124 results about "Wavelet filtering" patented technology

Changes occurring downhole during a fracturing process can create or reflect pressure signals. Capturing and evaluating such pressure waves during fracturing enables personnel to monitor, in real time or later, what happens downhole. When a fracture extends, a burst of acoustic noise is embodied in a pressure wave or signal, as is noise coming from other sources. By transforming time-based pressure signals to a frequency base, one can monitor this acoustic noise. In a particular implementation, a waterfall plot of frequency spectra at successive time slices of the original signal is used to determine frequency ridges, such as a ridge of decreasing frequencies indicates fracture extension and a ridge of increasing frequencies indicates either closure or proppant backing up in the fracture. Filtering, such as wavelet filtering, can be used. A fracturing process can be controlled in response to determining whether the fracture is extending.

The invention relates to an equipment health state evaluation and recession prediction method based on multi-channel sensing signals. The operation steps of the method are as follows: the first step: joint noise filtration of the multiple sensing signals: adopting a multi-scale wavelet filtering algorithm to perform the joint noise filtration on the multi-channel sensing signal collected from equipment so as to improve the quality of the multiple sensing signals; the second step: generation of an original characteristic set; the third step: extraction of local characteristics of the signals; the fourth step: modeling of a reference self-organizing mapping model; and the fifth step: health quantitative evaluation and residual life prediction. By adopting the method, the quantitative evaluation and the life prediction of performance recession of the equipment can be realized, the operation reliability of the equipment can be further improved and the maintenance cost can be reduced.

The invention relates to the technical field of computer application, specifically, relates to a heart sign obtaining method and a heart sign obtaining device. The heart sign obtaining method comprises the steps of: obtaining photoplethysmography (PPG) data; performing low-pass filtering processing with cut-off frequency less than or equal to 4 HZ on the PPG value and removing high-frequency noise; performing discrete wavelet filtering processing on the PPG value subjected to high-frequency noise removal and correcting baseline drift; obtaining the wave crest and wave trough positions of PPG waveform according to the PPG data subjected to baseline drift correction, and obtaining the time interval value between the wave crests according to the wave crest and wave trough positions and the sampling rate of the PPG data; obtaining RR interval data according to the time interval value between the wave crests, and calculating the heart sign according to the RR interval data. By adoption of the method and the device, a user can comprehensively obtain the heart sign value.

The invention discloses a sliding window auto-regression residual error model-based non-intrusive load switching action monitoring method. The method comprises the following steps of: obtaining current of a bus through NILM equipment such as an intelligent electric meter, and carrying out wavelet filtering and denoising on the current; and obtaining a current period maximum value sequence, obtaining a next current maximum residual error by using an AR regression model, and judging a load state according to a size relationship between the residual error and a threshold value. According to the method, double threshold values U and L are used for dividing corresponding events of load switching; the load switching action monitoring range is enlarged, and load switching actions with relativelylow power can be detected, so that missing detection, caused by power change, of load switching actions with small power is avoided; other electricity parameters of non-intrusive measurement equipmentdo not need to be obtained, and only current parameters of the bus are required, so that the requirement for non-intrusive performance is simplified; and through the category division of loads, the complexity of a whole NILM recognition algorithm can be simplified.

The invention relates to an optic fiber gyroscope random drift modeling method based on a locally variable integrated neural network. The method comprises the following steps: (1) establishing an integrated neural network model; (2) obtaining a learning sample; (3) training the integrated neural network; and (4) modeling to the random drift of an optic fiber gyroscope. The method firstly uses theFLP / wavelet filtering method to denoise the original output signals of the optic fiber gyroscope and form a data pretreatment system with good denoising effect and wide application range; secondly the locally variable integrated neural network is used in modeling, the integrated neural network overfits the training samples, the defects of low precision and poor stability of the sub-neural networkare overcome; the locally variable weight method based on temperature is introduced to ensure that the network can adjust the integrated weight matrix according to the change of temperature; and the optimum network modeling precision can be ensured at different temperatures, the generalization ability of network is increased, and the dynamic modeling of optic fiber gyroscope random drift is realized.

