CN114090950A - Method and device for compressing broadband noise equal phase of steady vibration signal - Google Patents
Method and device for compressing broadband noise equal phase of steady vibration signal Download PDFInfo
- Publication number
- CN114090950A CN114090950A CN202111417071.8A CN202111417071A CN114090950A CN 114090950 A CN114090950 A CN 114090950A CN 202111417071 A CN202111417071 A CN 202111417071A CN 114090950 A CN114090950 A CN 114090950A
- Authority
- CN
- China
- Prior art keywords
- sequence
- noise
- maximum value
- time domain
- index
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 56
- 230000006835 compression Effects 0.000 claims abstract description 61
- 238000007906 compression Methods 0.000 claims abstract description 61
- 108010076504 Protein Sorting Signals Proteins 0.000 claims abstract description 44
- 230000003595 spectral effect Effects 0.000 claims abstract description 33
- 230000008569 process Effects 0.000 claims description 3
- 238000004891 communication Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 abstract description 5
- 238000012544 monitoring process Methods 0.000 abstract description 4
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000001228 spectrum Methods 0.000 description 14
- 238000012545 processing Methods 0.000 description 11
- 238000004519 manufacturing process Methods 0.000 description 4
- 238000007781 pre-processing Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000001360 synchronised effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035939 shock Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention relates to the field of mechanical state monitoring, in particular to a broadband noise equal-phase compression method and a device for a steady vibration signal, wherein the method comprises the following steps of S1, acquiring a time domain signal sequence X (N) with the sequence length of l, wherein N is an element index and is a positive integer less than or equal to l, and indexes of all elements of the time domain signal sequence X (N) form an index complete set N; s2, carrying out Fourier transform on the time domain vibration sequence X (n) to obtain a frequency domain complex sequence F (n), and obtaining an amplitude sequence A (n) of the frequency domain complex sequence F (n); s3, identifying the characteristic frequency index according to the amplitude sequence A (n) to obtain a characteristic frequency index set; s4, subtracting the characteristic frequency index set from the index complete set to obtain a noise spectral line index set xi; and S5, setting a noise compression ratio, and compressing the noise in the frequency domain complex sequence F (n) according to the noise spectral line index set xi. The invention ensures that the phase information is not deformed, realizes the equal phase compression and is beneficial to the accurate positioning and the characteristic extraction of characteristic frequency spectral lines.
Description
Technical Field
The invention relates to the field of mechanical state monitoring, in particular to a method and a device for compressing broadband noise equal phase of a steady vibration signal.
Background
The fault diagnosis and monitoring technology based on signal processing is widely applied to the processing and manufacturing industry of China, and the equipment accident risk is greatly reduced and the equipment production efficiency is improved for the reliable operation of production equipment. The essence of signal processing is the transformation and extraction of information, and the more comprehensive the information extracted from the signal can reflect the equipment state, the more the relevant signal processing technology has engineering value.
The vibration signal is often used as the first-choice signal form for diagnosing and monitoring the running state of the equipment due to the characteristics of intuitive feeling, clear mechanism, flexible measuring point selection and the like. Studies have shown that mechanical vibration signals exhibit random characteristics that are not repeatable and predictable, and that due to the presence of a large number of rotating moving parts, the mean and variance of the random vibration signals will remain constant or vary periodically, when we call the vibration signal a stationary signal or a cyclostationary signal. The stable signal or the circulation stable signal acquired at high frequency contains rich equipment running state information, however, the production field is complex, and the sensor inevitably collects interference information such as background noise, electrical noise, control fluctuation, nonlinear pulse and the like in the stable vibration signal acquisition process, so that effective identification and compression of noise are important preprocessing steps for carrying out equipment state evaluation based on the stable vibration signal.
Common noise compression methods, such as EMD denoising, wavelet denoising, etc., decompose the vibration waveform from different angles, compress or discard part of the decomposed components according to application requirements, and then perform inverse reconstruction on the basis of the remaining components, thereby achieving the purpose of noise elimination or compression, and can be applied to various scenes, such as stationary, non-stationary, etc. However, the fourier transform is known to be one of the most reliable, robust and widely used methods for fault state analysis based on stationary vibration signals, and noise compression from the fourier transform perspective has higher practical value.
