47 results about "Daubechies wavelet" patented technology

A wavelet-based band difference-sum ratio method reduces the computation cost of classification and feature extraction (identification) tasks. A Generalized Difference Feature Index (GDFI), computed using wavelets such as Daubechies wavelets, is employed in a method to automatically generate a large sequence of generalized band ratio images. In select embodiments of the present invention, judicious data mining of the large set of GDFI bands produces a small subset of GDFI bands suitable to identify specific Terrain Category/Classification (TERCAT) features. Other wavelets, such as Vaidyanathan, Coiflet, Beylkin, and Symmlet and the like may be employed in select embodiments. The classification and feature extraction (identification) performance of the band ratio method of the present invention is comparable to that obtained with the same or similar data sets using much more sophisticated methods such as discriminants, neural net classification, endmember Gibbs-based partitioning, and genetic algorithms.

The present invention relates to a method of recognizing the human iris using the Daubechies wavelet transform. The dimensions of characteristic vectors are initially reduced by extracting iris features from the inputted iris image signals through the Daubechies wavelet transform. Then, the binary characteristic vectors are generated by applying quantization functions to the extracted characteristic values so that the utility of human iris recognition can be improved as the storage capacity and processing time thereof can be reduced by generating low capacity characteristic vectors. By measuring the similarity between the generated characteristic vectors and the previously registered characteristic vectors, characteristic vectors indicative of the iris patterns can be realized.

A music classification technique computes histograms of Daubechies wavelet coefficients at various frequency subbands with various resolutions. The coefficients are then used as an input to a machine learning technique to identify the genre and emotional content of music.

The invention relates to a transformer on-line fault detecting method base on a sampling integrated SVM (support vector machine) under wavelet GGD (general Gaussian distribution) feathers and unbalanced K-mean values, and belongs to the field of transformer fault detection. The method aims at overcoming the defects caused when the wavelet analysis is applied to the transformer fault detection for carrying out feather extraction in the prior art. The transformer on-line fault detecting method comprises the steps that 1, vibration signals of a transformer are collected; 2, low-pass filtering processing is carried out, high-frequency noise information is removed, and noise reduction vibration signals are obtained; 3, the noise reduction vibration signals are subjected to segment processing according to time series, db20 wavelets in Daubechies wavelet series are subjected to five-layer static wavelet analysis, each layer of wavelet conversion GGD parameters are extracted, five layers of GGD parameters are combined to be used as fault detection feather data, and the fault detection feather data is respectively used as training samples and testing samples; 4, the training samples are utilized for training a SVM detector; and 5, the testing samples are input into the trained SVM detector, and the on-line fault detection of the transformer is realized.

The invention relates to a fabric defect detection method based on a deep learning algorithm, and the method comprises the following steps: carrying out the noise filtering and noise enhancement of a fabric original image through employing the image preprocessing technology; carrying out the decomposition and reconstruction of the fabric image after preprocessing through a Daubechies wavelet transformation method, carrying out the feature extraction of images in horizontal, vertical and diagonal directions after reconstruction, and forming a feature vector of the images; Carrying out the analysis and processing of the extracted feature vector of the fabric image through employing a multi-hidden-layer BP neural network model, and achieving the recognition and classification of defects in the fabric image. The method can automatically recognize the defects, and classifies the recognized defects.

The invention discloses a criterion method for the protection startup of an extra-high voltage direct-current circuit according to a polar wave change rate. The method comprises the following steps of: when the direct-current circuit fails, determining a polar wave according to direct-current voltage and current measured at a protection installation position, wherein the sampling rate is 10 kHz and the time window is 5 ms; performing one-dimensional wavelet decomposition on the polar wave by using a db4 wavelet of a Daubechies wavelet system to obtain modulus maximums, and selecting a first modulus maximum serving as a voltage change rate startup criterion, wherein the electric quantity used by the criterion has high polar wave representation discrimination and is easy to judge; and performing one-dimensional wavelet decomposition on a fault signal by using the db4 wavelet of the Daubechies wavelet system, so that the influence of noise can be eliminated to a great extent. As proved by a large quantity of emulation results, the criterion method has a good effect.