The invention relates to a blade imbalance fault diagnosis method based on a stator current of a double-fed wind generating set. The blade imbalance fault diagnosis method comprises the following steps of: 1, controlling the double-fed wind generating set to operate in a detection mode, and keeping the wind generating set to operate at a constant rotating speed; 2, synchronously acquiring a set stator phase current iA and a generator rotating speed r in the constant-speed operating stage, which meet requirements; 3, according to the acquired generator rotating speed r, calculating a 1-multiple frequency of an impeller speed, carrying out signal processing by an envelope spectrum to obtain an envelope spectrum if signal after carrying out wavelet filtering processing on the stator phase current iA, extracting a 1-multiple frequency amplitude M (1) and a k-multiple frequency amplitude M(k) of the impeller speed in the envelope spectrum if signal and using the 1-multiple frequency amplitude M (1) and the k-multiple frequency amplitude M(k) of the impeller speed as features, wherein k represents a blade number; 4, judging sizes of the 1-multiple frequency amplitude M (1) and the k-multiple frequency amplitude M(k), and if M(1) is greater than a*M(k), determining that blade imbalance exists, wherein a represents a preset amplitude ratio. The blade imbalance fault diagnosis method is simple and easy to operate, excellent in accuracy, and good in applicability.

The invention discloses a device capable of improving measurement precision of gas analysis instrument, which comprises a filter wheel which is provided with a plurality of infrared filters; an electric motor which is used to drive the an infrared stationary light source and a monolithic system, wherein, a plurality of infrared filters are arranged on the filter wheel, the electric motor is used to drive the filter wheel to rotate against a preset frequency; an infrared stable light source which is used to irradiate a detecting gas cell through the infrared filters and is provided with an infrared sensor behind the detecting gas cell to receive the beam from the infrared stable light source; a single chip system which is used to control the drive of the electric motor and conduct wavelet contraction thresholds filtering on the signals received by the infrared sensor. The method and device of the invention are optimized by decomposing Haar wavelet and rebuilding wavelet, multiplication is transferred to addition in wavelet transform and meanwhile section wavelet filtering is adopted, which reduces operation volume, can be applied on a single chip system or imbedded system, effectively reducing the noise signals of gas analysis instrument, improving the measurement precision and decreasing the dependence on the hard circuit filter and model selection.

The invention discloses a laser radar weak signal extraction and decomposition method based on fractional order Fourier transformation, and belongs to the laser radar signal processing technical field; the method comprises the following steps: 1, carrying out fractional order Fourier transformation for noise-containing laser radar echo signals, obtaining a FRFT spectrum, and converting one dimension signals into a 2D image; 2, using a mean filtering, wavelet filtering and Butterworth filtering compound filtering method to make image enhancement for the FRFT spectrum; 3, clustering enhanced FRFT spectrum image data, and determining a clustering quantity and separating each piece of cluster data; 4, recording position and intensity of separated each piece of cluster data, screening valid data points, carrying out least square fitting for reserved data so as to obtain a flight time, a pulse width and an amplitude, thus finishing laser radar echo signal extraction and component decomposition. The 2D time-frequency analysis method for the laser radar signal processing has a stronger anti-noise ability and a more accurate decomposition result.

A method according to an aspect of the invention is for filtering coherent multipulse radar, sonar or lidar returns. The method comprises the steps of separating the returns into mutually orthogonal I & Q components, performing wavelet processing on the I and Q components to generate an I & Q wavelet filter output. The I & Q wavelet filter outputs are data compressed to remove clutter portions. Data compression may be removal of large-amplitude or earliest wavelet responses. An inverse wavelet filtering function is performed on the compressed wavelet filter outputs, to thereby generate edited coherent I & Q components. Doppler filtering, is performed on the edited coherent I and Q components to thereby generate frequency-domain target data in which the minimum discernible target velocity is improved.