Disclosure of Invention
The invention aims to provide a method and a device for compressing broadband noise and other phases of a stable vibration signal applied to mechanical equipment vibration signal processing from the perspective of novel Fourier transform. In order to achieve the above purpose, the invention provides the following technical scheme:
a broadband noise equiphase compression method for a steady vibration signal comprises the following steps:
s1, acquiring a time domain signal sequence X (N) with a sequence length of l, wherein N is an element index and is a positive integer less than or equal to l, and indexes of all elements of the time domain signal sequence X (N) form an index complete set N;
s2, carrying out Fourier transform on the time domain vibration sequence X (n) to obtain a frequency domain complex sequence F (n), and obtaining an amplitude sequence A (n) of the frequency domain complex sequence F (n);
s3, identifying a characteristic frequency index according to the amplitude sequence A (n) to obtain a characteristic frequency index set;
s4, subtracting the characteristic frequency index set from the index complete set to obtain a noise spectral line index set xi;
and S5, setting a noise compression ratio, and compressing the noise in the frequency domain complex sequence F (n) according to the noise spectral line index set xi.
Further, step S3 specifically includes the following steps:
s31, calculating a first order difference sequence d (n) of the amplitude sequence A (n);
s32, acquiring an index set I corresponding to the maximum value in the amplitude sequence A (n) based on the first-order difference sequence d (n), thereby obtaining a maximum value sequence A (I);
s33, obtaining the median of the maximum value sequence A (I), and marking as mjWherein j equals 1;
s34, using the maximum value sequence A (I) as a new amplitude value sequence;
s35, adding 1 to the value of j, then calculating the first order difference sequence of the new amplitude sequence, and acquiring an index set corresponding to the maximum value in the new amplitude sequence based on the first order difference sequence of the new amplitude sequence, thereby obtaining a new maximum value sequence;
s36, obtaining the median of the new maximum value sequence, and recording as mj；
S37, when the value of j is less than 2, the new maximum value sequence is used as a new amplitude value sequence, and then the step S35 is executed;
when the value of j is greater than or equal to 2, judging whether m is satisfiedj-mj-1≧2(mj-1-mj-2) If not, the new maximum value sequence is taken as a new amplitude value sequence, and then step S35 is performed; if yes, the indexes corresponding to the elements in the new maximum value sequence are formed into the characteristic frequency index set, and then the step S3 is ended.
Further, in step S32, based on the first-order difference sequence d (n), the method for obtaining the index set I corresponding to the maximum value in the amplitude sequence a (n) includes:
I＝find(d1·d2<0&d1>0)
wherein d is1Is a subsequence of the 1 st to the l-1 st elements of the first order difference sequence d (n), d2For subsequences of the 2 nd to l th elements of said first order difference sequence d (n), find represents a non-zero element in the search sequence and returns an index,. represents a bitwise multiplication of the sequence elements,&representing a bitwise logical and operation.
Further, in step S35, based on the first-order difference sequence of the new amplitude sequence, the method for obtaining the index set corresponding to the maximum value in the new amplitude sequence includes:
I＝find(d1·d2<0&d1>0)
wherein, I represents the index set corresponding to the maximum value in the new amplitude sequence, d1For subsequences of 1 st to l-1 st elements in the first order difference sequence of the new magnitude sequence, d2Is the new amplitude valueSubsequences of 2 nd to l th elements in a first order difference sequence of the sequence, find denotes searching for a non-zero element in the sequence and returning an index,. denotes bitwise multiplication of the sequence elements,&representing a bitwise logical and operation.
Preferably, before step S2, the method further includes a step of determining stationarity of the time domain signal sequence x (n), and when the signal satisfies a stationary signal determination condition, step S2 is executed, otherwise, the method is ended.
Further, the determining the stationarity of the time domain signal sequence x (n) includes:
selecting 0.01 l-0.05 l of local window width, and calculating a local mean sequence and a local variance sequence of the time domain signal sequence X (n), wherein the peak-to-peak value of the local mean sequence is PMThe peak value of the local variance sequence is PS(ii) a The peak-to-peak value of the time domain signal sequence X (n) is PX；
The stationary signal judgment condition is as follows: pM<0.01PXAnd P isS<0.01PX。
Further, in step S5, the noise is compressed according to the following formula:
F'(k)＝(1-η)F(k)
wherein, F' (k) represents the frequency domain complex sequence after compression, η is the noise compression ratio, F (k) is the frequency domain complex sequence, and k ∈ ξ.
Preferably, the method for equal-phase compression of wideband noise further includes: s6, inverse fourier transform is performed on the frequency domain complex sequence F '(k) after noise compression to obtain a time domain vibration sequence X' (n) after noise compression.