The invention discloses an alarm method for a monitoring system, which is based on Daubechies wavelets to decompose the monitoring information time-series data into a high-frequency signal and a low-frequency signal, wherein a forecast model of the high-frequency signal and a forecast model of the low-frequency signal are respectively constructed through adopting a least squares support vector machine arithmetic, and the wavelet inverse transformation is utilized to obtain the final forecast result as the forecast models determine forecast values. The alarm method adopts a particle swarm optimization algorithm to real-timely and automatically adjust model parameters according to the observed data and the estimation result, so the tracking of the 'slow' time varying physiological parameter time series is realized, and the model accuracy is ensured. According to the modeling result, the alarm method can forwardly forecast, ensures the forecast values and the upper and the lower thresholds and automatically set an alarm threshold, accordingly, different alarm models can be built according to different monitored people, and the alarm threshold is automatically set. The alarm method can be applied in a central monitoring system, an intensive care unit and the coronary heart disease monitoring of a hospital and a community remote monitoring system.

The invention discloses a method and a device for detecting the slag boundary viscosity temperature of a Shell gasification furnace. A Shell coal gasification furnace comprises a pressure shell, wherein a gasification chamber enclosed by the wall of a gasifier is formed in the pressure shell. The method comprises: measuring the temperature of the slag, receiving a vibration signal generated by the wall of the gasifier, performing second-order Daubechies wavelet decomposition, selecting a wavelet frequency band of which the Hurst value is smaller than 0.5 as a characteristic frequency band, and observing a curve of the change relationship between the energy distribution Rs of the characteristic frequency band and the temperature of the slag, wherein the corresponding temperature at which the Rs reduces sharply and begins to approach zero is the slag boundary viscosity temperature. In the invention, an extra vibration source is not needed, a vibration signal is generated by fluid itselfin a moving proves, safety and environmental friendliness are ensured, the measurement is simple and convenient, and the measurement error is small.

The invention relates to a modeling method for a quartz flexible accelerometer starting model, and belongs to the field of navigation and control research of delivery vehicles. The method comprises the following steps of: performing data sampling on a quartz flexible accelerometer starting process; de-noising the sampled quartz flexible accelerometer starting process data by using Daubechies wavelet to acquire time sequence data; extracting an exponential tread term and a linear trend term of the time sequence data; standardizing the time sequence data after the trend term is extracted; determining category of the time sequence model according to characteristics of an autocorrelation function and a partial autocorrelation function of the time sequence data and modeling by using a proper model; and estimating parameters of the time sequence data model and establishing the time sequence analysis-based quartz flexible accelerometer starting model. The invention provides the modeling method for the time sequence analysis-based quartz flexible accelerometer starting model. Research personnel are helped to more deeply analyze the accelerometer starting process.

The invention relates to a fault detection method based on Daubechies wavelet transform and an elastic network. The method includes the steps that training data and test data are obtained, and the test data are standardized; Daubechies wavelet transform is carried out on the training data, each set of the data serves as a pivot element column vector, elastic network regression is carried out on the pivot element column vectors and the training data, and different minimum evaluation values beta en are solved; by means of a probability density evaluation method, the optimal evaluation value beta en is solved to serve as a threshold value; Daubechies wavelet transform and elastic network regression are sequentially carried on the test data, and the solved evaluation value beta en of each set of the data is compared with the threshold value, and whether faults exist in the data is judged. Compared with the prior art, the method has the advantages that all feature values are taken into consideration, detection accuracy is improved, and applicability is good.

The invention provides an electrocardiosignal baseline leveling method based on wavelet decomposition and spline interpolation. The method comprises the steps that firstly, the position of a QRS wavegroup starting point is determined by utilizing a wavelet analysis method, then the QRS wave group starting point serves as an interpolation node, a baseline form is fitted by using a cubic spline interpolation function, and finally, the baseline is subtracted from original signals, so that drift components in the electrocardiosignals are removed, and the baseline is leveled. The electrocardiosignal baseline leveling method comprises the following steps that firstly, a second-order Daubechies wavelet is selected as a base function, and the original electrocardiosignals are decomposed into a six-stage wavelet, and a sixth-stage high-frequency component reconstruction signal is used for detecting a QRS wave group; secondly, the position of the QRS wave group starting point is determined; thirdly, the position of the QRS wave group starting point serves as an interpolation node, and the baseline form is fitted by using the cubic spline interpolation function; fourthly, the baseline component is subtracted from the original electrocardiosignals to obtain signals without the drift components.