The invention discloses a method for detecting component size based on matching pursuit (MP) wavelet filtering. The method comprises the steps of (1), calibrating a charge coupled device (CCD) camera; (2), acquiring a component image and performing MP wavelet filtering on an image; (3), performing differential operator edge detection on the image; (4) performing sub-pixel edge detection on an edge pixel; and (5) calculating the size of the component according to a sub-pixel coordinate to generate computer-aided design (CAD) graph. According to the method, the image of the component is subjected to the MP wavelet filtering, the filtered image is subjected to the sub-pixel edge detection, and the size of the component is calculated and extracted according to the sub-pixel coordinate, so that the detection accuracy can be correspondingly improved, and the detection speed is quickened. The invention discloses a detecting system based on the detecting method. The system comprises an image acquisition unit, an image processing unit and an image generating unit. Because the camera is combined with a personal computer (PC), image acquisition and image processing on components on the production line can be realized, so that the timeliness for detecting the components is guaranteed.

The embodiment of the invention discloses a sleep period dividing method and device. The method comprises the steps of acquiring a state signal, wherein the state signal at least comprises a brain electrical signal and an eye electrical signal; conducting time-domain amplitude filtering on the state signal to remove time-domain interference and conducting frequency-domain wavelet filtering on thestate signal to remove frequency-domain interference; utilizing Hjorth parameters to determine time-domain characteristics of the brain electrical signal; conducting time-frequency converting on the brain electrical signal to obtain time-frequency-domain characteristics of the sleep signal; determining the complexity of the brain electrical signal; extracting frequency change information of the eye electrical signal; combining the time-frequency characteristics, the time-frequency-domain characteristics, the complexity and the frequency change information and determining a currently sleep period; extracting characteristics of the state signal from many dimensions. Thus the judgment accuracy is improved, and meanwhile, compared with the prior art, the sleep period dividing method and devicealso have the advantage of being low in calculation complexity.

The invention relates to a filtering selection platform based on image content analysis. The platform comprises a median filter device, a self-adaptive recursive filter device and a wavelet filter device. The median filter device, the self-adaptive recursive filter device and the wavelet filter device are used for carry out median filtering, self-adaptive recursive filtering and wavelet filtering on an infrared image to be processed respectively. According to the invention, the customization of an image filtering type is changed, and the efficiency and the accuracy of image filtering are improved.

The invention provides a method for dynamic direction measurement error compensation while drilling based on microquartz angular rate sensor, belonging to probing technical field. The method comprises performing rapid wavelet filtering pretreatment to the original signal from the measurement system while drilling based on the characteristics of the output signal from the microquartz angular rate sensor and the noise characteristics of the scaling factor parameter, modeling in accordance with grey theory, applying operation method of accumulated generation to obtain regular nonlinear function corresponding to single value, building model of single valued mapping relation of the angular rate sensor electrical null in three-dimensional space, building drifting model of the angular rate sensor with time and temperature as input parameters, and approximating the single valued corresponding nonlinear function generated by accumulated generation method to improve the accuracy of dynamic measurement. The invention is suitable for directional well logging while drilling and micropore well logging.

The invention discloses a method for identifying sino atrial node electrogram based on integration of wavelet transform and a support vector machine, comprising the following steps of: firstly carrying out wavelet filtering processing on a sino atrial node electrogram signal based on a bivariate shrinkage function; then carrying out four-layer double density wavelet transform on the interference eliminated signal, carrying out V wave peak value positioning on a scaling function x(n) at the fourth layer by means of an adaptive slope method, based on the above processing, positioning an A wave peak value, calculating time width characteristic TA of the A wave, and positioning a starting point and an ending point of a P forward wave to obtain sino atrial node conduction time TSACT (Total Sino Atrial Conduction time); and respectively calculating cross entropies HA and HP of two double density wavelet coefficients of the A wave and the P forward wave, carrying out tandem fusion on the above characteristic parameters of TA, TSACT, HA and HP to obtain a fusion characteristic vector FSNE=(TATSACTHAHP), and finally sorting the characteristic vector using a wavelet support vector machine so as to realize the identification of the sino atrial node function.