Preferably, the method for equal-phase compression of wideband noise further includes: s7, performing energy compensation on the time domain vibration sequence X' (n) after the noise compression by using an energy compensation formula, wherein the energy compensation formula is as follows:
wherein, Xd(n) is the compensated time domain vibration sequence;
the compensation coefficient is calculated according to the following formula:
where RMS stands for root mean square operation.
Based on the same inventive concept, the invention provides a broadband noise equiphase compression device for a stable vibration signal, which comprises at least one processor and a memory which is in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform any of the methods described above.
Compared with the prior art, the invention has the beneficial effects that:
1. according to the method, a time domain signal sequence is converted into a frequency domain signal sequence through Fourier transform, a characteristic frequency index is identified according to an amplitude sequence of the frequency domain signal sequence, then a noise spectral line index set is obtained by subtracting the characteristic frequency index set from a full index set of the time domain signal sequence, and then noise compression is carried out according to the noise spectral line index set, so that spectrum background noise can be effectively reduced on the premise of not influencing main characteristic spectral lines of a frequency domain, and the integrity of key characteristic frequency is ensured; during compression, synchronous equal-proportion compression is carried out on the real part and the imaginary part of the frequency domain complex sequence, so that the frequency spectrum phase is not influenced, the phase information is not deformed, the equal-phase compression is realized, the accurate positioning and the characteristic extraction of a characteristic frequency spectral line are facilitated, and the method is an effective stable vibration signal preprocessing means.
2. The invention converts the time domain signal sequence into the frequency domain signal sequence through Fourier transform, searches the maximum value of the frequency domain signal sequence, searches the maximum value again in the sequence formed by the maximum values, and sets a proper judgment condition mj-mj-1≧2(mj-1-mj-2) The method can accurately screen out the characteristic frequency indexes without mistakenly rejecting useful characteristic frequencies, and can filter the noise spectral line indexes to ensure the integrity of the key characteristic frequencies.
3. Because the Fourier transform is one of the most reliable, robust and widely used methods for analyzing the fault state based on the stable vibration signal, the invention provides a broadband noise equal-phase compression method of the stable vibration signal applied to the vibration signal processing of mechanical equipment from the perspective of the Fourier transform, and ensures the usability on the signal processing.
Drawings
FIG. 1 is a flow chart of the method for compressing the broadband noise equal phase of the stationary vibration signal according to the present invention.
Fig. 2 is a flowchart of a method for compressing wideband noise of a stationary vibration signal in equal phase according to another embodiment.
Fig. 3 is a local mean and local variance waveform diagram of a time-domain signal sequence of example 2.
Fig. 4 is a passband amplitude spectrum of the amplitude sequence in the frequency domain complex sequence of example 2.
Fig. 5 is a graph showing the variation of the maximum value sequence in the cycle of example 2.
Fig. 6 is a graph of the median change of the maximum sequence in the cycle of example 2.
Fig. 7 shows the noise line identification result in example 2.
Fig. 8 is an amplitude spectrum before noise compression and a partially enlarged view in example 2.
Fig. 9 is an amplitude spectrum and a partial enlarged view after noise compression in example 2.
Fig. 10 is a comparison graph of the whole time domain waveform before and after noise compression in example 2.
Fig. 11 is a partial time domain waveform diagram before noise compression in example 2.
Fig. 12 is a local time domain waveform diagram after noise compression in example 2.
Fig. 13 is a local phase spectrum before noise compression in example 2.
Fig. 14 is a local phase spectrum after noise compression in example 2.
Detailed Description
The present invention will be described in further detail with reference to test examples and specific embodiments. It should be understood that the scope of the above-described subject matter is not limited to the following examples, and any techniques implemented based on the disclosure of the present invention are within the scope of the present invention.
Example 1
The embodiment provides a method for compressing wideband noise and other phases of a stationary vibration signal, as shown in fig. 1, including the following steps:
s1, acquiring a time domain signal sequence X (N) with a sequence length of l, wherein N is an element index and is a positive integer less than or equal to l, and indexes of all elements of the time domain signal sequence X (N) form an index complete set N;
specifically, the time domain signal sequence x (N), N ═ 1,2, …, l, the indices 1,2, …, l of all elements are the index corpus N, N ∈ N.
Before executing step S2, the method further includes a step of determining stationarity of the time domain signal sequence x (n), when the signal satisfies a stationary signal determination condition, executing step S2, otherwise, ending;
specifically, the determining the stationarity of the time domain signal sequence x (n) includes:
selecting 0.01 l-0.05 l of local window width, and calculating a local mean sequence and a local variance sequence of the time domain signal sequence X (n), wherein the peak-to-peak value of the local mean sequence is PMThe peak value of the local variance sequence is PS(ii) a The peak-to-peak value of the time domain signal sequence X (n) is PX；
The stationary signal judgment condition is as follows: pM<0.01PXAnd P isS<0.01PXIf the determination condition is satisfied, step S2 is executed, otherwise, the process ends.
