inter-turn short trouble of a permanent magnetic motor with fault tolerance, which is characterized in that the method comprises the following steps: (a) the method implements the transient power analysis of motor windings, and if the number of the windings that have short trouble is no less than 10, the method decides whether a trouble happens by computing the mean value change of the transient power before and after the trouble happens, if the number of the windings that have short troubles is less than 10, the method turns to next step; (b) the method takes the derivation of the first-order complex Gaussian wavelet function as the wavelet function, analyzes and locates the unsteady signal of the inter-turn short trouble; (c) the method compares the transformation situations of the wavelet modulus maximum value of the transient power signal in a trouble phase and in a normal phase, and if the situations are the same, the method decides that no trouble exists; if the situation are different, the method decides that the inter-turn short trouble exists. By adopting the modulus maximum value of the complex Gaussian wavelet, the invention can markedly observe a distortion point of the transient power before and after an inter-turn short trouble happens, accurately capture trouble characteristics when the inter-turn short trouble happens in the motor, and solve the difficult problems of the online trouble diagnosis of the inter-turn short trouble of small turns of a stator winding in a permanent magnetic motor.

The invention discloses a distributed optical fiber fence intrusion detection and location method based on a phi-OTDR (Optical Time Domain Reflectometer), and the method comprises the following steps of: firstly, subtracting an acquired ith signal by an acquired (i-1)th signal to obtain a preprocessing monitoring signal; secondly, carrying out at least four-layer wavelet decomposition on the preprocessing monitoring signal by adopting a plurality of wavelet functions; thirdly, judging whether actual intrusion exists according to the consistency of distribution positions of maximum-wavelet mode maximums in all scale signals; fourthly, judging that the actual intrusion exists if the distribution positions of the maximum-wavelet mode maximums in at least three scale signals are consistent; fifthly, corresponding multiplying the scale signals in which the intrusion exists and the distribution positions of the maximum-wavelet mode maximums are consistent to obtain a composite signal of which pseudo maximums are inhibited and mode maximums at a catastrophe point are enhanced, using a position corresponding to the maximum-mode maximum in the composite signal as a position of the determined intrusion point, determining the intrusion generation and the intrusion position by combining various composite signal judgment results of wavelet multi-scale analysis according to the majority voting criterion.

The invention discloses a cable partial discharging positioning method based on self-correlation-wavelet modulus maximum analysis, and belongs to the field of the electric system. The method comprises the following steps that firstly, collecting and primary processing is carried out on a partial discharging signal; secondly, a partial discharging incident wave-reflected wave pulse data segment is extracted; thirdly, self-correlation processing is carried out on the incident wave-reflected wave pulse data segment x(n) in the second step, and an estimated value dN of a wavefront time difference of an incident wave and a reflected wave is extracted; fourthly, N-layer wavelet decomposition and reconstitution are carried out on the incident wave-reflected wave pulse data segment x(n) in the third step, a maximum point of each layer wavelet reconstitution signal is extracted as a the wavefront position of the incident wave, and the estimated value dN, obtained by the self-correlation analysis in the third step, of the wavefront time difference of the incident wave and the reflected wave is regarded as the reference, the maximum value moment of the reflected wave wavelet modulus is extracted, the wavefront time difference of the incident wave and the reflected wave is calculated, and the partial discharging position is calculated. The method has the advantages that the positioning accuracy is high, and the anti-interference capacity is high.

The invention discloses a fault location method for wide area traveling wave signal below power grid, with a little of fault traveling wave monitoring points for the power grid, the fault signal is processed on the phase-mode transformation and the small wave mode maximum value process to obtain the initial wave header information; the extension double-end travelling wave fault location is realized in the complex power grid, the fault location method of general power grid is offered on this basis for fast finding the exact double-end ranging combination from the measured data, and raising the ranging precision by amending the result. The monitoring data is processed and checked for realizing extension double-end travelling wave fault location for any fault point of the system, and the method has good fault-tolerant capability.