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.

The invention discloses a high-frequency ocean radar target detection method based on a compact antenna array. A receiving antenna of a high-frequency ocean radar is an compact antenna array which consists of two identical monopole cross looped antennas. In the same time period, two groups of coherent radar echoes can be received. The radar echoes respectively comprise a target, sea clutter, zero frequency clutter and noise. The method can achieve the high-efficiency detection and extraction of a target through three steps: sequentially carrying out the PCA and extraction, wavelet filtering and adaptive threshold detection of the two groups of radar echoes. The method is advantageous in that the method can greatly improve the SNR (signal to noise ratio) and SCR (signal to clutter ratio) of the target; the method can achieve the better inhibition of the continuous clutter, such as the sea clutter and zero frequency clutter; and the method has better detection performances in a multi-target occasion.

The invention discloses a cold belt rolling mill panel thickness limited frequency domain feedforward control compensation method based on wave function, including the following steps: step one, a measuring device is utilized to measure process information online; step two, calculating is carried out by Psi(p.deltaL, t), wherein p is equal to -N, -N+1,..., N; step three, wavelet decomposition level is determined; step four, a wavelet decomposition method is adopted to decompose a Psi(p.deltaL, td) sequence, thus obtaining low frequency components and detailed components of each level; step five, Psi(p.deltaL, td) is made to be equal to GML(p.deltaL, td)-GMH(p.deltaL, td); step six, setting of cleaning between rolls is adjusted by the obtained Psi(0, td) by a feedforward compensator. The invention adopts the wavelet filtering method to reduce overlap of action range of feedforward and feedback frequency domains in the panel thickness control system, realizes orthogonalization of feedforward and feedback control actions to the utmost extent, reasonably reduces the action range of the feedforward frequency domain, is beneficial to optimization of parameters of the feedforward compensation system, improves the matching efficiency of two basic control manners of feedforward and feedback in the panel thickness control system, and finally can effectively improve panel thickness control precision.

The invention discloses an electrocardiosignal ST section automatic judging method and device based on the artificial intelligent technology. For filtered human body electrocardiosignals, on the basisof wavelet filtering and a triangular area method, key character points of the electrocardiosignals are extracted, positioning S waves and T waves are included, the ST section initial point and slopecharacteristics are accurately recognized, then, a base line removal method based on mean value filtering is put forwards, extracted base line sequences are subtracted, novel electrocardiosignals areobtained, ST section and base lines are extracted for the electrocardiosignals, the heart rate and the slope of each section base line are calculated, the standard reference line is selected comprehensively, finally ST section abnormal changes are recognized according to the standard reference line, the qualitative and quantitative ST section judging result is obtained, and whether the base lineis lifted, lowered or normal is determined. The problems that due to base line shifting, base line errors are caused, standard base line selection is not accurate, and the abnormal heartbeat number difference is large are solved, the ST section abnormal change judging accuracy is indirectly improved, calculation is easy, and the method is easy to implement.

The invention relates to a pipe pressure anomaly diagnostic method based on combined filtering and dynamic thresholds. The pipe pressure anomaly diagnostic method includes the steps of obtaining pressure signals of liquid to the head end and the tail end of a pipe in real time, carrying out decomposition reconstruction through low-pass filtering and discrete wavelet filtering, obtaining pressure signals generated after combined filtering, carrying out segment treatment on the pressure signals, calculating the dynamic thresholds of the segmented pressure signals obtained after combined filtering, judging whether the pressure signals of the pipe are abnormal or not in real time through the dynamic thresholds, judging whether the pressure anomaly signals at the head end and the tail end of the pipe are pressure anomaly signals caused by a same source or not with the Pearson related coefficient method, carrying out continuous wavelet transformation on the pressure anomaly signals with the same source to obtain moments corresponding to the maximum value points of wavelet coefficients, and carrying out anomaly point positioning through the difference of the moments corresponding to the maximum value points. By means of the pipe pressure anomaly diagnostic method, the pipe anomaly diagnostic sensitivity is improved, the misjudgment rate is reduced, and pressure anomaly points are accurately detected and positioned in the shortest time.