S2, performing Fourier transform on the time domain vibration sequence X (n) to obtain a frequency domain complex sequence F (n), and acquiring an amplitude sequence A (n) of the frequency domain complex sequence F (n).
S3, identifying a characteristic frequency index according to the amplitude sequence A (n) to obtain a characteristic frequency index set;
specifically, S3 includes the steps of:
s31, calculating a first order difference sequence d (n) of the amplitude sequence A (n);
s32, acquiring an index set I corresponding to the maximum value in the amplitude sequence A (n) based on the first-order difference sequence d (n), thereby obtaining a maximum value sequence A (I);
specifically, the method for obtaining the index set I corresponding to the maximum value in the amplitude sequence a (n) includes:
I＝find(d1·d2<0&d1>0)
wherein d is1Is a subsequence of the 1 st to the l-1 st elements of the first order difference sequence d (n), d2For subsequences of the 2 nd to l th elements of said first order difference sequence d (n), find represents a non-zero element in the search sequence and returns an index,. represents a bitwise multiplication of the sequence elements,&representing a bitwise logical and operation.
S33, obtaining the median of the maximum value sequence A (I), and marking as mjWherein j equals 1;
s34, using the maximum value sequence A (I) as a new amplitude value sequence;
s35, adding 1 to the value of j, then calculating the first order difference sequence of the new amplitude sequence, and acquiring an index set corresponding to the maximum value in the new amplitude sequence based on the first order difference sequence of the new amplitude sequence, thereby obtaining a new maximum value sequence;
similar to step S32, the method for obtaining the index set corresponding to the maximum value in the new amplitude sequence based on the first-order difference sequence of the new amplitude sequence includes:
I＝find(d1·d2<0&d1>0)
wherein, I represents the index set corresponding to the maximum value in the new amplitude sequence, d1For subsequences of 1 st to l-1 st elements in the first order difference sequence of the new magnitude sequence, d2For subsequences of 2 nd to l th elements in the first order difference sequence of the new magnitude sequence, find represents a non-zero element in the search sequence and returns an index, indicate that the sequence elements are in phaseThe result of the multiplication is,&representing a bitwise logical and operation.
S36, obtaining the median of the new maximum value sequence, and recording as mj；
S37, when the value of j is less than 2, the new maximum value sequence is used as a new amplitude value sequence, and then the step S35 is executed;
when the value of j is greater than or equal to 2, judging whether m is satisfiedj-mj-1≧2(mj-1-mj-2) If not, the new maximum value sequence is taken as a new amplitude value sequence, and then step S35 is performed; if yes, the indexes corresponding to the elements in the new maximum value sequence are formed into the characteristic frequency index set, and then the step S3 is ended.
The invention converts the time domain signal sequence into the frequency domain signal sequence through Fourier transform, searches the maximum value of the frequency domain signal sequence, searches the maximum value again in the sequence formed by the maximum values, and sets a proper judgment condition mj-mj-1≧2(mj-1-mj-2) The method can accurately screen out the characteristic frequency indexes without mistakenly rejecting useful characteristic frequencies, and can filter the noise spectral line indexes to ensure the integrity of the key characteristic frequencies.
S4, subtracting the characteristic frequency index set from the index complete set to obtain a noise spectral line index set xi;
s5, setting a noise compression ratio, and compressing the noise in the frequency domain complex sequence F (n) according to the noise spectral line index set xi;
specifically, the noise is compressed according to the following formula:
F'(k)＝(1-η)F(k)
wherein, F' (k) represents the frequency domain complex sequence after compression, η is the noise compression ratio, F (k) is the frequency domain complex sequence, and k ∈ ξ. F' (k) is a complex number, and the real part and the imaginary part are synchronously compressed in equal proportion without influencing the spectrum phase, namely, so-called equal phase compression.
After step S5 is completed, inverse fourier transform is performed on the frequency domain complex sequence F '(k) after noise compression to obtain a time domain vibration sequence X' (n) after noise compression.
Performing energy compensation on the time domain vibration sequence X' (n) after noise compression, wherein the formula of the energy compensation is as follows:
wherein, Xd(n) is the compensated time domain vibration sequence;
the compensation coefficient is calculated according to the following formula:
where RMS stands for root mean square operation.