The invention relates to an ear image recognition method combining wavelet analysis and moment features, which is characterized in that the method comprises following steps: (1) preprocessing collected ear images: using wavelet transform modulus maxima to denoise for the collected ear images, and extracting ear image edge to gain wavelet decomposition modulus maxima edge images in various scales; (2) extracting characteristic value of the ear images: making use of an improved wavelet moment invariants arithmetic to calculate the wavelet moment invariants used for extracting the characteristic value of the ear images; (3) recognizing the ear: processing weighting and classifying the characteristic value of the ear images gained in the step (2), and recognizing the ear. The invention processes to extract the characteristic value of the ear via wavelet modulus maxima and the improved wavelet moment invariants arithmetic, and not only has a high calculation accuracy and a fast calculation speed, but also can improve discrimination of the collected ear images in the environment of poor illumination, illumination change and noise jamming.

The invention relates to a least square and correlation analysis-based line arrester action recognition method, and belongs to the technical field of power system relay protection. The method comprises the following steps of: when a transmission line encounters a lightning stroke, measuring a lightning stroke current waveform by a measurement end; transforming the measured lightning stroke currentwaveform by utilizing a wavelet modulus maximum value so as to determine an arrival moment of a faulted initial traveling wave; fitting an unacted curve of an arrester by utilizing a least square method according to a wave head falling edge of the initial traveling wave; selecting a proper time window, carrying out correlation analysis on the lightning stroke current waveform measured by the measurement end and the fit unacted curve of the arrester to obtain a correlation analysis coefficient, the threshold value of which is 0.9, if the correlation analysis coefficient is greater than 0.9, considering that the line arrester does not act, and if the correlation analysis coefficient is smaller than or equal to 0.9, considering that the line arrester acts. The method is not influenced by thefactors such as lightning stroke type, lightning stroke line positions and lightning current amplitude, and is correct and reliable in judgement result.

The invention discloses a method for detecting the edge of a noisy image. The method comprises the steps of firstly, inputting the noisy image; secondly, improving wavelet modulus maximum edge detection operators, and obtaining improved edge detection operators in the horizontal direction and the vertical direction; thirdly, calculating partial derivatives of the noisy image in the horizontal direction and the vertical direction; fourthly, calculating an amplitude value matrix and an angle matrix of an image gradient; fifthly, determining edge points, and obtaining edge images; sixthly, fusing the edge images and obtaining a fused image; seventhly, determining the optimal edge and obtaining the final edge image; eighthly, outputting the final edge image. According to the method, shear transformation and traditional edge detection operators are combined for image edge detection, a de-noised image is obtained finally, namely noise can be restrained effectively, and the more complete and continuous image edge can be detected.

The invention discloses a ground penetrating radar signal quantitative analysis method and system. The method comprises steps: ground penetrating radar signals are acquired; the ground penetrating radar signals are filtered to determine a detected target in a radar image; the detected target is extracted to obtain multiple single channel signals containing detected target reflected wave information; according to the used ground penetrating radar antenna pulse function type, a corresponding quantitative analysis-used optimal wavelet basis is selected or constructed; the optimal wavelet basis isadopted to process each single channel signal to obtain multiple detail coefficient component time-modulus curves; a wavelet singularity analysis method is adopted to determine multiple detected target reflected waves in each detail coefficient component time-modulus curve; a wavelet modulus maximum method and the multiple detected target reflected waves are used to determine the coordinates of alocal modulus maximum point; and a two-way travel time formula and the coordinates of the local modulus maximum point are used to calculate the distance between different detected target reflected waves. By adopting the method or the system disclosed in the invention, accurate quantitative analysis on the ground penetrating radar signals can be realized.

The invention discloses an elevator running health degree evaluation method based on elevator cabin acceleration signal analysis. The elevator running health degree evaluation method aims at solving the problem that a current elevator lacks real-time and accurate elevator running health evaluation standards. The elevator cabin acceleration signal reflecting the elevator vibration condition is fully utilized, the acceleration of the elevator cabin of the elevator in the actual running process is collected in real time, a vibration signal which is closely related to the elevator running performance is obtained by using the wavelet modulus maximum value noise reduction and de-trend method, then a plurality of characteristic variables are selected as evaluation indexes, and finally, a scoringfunction is determined to be combined with the plurality of evaluation indexes to give quantitative evaluation results of aspects of the elevator running health degree. According to the method, timelyand quantitative evaluation of the comprehensive operation health degree of the elevator is realized, the change condition of the operation performance of the elevator can be effectively mastered, onone hand, the comfort degree of the passenger can be effectively improved, on the other hand, valuable reference basis is provided for maintenance of the elevator, occurrence of some important safetyaccidents can be prevented, and safe and reliable operation of the elevator is guaranteed.

