[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

CN109583350A - A kind of high-precision denoising method of local ultrasound array signal - Google Patents

A kind of high-precision denoising method of local ultrasound array signal Download PDF

Info

Publication number
CN109583350A
CN109583350A CN201811398953.2A CN201811398953A CN109583350A CN 109583350 A CN109583350 A CN 109583350A CN 201811398953 A CN201811398953 A CN 201811398953A CN 109583350 A CN109583350 A CN 109583350A
Authority
CN
China
Prior art keywords
signal
array signal
array
precision
local
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201811398953.2A
Other languages
Chinese (zh)
Other versions
CN109583350B (en
Inventor
马君鹏
王成亮
亓彦珣
吴晗
刘叙笔
杨贤彪
谢庆
岳贤强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Corp of China SGCC
State Grid Jiangsu Electric Power Co Ltd
North China Electric Power University
Jiangsu Fangtian Power Technology Co Ltd
Original Assignee
State Grid Corp of China SGCC
State Grid Jiangsu Electric Power Co Ltd
North China Electric Power University
Jiangsu Fangtian Power Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by State Grid Corp of China SGCC, State Grid Jiangsu Electric Power Co Ltd, North China Electric Power University, Jiangsu Fangtian Power Technology Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN201811398953.2A priority Critical patent/CN109583350B/en
Publication of CN109583350A publication Critical patent/CN109583350A/en
Application granted granted Critical
Publication of CN109583350B publication Critical patent/CN109583350B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing
    • G06F2218/04—Denoising
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
    • G01R31/1209—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing using acoustic measurements

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Signal Processing (AREA)
  • General Engineering & Computer Science (AREA)
  • Acoustics & Sound (AREA)
  • Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
  • Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
  • Ultra Sonic Daignosis Equipment (AREA)

Abstract

本发明公开了一种局部超声阵列信号的高精度去噪方法，包含以下步骤：运用快速独立分量FastICA将多通道的阵列信号分离为统计的独立分量，对这些独立分量进行自相关分析，初步消除噪声的影响；运用集合经验模态分解算法EEMD对每个通道的信号进行进一步去噪；对精去噪后的信号进行重组，得到高精度的去噪后的信号。本发明解决了传统的EMD分解只能对单个传感器、单通道信号去噪、FastICA只能对阵列信号进行粗略去噪的弊端，避免了因现场噪声信号的干扰而造成的测向定位误差太大，从而提高了超声阵列信号去燥效果和测向定位精度。

The invention discloses a high-precision denoising method for local ultrasonic array signals. The influence of noise; using the ensemble empirical mode decomposition algorithm EEMD to further de-noise the signal of each channel; recombining the fine-denoised signal to obtain a high-precision de-noised signal. The invention solves the drawbacks that the traditional EMD decomposition can only denoise a single sensor, a single channel signal, and the FastICA can only roughly denoise the array signal, and avoids the direction finding and positioning error caused by the interference of the on-site noise signal. , thereby improving the ultrasonic array signal de-drying effect and direction finding positioning accuracy.