According to the method, a time domain signal sequence is converted into a frequency domain signal sequence through Fourier transform, a characteristic frequency index is identified according to an amplitude sequence of the frequency domain signal sequence, then a noise spectral line index set is obtained by subtracting the characteristic frequency index set from a full index set of the time domain signal sequence, and then noise compression is carried out according to the noise spectral line index set, so that spectrum background noise can be effectively reduced on the premise of not influencing main characteristic spectral lines of a frequency domain, and the integrity of key characteristic frequency is ensured; during compression, synchronous equal-proportion compression is carried out on the real part and the imaginary part of the frequency domain complex sequence, so that the frequency spectrum phase is not influenced, the phase information is not deformed, the equal-phase compression is realized, the accurate positioning and the characteristic extraction of a characteristic frequency spectral line are facilitated, and the method is an effective stable vibration signal preprocessing means.
Because the Fourier transform is one of the most reliable, robust and widely used methods for analyzing the fault state based on the stable vibration signal, the invention provides a broadband noise equal-phase compression method of the stable vibration signal applied to the vibration signal processing of mechanical equipment from the perspective of the Fourier transform, and ensures the usability on the signal processing.
Example 2
Taking the acceleration vibration signal of the electric spindle of a certain machine tool as an example, the steps of the method of the invention are applied, and as shown in fig. 2, the noise is compressed in equal phase.
The length of the time domain signal sequence obtained in step S1 is 500000, the sampling frequency is 25600Hz, i.e., the sampling duration is about 19.5S, the window width is 5000, the step length is 50, and the local mean value and the local variance of the signal are calculated as shown in fig. 3. Step S2 is to perform FFT to obtain a frequency domain complex sequence, and draw an amplitude spectrum for the amplitude sequence of the frequency domain complex sequence as shown in fig. 4, which shows that the spectral line has uneven amplitude distribution, and has the characteristics of "low-frequency low-amplitude, high-frequency high-amplitude", and a large amount of background noise exists at the bottom.
Entering a noise spectral line identification loop, that is, step S3 described in embodiment 1, and drawing a maximum value sequence in the first 3 loops (that is, j is 1,2,3) around the crossover harmonic 500Hz, as shown in fig. 5, it can be seen that as the loop progresses, the length of the maximum value sequence is continuously shortened, and the noise spectral line around the 500Hz spectral line is gradually identified and removed. Calculating median m of maximum sequence in each cyclejAs shown in fig. 6, the median of the maximum value sequence increases with the progress of the cycle and tends to increase in an accelerated manner, and the main cause of this phenomenon is that the ratio of the high amplitude characteristic frequency representing the principal axis state in the sequence increases. Wherein m is5-m4＝0.0039-0.0023＝0.0017，m4-m3＝0.0023-0.0017＝0.0008，m5-m4>2(m4-m3) Triggering the cycle termination condition of step S37 in embodiment 1 to obtain a characteristic frequency index set, continuing to execute step S4 to obtain a noise spectral line index set, and outputting a noise spectral line identification result as shown in fig. 7. As can be seen from fig. 7, the noise line is composed of two parts: the low-amplitude background noise line of the passband and a plurality of interference lines with slightly larger amplitudes, which all have spectral lines corresponding to each other in fig. 8, are usually generated by slight fluctuation of the rotation speed or insufficient resolution, and interfere with the accurate positioning of the spectral lines. Comparison ofAs can be seen from fig. 8 and 9, the present invention converts the time domain signal sequence into the frequency domain signal sequence by fourier transform, searches for the maximum value in the frequency domain signal sequence, searches for the maximum value again in the sequence consisting of the maximum values, and sets the appropriate judgment condition mj-mj -1≧2(mj-1-mj-2) The method can accurately screen out the characteristic frequency indexes without mistakenly rejecting useful characteristic frequencies, can filter the noise spectral line indexes, effectively reduces the background noise of the frequency spectrum on the premise of not influencing the main characteristic spectral lines of the frequency domain, and ensures the integrity of the key characteristic frequency.
In step S5, noise compression is performed with the noise compression ratio η set to 0.95, and when the time domain waveforms before and after compression are compared as shown in fig. 10, it is found that the signal energy after noise compression is significantly lost, and when the time domain waveforms before and after compression are amplified as shown in fig. 11 and 12, respectively, it is found that the energy after compression is lost, but the periodicity of the shock is improved to some extent. In step S6, compensation is performed for the loss energy, and the compensation coefficient is calculated as
As shown in fig. 13 and 14, the phase step changes at characteristic frequencies of 250 Hz and 500Hz are completely the same as the local phases before and after compression against noise.