The invention discloses a method for detecting the rotate speed of the flooding point of the mixing kettle, comprising: 1) mounting a hydrophone probe head at the outer wall surface of the mixing kettle or in the mixing kettle;2) receiving the vibration signal from the inner of the mixing kettle;3) analyzing the received vibration signal by the second order Daubechies wavelet analyzing technology and selecting the vibration signal in the frequency band with the Hurst value being less than 0. 5;4)obtaining the corresponding mixing rotate speed when the energy fraction R[i] of the feature signal is increased rapidly and start to be stabilized, namely the rotate speed of the flooding point. The vibration receiving device adopting the method can be in the insertion type or in the non-insertion type, has simple and easy installation and no influence for the movement and response of the fluid; the vibration signal is generated in the movement procedure of the fruit without the emission source, thus being safe and environment-friendly; the device has lower request for the condition of measurement and can adapt to the poor condition, and has sensitive response and small measurement error and wide application range.

The invention provides a novel method for extracting fault characteristics of a rolling bearing based on a Daubechies wavelet energy base. The method comprises a step of performing Daubechies waveletdecomposition reconstruction on a rolling bearing vibration signal, a step of determining a reconstructed wavelet layer number i according to a set error value, a step of extracting first ith layer ofDaubechies wavelet with maximum specific weight and carrying out orthogonal normalization, a step of calculating the first ith layer of Daubechies wavelet and establishing a fault mode classificationspace, a step of calculating the projection coordinates of a time domain signal in the fault mode classification space under different working conditions and calibrating fault characteristics, a stepof carrying out space division on different working condition signal characteristics by using a support vector machine and dividing a fault characteristic area in the fault mode classification space,and a step of carrying out Daubechies wavelet decomposition, reconstruction and orthogonal normalization on a newly obtained working condition signal, calculating a power spectrum, calculating the fault mode classification space coordinates, and judging the fault characteristic area. According to the method, the single-point fault characteristic signal of the rolling bearing can be effectively extracted, and a diagnosis result has high precision.

The invention discloses an electroencephalogram abnormal signal detection device and method. The device comprises an electroencephalogram signal preprocessing unit which is used for obtaining an original electroencephalogram signal and carrying out the denoising of the original electroencephalogram signal, and obtaining a target electroencephalogram signal, a wavelet decomposition and reconstruction unit which is used for acquiring the target electroencephalogram signal, and performing X-layer decomposition by adopting Daubechies wavelets according to the coverage frequency of the abnormal waveform and the sampling frequency of the electroencephalogram detection equipment to obtain X-layer frequency bands and characteristic components of each frequency band, a nonlinear kinetic parameter estimation unit which is used for calculating sample entropy characteristics of the electroencephalogram signals of each frequency band after wavelet decomposition, a normalization unit which is used for carrying out normalization processing on the feature components and the sample entropy features to obtain feature vectors, and a detection and classification unit which is used for detecting and classifying the feature vectors. According to the method, features after wavelet transform and features of nonlinear dynamics are combined, comprehensive consideration is carried out, and classificationdetection is carried out on final waveforms.

The invention discloses an operating parameter detection method of stirring tank, which comprises (1), arranging an acoustic emission signal receiver on the wall of a stirring tank, (2), receiving the acoustic emission signal from the stirring tank, (3), using second-order Daubechies wavelet decomposition technique to analyze the received acoustic emission signal, using Hurst analysis to select the acoustic emission signal which Hurst value is lower than 0.5, (4), when the energy consumption ratio Ri is quickly reduced and gradually stabilized, setting the relative stirring speed as critical stirring speed. The inventive vibration receiver is non-inserting type, with simple and easy mounting, without affecting the motion or internal reaction of multi-phase fluid and without demanding emission resource. The vibration signal is generated in fluid motion, which is safe and environment friend with low demand on measurement conditions and the application in bad weather, therefore, the invention can normally work under high-temperature high-pressure condition, with sensitive reaction, low measurement error and wide application.