The invention provides a multi-carrier spread spectrum power line communication system. Communication equipment in the system comprises a signal transmission device and a signal receiving device; the communication equipment is characterized in that the signal transmission device comprises an encoder, a spread spectrum modulating unit and a carrier modulating unit, which are connected in sequence; the signal receiving device comprises a wavelet filter, a carrier frequency separator, a despreading unit, a data demodulator and a data decoder, which are connected in sequence; when in use, the spread spectrum modulating unit in the signal transmitting device carries out spread spectrum modulation on the data encoded by the encoder, and then the carrier modulating unit carries out multi-carrier modulation on the signal subjected to the spread spectrum in order to meet the orthogonality conditions, and transmits the signal subjected to the spread spectrum to the signal receiving device; the wavelet filter in the signal receiving module carries out wavelet filtering on a received signal, and then the carrier frequency separator carries out frequency separation on each filtered subcarrier, the despreading unit despreads, and the data demodulator demodulates the signal, sequentially, and finally the decoder decodes in order to gain the data.

The invention discloses a quality-fluctuation prediction method for multi-procedure processing process for solving the technical problem that an existing quality control method for multi-procedure processing process is poor in accuracy. The technical scheme of the quality-fluctuation prediction method includes steps of designing a quality fluctuation model for multi-procedure manufacturing process by means of Bayesian state space; processing noise item in the quality fluctuation model for the manufacturing process by means of wavelet filtering, and eliminating noise; designing multielement quality control graph to monitor processing process on the basis of state value after noise elimination, and realizing control of multielement processing process by designing statistics quantity of multielement fluctuation quantity. By taking ternary quality features as object, when the process is at stable state, the average running step length ARL0 of the control graph is increased to 199.624 from 186.535 in the background technology, and false alarm rate is reduced to 0.2% from 6.73% in the background technology. When the process is out of control and offset quantity is (0,5, 0, and 0), the average running step length ARL1 of the control graph is reduced to 168.354 from 170.651 in the background technology, and the false alarm rate is increased to 15.823% from 14.675% in the background technology.

The invention discloses a high-speed wavelet multi-carrier spread-spectrum communication system and method suitable for a 5G network. The system comprises a plurality of communication devices which are in mutual communication and are arranged in a wireless network. The communication devices comprise the signal emission device and the signal receiving device. The signal emission device comprises a coding unit, a spread-spectrum unit, a carrier modulation unit and a parallel-serial conversion unit which are sequentially connected. The signal receiving device comprises a series-parallel conversion unit, a wavelet filtering unit, a dispreading unit, a data demodulation unit and a decoding unit which are sequentially connected. High adaptability is achieved to a time-frequency selectivity wireless channel. Meanwhile, due to the facts that spreading codes with different lengths are used for corresponding to various wavelet subcarriers with different bandwidths, and the spreading codes with the same length are further arranged, so that multiple choices are achieved, the high-speed wavelet multi-carrier spread-spectrum communication system can obtain more higher data rates than that of a conventional multi-carrier spread-spectrum communication system in which fast Fourier transformation is used, and thus the high-speed wavelet multi-carrier spread-spectrum communication system and method are suitable for the high-speed 5G network.

The invention discloses a BP neural network-based MEMS gyro random error compensation method. The method comprises the following steps: collecting original data of an MEMS gyro; preprocessing the original data through wavelet filtering; setting the input quantity and the output quantity of a BP neural network; training the data to establish a BP neural network-based MEMS gyro random error model; and compensating an MEMS gyro random error through the BP neural network-based MEMS gyro random error model. The input quantity of the BP neural network is constructed through data difference, the algorithm is simple, and the precision is high, so the method can effectively improve the accuracy and the reliability of MEMS gyro random error compensation, is simple and is easy to realize.