The invention relates to an electrocardiogram data classification method based on 12 leads and a convolution neural network. The method comprises the following steps: acquiring 12-lead electrocardiogram data signals from a PTB diagnostic electrocardiogram database; performing noise reduction processing on the signals acquired in the step 1 by using a wavelet transform denoising algorithm; processing the signals subjected to the noise reduction processing in the step 2 by using a wavelet modulus maximum value and variable threshold method; decomposing periods of a 12-lead electrocardiogram by using R wave peak position information acquired in the step 3, and then extracting a P-QRS-T characteristic segment of each period; selecting appropriate electrocardiosignals and sampling electrocardiosignals according to set sampling points; and constructing a one-dimensional convolution neural network, setting the number of nodes of an input layer, an implicit layer and an output layer of the one-dimensional convolution neural network, training the one-dimensional convolution neural network, and building a 12-lead electrocardiogram classification model. The method can quickly identify electrocardiosignals of a patient suffering from cardiovascular diseases.

The invention discloses a distributed image feature extraction algorithm suitable for collaborative learning. The algorithm is based on Radon transform and singular value decomposition. According to the geometric characteristics of Radon transform, a new invariant texture feature extraction based on Radon transform and wavelet transform is constructed. method, first extract the edge contour of the original image according to the principle of wavelet modulus maxima, construct the central moment of Radon transform for the edge image, obtain translation invariance, and then construct the scale invariant moment according to the statistical characteristics of Radon transform on the basis of the central moment , and finally, find the descending singular value vector of the even moment of the scale invariant moment, the singular value eigenvector has translation, scale and rotation invariance. The invention enhances the anti-noise ability while maintaining translation, rotation and scale invariance; and overcomes the influence of noise on image shape.

The invention discloses a method for judging resistance encountering of an electric lifting table by utilizing a wavelet modulus maximum value. The method comprises the following steps: (1) acquiringcurrent data of the electric lifting table under different working conditions; the method comprises the steps of (1) obtaining an original signal, (2) performing preprocessing of filtering high frequency and averaging on the original signal to obtain a preprocessed signal, (2) applying CWT to the preprocessed signal in the step (1) to obtain a wavelet coefficient, and (3) positioning a hand clamping moment or a hand clamping region by using gray scale images of the wavelet transform coefficient under different scales. Wherein when an obstacle is encountered, singularity can be generated alongwith the occurrence of the modulus maximum of the wavelet coefficient, so that an obvious stripe cone can be formed in the grayscale image of the wavelet transform coefficient, and (4) searching the wavelet modulus maximum in each scale within the influence range. Drawing a wavelet modulus maximum value line on the scale based on the found wavelet modulus maximum value point, wherein all points inthe line are wavelet transform modulus maximum values. By utilizing the method, the load condition can be effectively identified under each load condition.

The invention relates to a novel power signal compressed sensing method based on signal singularity detection improvement. The method is characterized by comprising the following steps of: 1) de-noising power transmission signals based on combination of the maximum value of a wavelet modulus and compressed sensing; 2) designing a compressed sensing measurement matrix in combination with the characteristics of the power transmission signals; and 3) carrying out compressed sensing and reconstruction on the power transmission signals based on the signal singularity detection improvement. According to the method, in a maximum domain of the wavelet modulus, the positive and negative nature of a Lipschitz index is taken as a judgment basis to remove the modulus maximum of noise contained in the power transmission signals, a Chebyshev sparse cycle measurement matrix is designed to carry out compressed sampling on the denoised modulus maximum, and finally, the power transmission signals are reconstructed through a reconstruction algorithm. Thus, the accuracy and efficiency of power line communication information transmission are improved, and the engineering application value of a PLC is enhanced.