Description

A kind of high-precision denoising method of local ultrasound array signal
Technical field
The present invention relates to a kind of denoising method, especially a kind of high-precision denoising method of local ultrasound array signal.
Background technique
Often there is the influence of noise when electrical equipment type local-discharge ultrasonic array detection, it is therefore necessary to calculate the denoising of signal Method is studied.Existing Denoising Algorithm is broadly divided into two classes, and the first kind is the denoising of individual signals, as Wavelet Denoising Method, EMD are gone Make an uproar etc., such algorithm is only preferable to the denoising effect of single channel signal, when denoising with this method to array signal It will affect the phase difference between each array signal, to causing significant impact in the processing of subsequent array signal；Second class is The denoising of multichannel array signal, although such method will not influence the phase difference of array signal, but such algorithm is mainly Based on blind source separating principle, restrictive condition is relatively more, and it is unobvious to denoise effect.
Summary of the invention
Technical problem to be solved by the invention is to provide a kind of high-precision denoising method of local ultrasound array signal, It is good that effect is removed dryness to array signal.
In order to solve the above technical problems, the technical scheme adopted by the invention is that:
A kind of high-precision denoising method of local ultrasound array signal, it is characterised in that comprise the steps of:
The array signal of multichannel: being separated into the isolated component of statistics with quick isolated component FastICA by step 1, Autocorrelation analysis, the preliminary influence for eliminating noise are carried out to these isolated components；
Step 2: the signal in each channel is further denoised with set empirical mode decomposition algorithm EEMD；
Step 3: the signal after essence denoising is recombinated, the signal after obtaining high-precision denoising.
Further, the quick isolated component FastICA is
It sets a trap and puts source signal as S (t)=[S1(t),S2(t),…SN(t)]T, by N number of mutually independent Partial discharge signal group At the collected array signal of sensor array is expressed as follows shown in formula:
X (t)=AS (t)+n (t) (1)
Wherein, t is time variable, and X (t) is that M ties up the received N number of array signal of sensor, i.e. X (t)=[x1 (t), x2 (t) ... xN (t)] T, A is the unknown hybrid matrix of M × N rank, and n (t) is that M ties up observation noise vector；
With the array signal X (t) observed, seek separation matrix W, to obtain the estimation signal Y (t) of source signal.
Further, the step 1 is specially
1.1 pre-processing first to array signal X (t), pretreatment mainly comprising going two steps of mean value and albefaction, is gone Mean value makes observation signal meet zero-mean, and the correlation between each data is eliminated in albefaction, so that each component is as independent as possible；
Using the mean value of following formula removal sample:
Array signal after going mean value is X0(t), to going the data after mean value to carry out whitening processing, it may be assumed that
X ' (t)=TX0(t) (3)
Array signal after whitening processing is X ' (t), and T is linear change matrix；
The initial value of 1.2 setting separation matrix W；
1.3 building quadratic function G1, G2, so that
A1=1.5, a2=2 in formula carry out derivation to G1, G2, and obtaining derived function is g1, g2；
1.4 carry out loop iteration with Newton iteration method, and iteration formula is W ← E (xg1(WTx′))-E(g2(WTX ')) W, Until convergence；
1.5 seek the array signal after separation, it may be assumed that Y (t)=WTX (t).
Further, the set empirical mode decomposition algorithm EEMD is
The array signal obtained after denoising with FastICA, each channel do not have correlation, go to it with single channel It will not influence the phase difference between array signal after algorithm of making an uproar denoising；
Y ' i (t) is obtained after different white noise sequence fi (t) is added in Yi (t), empirical modal point is carried out to Y ' i (t) Solution, obtains each rank intrinsic mode function component (IMF), at this timeWherein Ck (t) is each rank natural mode State function component, rnIt (t) is surplus.
Further, the step 2 is specially
2.1 calculate all Local Extremums of Y ' i (t)；
The envelope up and down that all maximum points and minimum point are constituted is denoted as u by 2.2 respectively0i(t) and v0i(t)；
The mean value of about 2.3 envelopes isThe envelope up and down of single channel Partial discharge signal it is equal The difference of value is H0i(t)=Yi′(t)-m0i(t)；
2.4 judge H0i(t) whether meet extreme point number or zero crossing number is equal or most differences one, and by part The average value of coenvelope line and the lower envelope line being made of local minimum that maximum is constituted is zero, if meeting conditions above, Then H0i(t) it is natural mode of vibration component, otherwise repeats 2.1~2.3 steps, until finding first natural mode of vibration component, is denoted as C1 (t)；
2.5 note r1(t)=Yi′(t)-C1(t) it is new amount to be analyzed, repeats above step and obtain second IMF Amount ... ... and so on obtains n IMF amount, finally remains next monotonic signal rn(t), therefore single channel original signal can be with It indicates are as follows:
Compared with prior art, the present invention having the following advantages that and effect: EEMD is decomposed and is tied with FastICA phase by the present invention The denoising for being applied to type local-discharge ultrasonic array signal is closed, solving traditional EMD decomposition can only be to single sensor, single channel signal The drawbacks of denoising, FastICA can only denoise roughly array signal, avoids and causes because of the interference of live noise signal DF and location error it is too big, so that improving supersonic array signal removes dryness effect and DF and location precision.
Detailed description of the invention
Fig. 1 is that a kind of high-precision of local ultrasound array signal of the invention removes the flow chart of drying method.
Fig. 2 is FastICA disassembler reconfiguration principle figure of the invention.
Fig. 3 is EEMD decomposition process figure of the invention.
Specific embodiment
Below by embodiment, the present invention is described in further detail, following embodiment be explanation of the invention and The invention is not limited to following embodiments.
As shown in Figure 1, a kind of high-precision denoising method of local ultrasound array signal of the invention comprising the steps of:
The array signal of multichannel: being separated into the isolated component of statistics with quick isolated component FastICA by step 1, Autocorrelation analysis, the preliminary influence for eliminating noise are carried out to these isolated components；
Quickly isolated component FastICA is
It sets a trap and puts source signal as S (t)=[S1(t),S2(t),…SN(t)]T, by N number of mutually independent Partial discharge signal group At the collected array signal of sensor array is expressed as follows shown in formula:
X (t)=AS (t)+n (t) (1)
Wherein, t is time variable, and X (t) is that M ties up the received N number of array signal of sensor, i.e. X (t)=[x1 (t), x2 (t) ... xN (t)] T, A is the unknown hybrid matrix of M × N rank, and n (t) is that M ties up observation noise vector；
With the array signal X (t) observed, seek separation matrix W, to obtain the estimation signal Y (t) of source signal.
It is broadly divided into following steps:
1.1 first pre-process array signal X (t), and pretreatment mainly comprising going two steps of mean value and albefaction, is gone Mean value makes observation signal meet zero-mean, and the correlation between each data is eliminated in albefaction, so that each component is as independent as possible；
Using the mean value of following formula removal sample:
Array signal after going mean value is X0(t), to going the data after mean value to carry out whitening processing, it may be assumed that
X ' (t)=TX0(t) (3)
Array signal after whitening processing is X ' (t), and T is linear change matrix；
The initial value of 1.2 setting separation matrix W；
1.3 building quadratic function G1, G2, so that
A1=1.5, a2=2 in formula carry out derivation to G1, G2, and obtaining derived function is g1, g2；
1.4 carry out loop iteration with Newton iteration method, and iteration formula is W ← E (xg1(WTx′))-E(g2(WTX ')) W, Until convergence；
1.5 seek the array signal after separation, it may be assumed that Y (t)=WTX (t).
Step 2: the signal in each channel is further denoised with set empirical mode decomposition algorithm EEMD；
Gathering empirical mode decomposition algorithm EEMD is
The array signal obtained after denoising with FastICA, each channel do not have correlation, go to it with single channel It will not influence the phase difference between array signal after algorithm of making an uproar denoising；
Y ' i (t) is obtained after different white noise sequence fi (t) is added in Yi (t), empirical modal point is carried out to Y ' i (t) Solution, obtains each rank intrinsic mode function component (IMF), at this timeWherein Ck (t) is each rank natural mode State function component, rnIt (t) is surplus.
Specific step is as follows:
2.1 calculate all Local Extremums of Y ' i (t)；
The envelope up and down that all maximum points and minimum point are constituted is denoted as u by 2.2 respectively0i(t) and v0i(t)；
The mean value of about 2.3 envelopes isThe envelope up and down of single channel Partial discharge signal it is equal The difference of value is H0i(t)=Yi′(t)-m0i(t)；
2.4 judge H0i(t) whether meet extreme point number or zero crossing number is equal or most differences one, and by part The average value of coenvelope line and the lower envelope line being made of local minimum that maximum is constituted is zero, if meeting conditions above, Then H0i(t) it is natural mode of vibration component, otherwise repeats 2.1~2.3 steps, until finding first natural mode of vibration component, is denoted as C1 (t)；
2.5 note r1(t)=Yi′(t)-C1(t) it is new amount to be analyzed, repeats above step and obtain second IMF Amount ... ... and so on obtains n IMF amount, finally remains next monotonic signal rn(t), therefore single channel original signal can be with It indicates are as follows:
Step 3: the signal after essence denoising is recombinated, the signal after obtaining high-precision denoising.
EEMD is decomposed the denoising combined with FastICA applied to type local-discharge ultrasonic array signal by the present invention, solves biography The EMD of system, which is decomposed, single sensor, single channel signal denoising, FastICA can only be denoised roughly to array signal The drawbacks of, it is too big to avoid DF and location error caused by due to the interference of live noise signal, to improve supersonic array Signal removes dryness effect and DF and location precision.
Above content is only illustrations made for the present invention described in this specification.Technology belonging to the present invention The technical staff in field can do various modifications or supplement or is substituted in a similar manner to described specific embodiment, only It should belong to guarantor of the invention without departing from the content or beyond the scope defined by this claim of description of the invention Protect range.