The invention discloses a design method of a multivariate exponentially weighted moving average controller and aims to solve the technical problem that the AVEWMA (multivariate exponentially weighted moving average) controller designed by the existing methods is high in noise. According to the technical scheme, the design method includes: firstly, establishing a product process quality fluctuation model; and secondly, processing quality fluctuation data by real-time wavelet denoising, and designing a WRD-MEWMA controller based on filtering state. Various error source input items are comprehensively considered in the process, the process quality fluctuation model is established by a Bayes state space method, noise information in the measured data is processed by the real-time wavelet filtering method based on the model so that more accurate process measured data are obtained, and the WRD-MEWMA controller is designed based on the data. By comparing the performance of the designed multivariate exponentially weighted moving average controller to that of the prior MEWMA controller, characteristic posture error vector average and variance norm of the controlled are respectively 11.37% and 8.01% lower than those of the controller designed by the prior MEWMA method.

The invention provides a method for discriminating stress concentration of ferromagnetic materials by utilizing detection of metal magnetic memory, which belongs to the field of metal nondestructive detection. The method adopts a stress concentration magnetic detector at least comprising two scan channels, and specifically comprises the following: step 1, preparing for pre-detection; step 2, using the stress concentration magnetic detector to measure normal magnetic field intensity Hy(x) and tangential magnetic field intensity Hx(x) on the surface of a detected ferromagnetic material along a length direction; step 3, performing arithmetic mean on the acquired normal magnetic field intensity Hy(x) and tangential magnetic field intensity Hx(x) of two or more channels respectively; Step 4, performing smoothed data filtering on data after arithmetic mean; step 5, performing wavelet filtering on the data after smoothed data filtering; step 6, calculating normal magnetic field gradient, tangential magnetic field gradient and peak-to-peak value according to the data after wavelet filtering calculation; and step 7, discriminating stress concentration.

The invention discloses an ultrasonic guided wave-based pipeline flaw imaging method, which is used for imaging flaws in a pipeline by using guided wave detection technology. The ultrasonic guided wave-based pipeline flaw imaging method comprises the following steps: under optimal excitation frequency, acquiring detection signals of all receiving sensors, denoising a waveform by using a wavelet filtering technology, and analyzing and calculating detection data by using a Fourier transformation and variable transformation method, so as to obtain data of sound fields at all positions of the pipeline. The invention provides a new research method for detecting pipeline flaws through guided waves with high efficiency; the imaging method does not need a focusing system, so that the influence of circumferential guided waves and other modal guided waves can be eliminated, and the long guided wave wavelength and low focused imaging efficiency can be overcome. The specific information of the flaws can be directly obtained from imaging results, no recalculation is needed. The ultrasonic guided wave-based pipeline flaw imaging method is strong in practicability, high in detection sensitivity, and precise for flaw positioning, and has wide application prospects in the field of nondestructive testing of pipelines.

The invention discloses a redundancy design method of fault diagnosis of a flight control system. According to the method, equivalence is carried out on a sensor fault model, a system model is simplified, an extended Kalman filter is designed by taking the simplified model as an object, the systems state and sensor fault parameters are estimated in a combined way, a noise variance in a filtering algorithm is increased properly to improve the stability of the filter and the estimation precision and tracking speed for the sensor fault parameters, wavelet filtering is carried out on an estimationresult to obtain a relatively smooth estimated value, and whether a fault occurs in the sensor can be detected rapidly according to an output value. The system stability is ensured, a fault can be determined rapidly and accurately, and the possibility of fault misjudgment can be reduced.

The invention discloses a data processing method for reducing uncertainty of measurement data. According to the method, various influence factors which are finally and comprehensively reflected on the measurement data are fully considered, thus the analysis and compensation of a statistical law are carried out on the measurement data, a system error is compensated by use of a method of averaging multiple measurements, a random error remained after the compensation of the system error is filtered by use of a wavelet filtering method, the combined compensation data are approximate to real data, thus the uncertainty of the measurement data is reduced, the processing process is compact and efficient; the method takes the measurement data processing of a three-coordinate measuring machine as an example, but is not limited to three-coordinate data processing; the method is a universal data processing method and has high engineering application value.