According to the method, a time domain signal sequence is converted into a frequency domain signal sequence through Fourier transform, a characteristic frequency index is identified according to an amplitude sequence of the frequency domain signal sequence, then a noise spectral line index set is obtained by subtracting the characteristic frequency index set from a full index set of the time domain signal sequence, and then noise compression is carried out according to the noise spectral line index set, so that spectrum background noise can be effectively reduced on the premise of not influencing main characteristic spectral lines of a frequency domain, and the integrity of key characteristic frequency is ensured; during compression, synchronous equal-proportion compression is carried out on the real part and the imaginary part of the frequency domain complex sequence, so that the frequency spectrum phase is not influenced, the phase information is not deformed, the equal-phase compression is realized, the accurate positioning and the characteristic extraction of a characteristic frequency spectral line are facilitated, and the method is an effective stable vibration signal preprocessing means.
Because the Fourier transform is one of the most reliable, robust and widely used methods for analyzing the fault state based on the stable vibration signal, the invention provides a broadband noise equal-phase compression method of the stable vibration signal applied to the vibration signal processing of mechanical equipment from the perspective of the Fourier transform, and ensures the usability on the signal processing.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (10)
1. A broadband noise equiphase compression method for a steady vibration signal is characterized by comprising the following steps:
s1, acquiring a time domain signal sequence X (N) with a sequence length of l, wherein N is an element index and is a positive integer less than or equal to l, and indexes of all elements of the time domain signal sequence X (N) form an index complete set N;
s2, carrying out Fourier transform on the time domain vibration sequence X (n) to obtain a frequency domain complex sequence F (n), and obtaining an amplitude sequence A (n) of the frequency domain complex sequence F (n);
s3, identifying a characteristic frequency index according to the amplitude sequence A (n) to obtain a characteristic frequency index set;
s4, subtracting the characteristic frequency index set from the index complete set to obtain a noise spectral line index set xi;
and S5, setting a noise compression ratio, and compressing the noise in the frequency domain complex sequence F (n) according to the noise spectral line index set xi.
2. The method as claimed in claim 1, wherein the step S3 comprises the following steps:
s31, calculating a first order difference sequence d (n) of the amplitude sequence A (n);
s32, acquiring an index set I corresponding to the maximum value in the amplitude sequence A (n) based on the first-order difference sequence d (n), thereby obtaining a maximum value sequence A (I);
s33, obtaining the median of the maximum value sequence A (I), and marking as mjWherein j equals 1;
s34, using the maximum value sequence A (I) as a new amplitude value sequence;
s35, adding 1 to the value of j, then calculating the first order difference sequence of the new amplitude sequence, and acquiring an index set corresponding to the maximum value in the new amplitude sequence based on the first order difference sequence of the new amplitude sequence, thereby obtaining a new maximum value sequence;
s36, obtaining the median of the new maximum value sequence, and recording as mj；
S37, when the value of j is less than 2, the new maximum value sequence is used as a new amplitude value sequence, and then the step S35 is executed;
when the value of j is greater than or equal to 2, judging whether m is satisfiedj-mj-1≧2(mj-1-mj-2) If not, the new maximum value sequence is taken as a new amplitude value sequence, and then step S35 is performed; if yes, the indexes corresponding to the elements in the new maximum value sequence are formed into the characteristic frequency index set, and then the step S3 is ended.
3. The method as claimed in claim 2, wherein in step S32, based on the first order difference sequence d (n), the method for obtaining the index set I corresponding to the maximum value in the amplitude sequence a (n) comprises:
I＝find(d1·d2<0&d1>0)
wherein d is1Is a subsequence of the 1 st to the l-1 st elements of the first order difference sequence d (n), d2For subsequences of the 2 nd to l th elements of said first order difference sequence d (n), find represents a non-zero element in the search sequence and returns an index,. represents a bitwise multiplication of the sequence elements,&representing a bitwise logical and operation.
4. The method as claimed in claim 2, wherein in step S35, the method for obtaining the index set corresponding to the maximum value in the new amplitude sequence based on the first order difference sequence of the new amplitude sequence includes:
I＝find(d1·d2<0&d1>0)
wherein, I represents the index set corresponding to the maximum value in the new amplitude sequence, d1For subsequences of 1 st to l-1 st elements in the first order difference sequence of the new magnitude sequence, d2Find represents the non-zero element in the search sequence and returns the index, for the subsequences of the 2 nd to l th elements in the first order difference sequence of the new magnitude sequence,. represents the bitwise multiplication of the sequence elements,&representing a bitwise logical and operation.