Claims (5)

1.一种局部超声阵列信号的高精度去噪方法，其特征在于包含以下步骤：1. a high-precision denoising method of a local ultrasonic array signal is characterized in that comprising the following steps: 步骤一：运用快速独立分量FastICA将多通道的阵列信号分离为统计的独立分量，对这些独立分量进行自相关分析，初步消除噪声的影响；Step 1: Use the fast independent component FastICA to separate the multi-channel array signal into statistical independent components, perform autocorrelation analysis on these independent components, and initially eliminate the influence of noise; 步骤二：运用集合经验模态分解算法EEMD对每个通道的信号进行进一步去噪；Step 2: Use the ensemble empirical mode decomposition algorithm EEMD to further denoise the signal of each channel; 步骤三：对精去噪后的信号进行重组，得到高精度的去噪后的信号。Step 3: Recombining the finely denoised signal to obtain a high-precision denoised signal. 2.按照权利要求1所述的一种局部超声阵列信号的高精度去噪方法，其特征在于：所述快速独立分量FastICA为2. according to the high-precision denoising method of a kind of local ultrasonic array signal according to claim 1, it is characterized in that: described fast independent component FastICA is 设局放源信号为S(t)＝[S1(t),S2(t),…SN(t)]T，其由N个相互独立的局放信号组成，阵列传感器采集到的阵列信号表示如下式所示：Suppose the PD source signal is S(t)=[S 1 (t), S 2 (t),…S N (t)] T , which consists of N mutually independent PD signals, and the The array signal representation is as follows: X(t)＝AS(t)+n(t) (1)X(t)=AS(t)+n(t) (1) 其中，t为时间变量，X(t)为M维传感器接收的N个阵列信号，即X(t)＝[x1(t),x2(t),…xN(t)]T,A为M×N阶未知混合矩阵，n(t)为M维观测噪声向量；Among them, t is the time variable, X(t) is the N array signals received by the M-dimensional sensor, that is, X(t)=[x1(t), x2(t),...xN(t)]T, A is M ×N-order unknown mixing matrix, n(t) is the M-dimensional observation noise vector; 运用观测到的阵列信号X(t)，寻求分离矩阵W，从而得到源信号的估计信号Y(t)。Using the observed array signal X(t), the separation matrix W is found to obtain the estimated signal Y(t) of the source signal. 3.按照权利要求1或2所述的一种局部超声阵列信号的高精度去噪方法，其特征在于：所述步骤一具体为3. The high-precision denoising method for a local ultrasonic array signal according to claim 1 or 2, wherein the step 1 is specifically: 1.1首先对阵列信号X(t)进行预处理，预处理主要包含去均值和白化两个步骤，去均值使观测信号满足零均值，白化消除各个数据之间的相关性，使得各个分量尽可能独立；1.1 First, preprocess the array signal X(t). The preprocessing mainly includes two steps: de-averaging and whitening. De-averaging makes the observed signal meet zero mean, and whitening eliminates the correlation between each data, so that each component is as independent as possible. ; 采用下式去除样本的均值：Use the following formula to remove the mean of the sample: 去均值后的阵列信号为X0(t)，对去均值后的数据进行白化处理，即：The array signal after de-averaging is X 0 (t), and the data after de-averaging is whitened, that is: X′(t)＝TX0(t) (3)X'(t)=TX 0 (t) (3) 白化处理后的阵列信号为X′(t)，T为线性变化矩阵；The array signal after whitening is X'(t), and T is a linear change matrix; 1.2设置分离矩阵W的初始值；1.2 Set the initial value of the separation matrix W; 1.3构建二次函数G1，G2，使得1.3 Construct quadratic functions G1, G2 such that 式中a1＝1.5,a2＝2，对G1，G2进行求导，得到导函数为g1，g2；In the formula, a1=1.5, a2=2, take the derivative of G1 and G2, and obtain the derivative function as g1, g2; 1.4运用牛顿迭代法进行循环迭代，迭代式子为W←E(xg1(WTx′))-E(g2(WTx′))W，直到收敛为止；1.4 Use the Newton iteration method for loop iteration, the iterative formula is W←E(xg 1 (W T x'))-E(g 2 (W T x'))W, until convergence; 1.5求分离后的阵列信号，即：Y(t)＝WTX(t)。