The present application provides a prediction method and a prediction device for food safety risk level and an electronic apparatus. The method includes: classifying food safety risk level based on historical test data for food safety, to obtain historical data for food safety risk level; performing wavelet decomposition on the historical data for food safety risk level based on Daubechies wavelet basis, to obtain a plurality of historical data components for food safety risk level; and inputting the plurality of historical data components for food safety risk level into an LSTM model and predicting a food safety risk level, to obtain a predicted value of the food safety risk level. By the prediction method and the prediction device for food safety risk level and the electronic apparatus according to the present application, the food safety risk level may be effectively predicted.

The invention discloses a method for correcting a test deviation of a relative reduction rate of pellet. The method is characterized by installing Matlab software on a computer, correcting the final reduction rate of a reduction signal at the end of the reduction test by using Daubechies wavelet in the Matlab software, separating noise of the reduction signal from the original signal, eliminating the noise in the original signal of the reduction signal in the presence of external interference, decomposing the mixed signal to different frequencies of block signals, reducing the sawtooth reduction signal to give a smooth curve, and obtaining the reduction rate the same with that of the original noise-free reduction signal curve. The method has the advantages that the noise of the pellet reduction curve is reduce conveniently by using the Daubechies wavelet combined with a GUI function of a Matlab wavelet analytical tool box, to achieve result correction. By correcting the reduction curve via the method, the detection time can be shortened by at least 16h (i.e. half time of retest), and the material cost can be saved by at least 1800 rmb in each test at the same time.

The invention discloses a method for detecting a commutation failure of direct current transmission. According to the method, characteristics of direct current waveforms before and after the commutation failure are collected, Daubechies wavelets with second-order vanishing moments are adopted for analysis, a normal state and slight faults of an alternating current system are excluded, and the purpose of quickly detecting the time point of the commutation failure is achieved. The method has the advantages that in the process of detecting the commutation failure of a direct current transmissionsystem, the calculation process is simple, the calculation speed is high and the detection precision is high; the method can be applied online, the accurate detection of the commutation failure is facilitated, commutation failure information can be quickly obtained, the situation that a power transmission line and converter station equipment fail is avoided, and the running safety of a power gridis improved.

The invention discloses a method and an apparatus used for detecting critical viscosity temperature of Shell gasifier slag. The Shell coal gasifier comprises a pressure shell; a gasification chamber is formed by a gasifier wall, and is arranged in the pressure shell. The method comprises following step: the temperature of slag is determined; vibration signals sent by the gasifier wall are received and are subjected to two-order Daubechies wavelet decomposition, a wavelet frequency segment with a Hurst value less than 0.5 is selected as the characteristic frequency segment; and a relation curve representing the relationship of the change of energy resolution Rs of the characteristic frequency segment and the temperature of slag is discussed. When Rs decreases rapidly and begins to approach zero, the corresponding temperature in the relation curve is the critical viscosity temperature of Shell gasifier slag. According to the method of the invention, it does not need to prepare another vibration source, the vibration signals are generated by fluid in transporting processes, the method is safe and is friendly to the environment, detection is simple and convenient, and detection error is small.

The invention discloses a sensor fault signal feature extraction method based on wavelet analysis, and the method comprises the steps: firstly obtaining sensor fault signal data, and carrying out the dimensionless preprocessing; carrying out N-layer wavelet decomposition on the dimensionless sensor fault signal by adopting Daubechies wavelet, and storing a wavelet decomposition coefficient; respectively carrying out modulus maximum feature extraction processing and high-frequency relative wavelet energy feature extraction processing on the wavelet decomposition coefficient to obtain a modulus maximum feature vector and a high-frequency relative wavelet energy feature vector, and finally forming a sensor fault signal feature matrix; according to the sensor fault signal feature extraction method provided by the invention, wavelet transform is only carried out on some discrete points on wavelet time and scale planes by adopting multi-scale one-dimensional wavelet decomposition, so that information redundancy is reduced, the feature extraction speed is improved, time domain and frequency domain features of fault data can be obtained at the same time, and the feature stability is enhanced; and the validity of the characteristic parameters and the reliability of subsequent fault identification are improved.