5. The method as claimed in claim 1, further comprising a step of determining the stationarity of the time domain signal sequence x (n) before the step S2, wherein the step S2 is executed when the signal satisfies the stationary signal determination condition, otherwise, the process is ended.
6. The method as claimed in claim 5, wherein the determining the stationarity of the time domain signal sequence x (n) comprises:
selecting 0.01 l-0.05 l of local window width, and calculating a local mean sequence and a local variance sequence of the time domain signal sequence X (n), wherein the peak-to-peak value of the local mean sequence is PMThe peak value of the local variance sequence is PS(ii) a The peak-to-peak value of the time domain signal sequence X (n) is PX；
The stationary signal judgment condition is as follows: pM<0.01PXAnd P isS<0.01PX。
7. The method for compressing broadband noise of steady vibration signals in equal phase as claimed in any one of claims 1 to 6, wherein in step S5, the noise is compressed according to the following formula:
F'(k)＝(1-η)F(k)
wherein, F' (k) represents the frequency domain complex sequence after compression, η is the noise compression ratio, F (k) is the frequency domain complex sequence, and k ∈ ξ.
8. The method of claim 7, wherein the method further comprises: s6, inverse fourier transform is performed on the frequency domain complex sequence F '(k) after noise compression to obtain a time domain vibration sequence X' (n) after noise compression.
9. The method of claim 8, wherein the method further comprises: s7, performing energy compensation on the time domain vibration sequence X' (n) after the noise compression by using an energy compensation formula, wherein the energy compensation formula is as follows:
wherein, Xd(n) is the compensated time domain vibration sequence;
the compensation coefficient is calculated according to the following formula:
where RMS stands for root mean square operation.
10. A broadband noise equiphase compression device for a stationary vibration signal is characterized by comprising at least one processor and a memory which is in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111417071.8A CN114090950B (en) | 2021-11-25 | Broadband noise equal-phase compression method and device for stable vibration signals |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111417071.8A CN114090950B (en) | 2021-11-25 | Broadband noise equal-phase compression method and device for stable vibration signals |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114090950A true CN114090950A (en) | 2022-02-25 |
| CN114090950B CN114090950B (en) | 2024-07-16 |
Family
ID=
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6339756B1 (en) * | 1995-04-10 | 2002-01-15 | Corporate Computer Systems | System for compression and decompression of audio signals for digital transmission |
| RU2364957C1 (en) * | 2007-12-27 | 2009-08-20 | Общество с ограниченной ответственностью "Стэл-Компьютерные Системы" | Determination method of parameters of lined voiced sounds spectrums and system for its realisation |
| CN105158740A (en) * | 2015-08-24 | 2015-12-16 | 西安电子科技大学 | High-precision frequency estimation-based noise amplitude-modulation interference suppression method |
| CN105651376A (en) * | 2014-11-10 | 2016-06-08 | 上海宝钢工业技术服务有限公司 | Mechanical equipment offline detection system vibration signal spectrum analysis and alarm method |
| CN107874750A (en) * | 2017-11-28 | 2018-04-06 | 华南理工大学 | Pulse frequency variability and the psychological pressure monitoring method and device of sleep quality fusion |
| CN109190598A (en) * | 2018-09-29 | 2019-01-11 | 西安交通大学 | A kind of rotating machinery monitoring data noise detection method based on SES-LOF |
| CN110068727A (en) * | 2019-04-09 | 2019-07-30 | 东南大学 | A kind of simple signal frequency estimating methods based on the comprehensive interpolation of Candan-Rife |
| CN110221125A (en) * | 2019-07-15 | 2019-09-10 | 福州大学 | A kind of direct current system harmonic analysis method restored using DC component |
| CN110231117A (en) * | 2019-06-10 | 2019-09-13 | 中设设计集团股份有限公司 | A kind of drag-line fundamental frequency feature identification method based on S-transformation |
| CN110333389A (en) * | 2019-05-07 | 2019-10-15 | 东南大学 | Sinusoidal signal frequency estimation method based on interpolated DFT |
| CN112968886A (en) * | 2021-02-02 | 2021-06-15 | 东北大学 | Vibration signal compression method and device, storage medium and computer equipment |
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6339756B1 (en) * | 1995-04-10 | 2002-01-15 | Corporate Computer Systems | System for compression and decompression of audio signals for digital transmission |