1.5 Find the separated array signal, namely: Y(t)=WTX(t). 4.按照权利要求1所述的一种局部超声阵列信号的高精度去噪方法，其特征在于：所述集合经验模态分解算法EEMD为4. The high-precision denoising method of a local ultrasonic array signal according to claim 1, wherein: the collective empirical mode decomposition algorithm EEMD is: 运用FastICA去噪后得到的阵列信号，各个通道不具有相关性，对其运用单通道去噪算法去噪后不会影响阵列信号之间的相位差；The array signal obtained after denoising with FastICA has no correlation between each channel, and the phase difference between the array signals will not be affected after denoising by the single-channel denoising algorithm; 在Yi(t)中加入不同的白噪声序列fi(t)后得到Y’i(t)，对Y’i(t)进行经验模态分解，得到各阶固有模态函数分量(IMF)，此时其中Ck(t)为各阶固有模态函数分量，rn(t)为余量。After adding different white noise sequences fi(t) to Yi(t), Y'i(t) is obtained, and Y'i(t) is empirically decomposed to obtain the intrinsic modal function components (IMF) of each order, at this time where Ck(t) is the natural mode function component of each order, and rn( t ) is the margin. 5.按照权利要求1或4所述的一种局部超声阵列信号的高精度去噪方法，其特征在于：所述步骤二具体为5. The high-precision denoising method of a local ultrasonic array signal according to claim 1 or 4, wherein the step 2 is specifically: 2.1计算出Y’i(t)的所有局部极值点；2.1 Calculate all local extreme points of Y'i(t); 2.2将所有的极大值点和极小值点构成的上下包络线分别记为u0i(t)和v0i(t)；2.2 Denote the upper and lower envelopes formed by all the maximum and minimum points as u 0i (t) and v 0i (t) respectively; 2.3上下包络线的均值为单通道局放信号的上下包络线的均值的差为H0i(t)＝Yi′(t)-m0i(t)；2.3 The mean of the upper and lower envelopes is The difference between the mean values of the upper and lower envelopes of the single-channel PD signal is H 0i (t)=Y i ′(t)-m 0i (t); 2.4判断H0i(t)是否满足极值点数目或过零点数目相等或最多相差一个，且由局部极大值构成的上包络线和由局部极小值构成的下包络线的平均值为零，若满足以上条件，则H0i(t)为固有模态分量，反之重复2.1～2.3步骤，直至找到第一个固有模态分量，记为C1(t)；2.4 Judging whether H 0i (t) satisfies the number of extreme points or the number of zero-crossing points is equal or differs by at most one, and the average value of the upper envelope composed of local maxima and the lower envelope composed of local minima is zero, if the above conditions are met, then H 0i (t) is the natural modal component, otherwise, repeat steps 2.1 to 2.3 until the first natural modal component is found, denoted as C1(t); 2.5记r1(t)＝Yi′(t)-C1(t)为新的待分析的量，重复以上步骤得到第二个IMF量，……以此类推得到n个IMF量，最终剩下一个单调信号rn(t)，因此单通道原始信号可以表示为： 2.5 Denote r 1 (t)=Y i ′(t)-C 1 (t) as the new quantity to be analyzed, repeat the above steps to obtain the second IMF quantity, and so on to obtain n IMF quantities, and finally A monotonic signal r n (t) is left, so the single channel original signal can be expressed as:
CN201811398953.2A 2018-11-22 2018-11-22 High-precision denoising method for local ultrasonic array signal Active CN109583350B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811398953.2A CN109583350B (en) 2018-11-22 2018-11-22 High-precision denoising method for local ultrasonic array signal