| RU2364957C1 (en) * | 2007-12-27 | 2009-08-20 | Общество с ограниченной ответственностью "Стэл-Компьютерные Системы" | Determination method of parameters of lined voiced sounds spectrums and system for its realisation |
| CN105651376A (en) * | 2014-11-10 | 2016-06-08 | 上海宝钢工业技术服务有限公司 | Mechanical equipment offline detection system vibration signal spectrum analysis and alarm method |
| CN105158740A (en) * | 2015-08-24 | 2015-12-16 | 西安电子科技大学 | High-precision frequency estimation-based noise amplitude-modulation interference suppression method |
| CN107874750A (en) * | 2017-11-28 | 2018-04-06 | 华南理工大学 | Pulse frequency variability and the psychological pressure monitoring method and device of sleep quality fusion |
| CN109190598A (en) * | 2018-09-29 | 2019-01-11 | 西安交通大学 | A kind of rotating machinery monitoring data noise detection method based on SES-LOF |
| CN110068727A (en) * | 2019-04-09 | 2019-07-30 | 东南大学 | A kind of simple signal frequency estimating methods based on the comprehensive interpolation of Candan-Rife |
| CN110333389A (en) * | 2019-05-07 | 2019-10-15 | 东南大学 | Sinusoidal signal frequency estimation method based on interpolated DFT |
| CN110231117A (en) * | 2019-06-10 | 2019-09-13 | 中设设计集团股份有限公司 | A kind of drag-line fundamental frequency feature identification method based on S-transformation |
| CN110221125A (en) * | 2019-07-15 | 2019-09-10 | 福州大学 | A kind of direct current system harmonic analysis method restored using DC component |
| CN112968886A (en) * | 2021-02-02 | 2021-06-15 | 东北大学 | Vibration signal compression method and device, storage medium and computer equipment |
Non-Patent Citations (1)
| Title |
|---|
| 季鸣;何玉珠;: "基于DFT插值的频率测量技术", 电子测量技术, vol. 39, no. 6, 15 June 2016 (2016-06-15), pages 48 - 50 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bin et al. | Early fault diagnosis of rotating machinery based on wavelet packets—Empirical mode decomposition feature extraction and neural network | |
| CN107356432A (en) | Fault Diagnosis of Roller Bearings based on frequency domain window experience small echo resonance and demodulation | |
| CN110991481B (en) | High-voltage shunt reactor internal loosening fault diagnosis method based on cross wavelet transformation | |
| CN109374119A (en) | Transformer vibration signal Characteristic Extraction method | |
| CN104714075B (en) | A kind of electric network voltage flicker envelope parameters extracting method | |
| CN111256993A (en) | Method and system for diagnosing fault type of main bearing of wind turbine generator | |
| CN105147252A (en) | Heart disease recognition and assessment method | |
| CN106137184B (en) | Electrocardiosignal QRS complex detection method based on wavelet transformation | |
| CN114486263B (en) | Noise reduction demodulation method for vibration signal of rolling bearing of rotary machine | |
| CN116049632A (en) | Wind power main shaft bearing fault diagnosis method, device and application | |
| CN114486260B (en) | Bearing fault diagnosis method based on self-adaptive variational modal decomposition | |
| CN102988041A (en) | Selective cardiac-magnetic signal averaging method in signal noise suppression | |
| Lv et al. | Generalized synchroextracting-based stepwise demodulation transform and its application to fault diagnosis of rotating machinery | |
| CN117349615B (en) | Self-adaptive enhancement envelope spectrum method for fault diagnosis of rolling bearing under strong noise condition | |
| CN117571316A (en) | Composite fault diagnosis method and system | |
| Zheng et al. | Zero-phase filter-based adaptive Fourier decomposition and its application to fault diagnosis of rolling bearing | |
| CN110376437B (en) | Order analysis method for overcoming non-order frequency component interference | |
| CN114090950A (en) | Method and device for compressing broadband noise equal phase of steady vibration signal | |
| CN114090950B (en) | Broadband noise equal-phase compression method and device for stable vibration signals | |
| CN108008262B (en) | Rapid partial discharge detection method | |
| CN114061746B (en) | Repeated transient signal extraction method in rotary machinery fault diagnosis | |
| CN106772032B (en) | Fault feature extraction method for hydroelectric generating set | |
| CN112116917B (en) | Phase jump degree-based method for separating acoustic signals of reactor body and fan | |
| CN114019236A (en) | Power grid harmonic single-channel aliasing target signal detection method and device | |
| CN110263711B (en) | Coupling signal impact feature extraction method based on improved spectral kurtosis |
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 |