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811398953.2A CN109583350B (en) 2018-11-22 2018-11-22 High-precision denoising method for local ultrasonic array signal

Publications (2)

Publication Number Publication Date
CN109583350A true CN109583350A (en) 2019-04-05
CN109583350B CN109583350B (en) 2022-06-10

Family

ID=65923763

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811398953.2A Active CN109583350B (en) 2018-11-22 2018-11-22 High-precision denoising method for local ultrasonic array signal

Country Status (1)

Country Link
CN (1) CN109583350B (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110221147A (en) * 2019-06-11 2019-09-10 东华大学 Power Quality Detection analysis method based on more composite optimization algorithms
CN110248325A (en) * 2019-04-22 2019-09-17 西安邮电大学 A kind of bluetooth indoor locating system based on the multiple de-noising of signal
CN110261305A (en) * 2019-06-18 2019-09-20 南京东南建筑机电抗震研究院有限公司 Based on across footpaths continuous bridge damnification recognition methods such as the multispan for influencing line
CN110940409A (en) * 2019-12-02 2020-03-31 天津市计量监督检测科学研究院 Ultrasonic signal measurement method based on ICEEMDAN and ICA combined denoising
CN112014692A (en) * 2020-07-20 2020-12-01 国网安徽省电力有限公司电力科学研究院 Blind source separation and denoising method of partial discharge UHF signal based on principal component analysis
CN112137649A (en) * 2019-06-28 2020-12-29 深圳市恩普电子技术有限公司 Ultrasonic Doppler fluid signal processing method and device
CN112710928A (en) * 2020-12-10 2021-04-27 国网宁夏电力有限公司电力科学研究院 Direct-current partial discharge waveform interference removing method and system based on autocorrelation analysis
CN118731608A (en) * 2024-07-08 2024-10-01 保定华创电气有限公司 Partial discharge detection method and device

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104375973A (en) * 2014-11-24 2015-02-25 沈阳建筑大学 Blind source signal denoising method based on ensemble empirical mode decomposition
CN108288058A (en) * 2018-04-12 2018-07-17 大连理工大学 A kind of improved wavelet threshold knee joint swinging signal Denoising Algorithm
CN108564046A (en) * 2018-04-19 2018-09-21 南京大学 Based on the steel construction dynamic strain signal processing method for improving EEMD

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104375973A (en) * 2014-11-24 2015-02-25 沈阳建筑大学 Blind source signal denoising method based on ensemble empirical mode decomposition
CN108288058A (en) * 2018-04-12 2018-07-17 大连理工大学 A kind of improved wavelet threshold knee joint swinging signal Denoising Algorithm
CN108564046A (en) * 2018-04-19 2018-09-21 南京大学 Based on the steel construction dynamic strain signal processing method for improving EEMD

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110248325A (en) * 2019-04-22 2019-09-17 西安邮电大学 A kind of bluetooth indoor locating system based on the multiple de-noising of signal
CN110248325B (en) * 2019-04-22 2022-11-25 西安邮电大学 A Bluetooth indoor positioning system based on signal multiple denoising
CN110221147A (en) * 2019-06-11 2019-09-10 东华大学 Power Quality Detection analysis method based on more composite optimization algorithms
CN110221147B (en) * 2019-06-11 2021-03-30 东华大学 Power quality detection and analysis method based on multi-composite optimization algorithm
CN110261305A (en) * 2019-06-18 2019-09-20 南京东南建筑机电抗震研究院有限公司 Based on across footpaths continuous bridge damnification recognition methods such as the multispan for influencing line
CN112137649A (en) * 2019-06-28 2020-12-29 深圳市恩普电子技术有限公司 Ultrasonic Doppler fluid signal processing method and device
CN110940409A (en) * 2019-12-02 2020-03-31 天津市计量监督检测科学研究院 Ultrasonic signal measurement method based on ICEEMDAN and ICA combined denoising
CN112014692A (en) * 2020-07-20 2020-12-01 国网安徽省电力有限公司电力科学研究院 Blind source separation and denoising method of partial discharge UHF signal based on principal component analysis
CN112710928A (en) * 2020-12-10 2021-04-27 国网宁夏电力有限公司电力科学研究院 Direct-current partial discharge waveform interference removing method and system based on autocorrelation analysis
CN112710928B (en) * 2020-12-10 2023-02-21 国网宁夏电力有限公司电力科学研究院 Direct-current partial discharge waveform interference removing method and system based on autocorrelation analysis
CN118731608A (en) * 2024-07-08 2024-10-01 保定华创电气有限公司 Partial discharge detection method and device

Also Published As

Publication number Publication date
CN109583350B (en) 2022-06-10

Similar Documents

Publication Publication Date Title
CN109583350A (en) A kind of high-precision denoising method of local ultrasound array signal
CN109884591B (en) Microphone array-based multi-rotor unmanned aerial vehicle acoustic signal enhancement method
CN106847302B (en) Single-channel mixed speech time-domain separation method based on convolutional neural network
CN108364659B (en) Frequency-domain convolution blind signal separation method based on multi-objective optimization
CN105654963B (en) Spectrum Correction and Data Density Clustering Method and Device for Underdetermined Blind Speech Recognition
CN110610717B (en) Separation method of mixed signals in complex frequency spectrum environment
CN106023996B (en) Acoustic recognition method based on cross-shaped acoustic array broadband beamforming
CN111986695B (en) Non-overlapping sub-band division rapid independent vector analysis voice blind separation method and system
CN106448694B (en) A Time-Frequency Single Source Point Extraction Method in Underdetermined Blind Source Separation Based on Complex Angle Detection
CN104573013A (en) Category weight combined integrated learning classifying method
CN103854660B (en) A kind of four Mike's sound enhancement methods based on independent component analysis
CN110176250B (en) A Robust Acoustic Scene Recognition Method Based on Local Learning
CN102231280A (en) Frequency-domain blind separation sequencing algorithm of convolutive speech signals
CN109166590B (en) A 2D Time-Frequency Mask Estimation Modeling Method Based on Spatial Domain Correlation
CN109238447A (en) A kind of blind source separation method of tether vibration signal
CN104332161B (en) It is a kind of that blind discrimination method is determined based on reception priori and the deficient of single source point detection
CN101640043A (en) Speaker recognition method based on multi-coordinate sequence kernel and system thereof
CN106098079A (en) Method and device for extracting audio signal
CN112992172A (en) Single-channel time domain bird song separating method based on attention mechanism
CN112992173B (en) Signal separation and denoising method based on improved BCA blind source separation
CN102760435A (en) Frequency-domain blind deconvolution method for voice signal
CN103994820A (en) Moving target identification method based on micro-aperture microphone array
CN115902959A (en) Method for testing radio frequency signals of satellite navigation receiver in various environments
CN111429937A (en) Voice separation method, model training method and electronic equipment
CN112201276B (en) Microphone array speech separation method based on TC-ResNet network

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant