CN113688669A - Hydroelectric generating set vibration signal energy characteristic identification method - Google Patents
Hydroelectric generating set vibration signal energy characteristic identification method Download PDFInfo
- Publication number
- CN113688669A CN113688669A CN202110789816.7A CN202110789816A CN113688669A CN 113688669 A CN113688669 A CN 113688669A CN 202110789816 A CN202110789816 A CN 202110789816A CN 113688669 A CN113688669 A CN 113688669A
- Authority
- CN
- China
- Prior art keywords
- energy
- signal
- eigenmode
- signals
- generating set
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 44
- 239000011159 matrix material Substances 0.000 claims abstract description 29
- 238000012545 processing Methods 0.000 claims abstract description 24
- 230000004069 differentiation Effects 0.000 claims description 11
- 239000013598 vector Substances 0.000 claims description 10
- 230000008569 process Effects 0.000 description 15
- 238000010586 diagram Methods 0.000 description 8
- 238000012216 screening Methods 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 5
- 238000003745 diagnosis Methods 0.000 description 5
- 238000000605 extraction Methods 0.000 description 5
- 238000001514 detection method Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
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
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
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)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a hydroelectric generating set vibration signal energy characteristic identification method, which comprises the following steps: acquiring an original signal to obtain an original signal of the vibration of the hydroelectric generating set; differential signals are obtained, and differential processing is carried out on the collected original signals to obtain differential signals; empirical mode decomposition, namely performing empirical mode decomposition on the signals, and forming an eigenmode function set by using a plurality of obtained eigenmode functions; calculating energy ratio, wherein the total energy of the signals in the eigenmode function set and the ratio of the energy of each eigenmode function in the total energy are calculated; and combining the energy characteristic matrix, the total signal energy and the eigenmode function energy proportion into an energy characteristic matrix. The invention enhances the high-frequency component of the signal by carrying out multiple differential processing on the original signal, recovers the high-frequency extreme point submerged in the original signal, and then carries out empirical mode decomposition on the original signal and the differential signal to solve the problem of mode confusion, thereby improving the adaptability and accuracy of the empirical mode decomposition.
Description
Technical Field
The invention relates to the field of hydroelectric generating set fault diagnosis, in particular to a method for identifying energy characteristics of a vibration signal of a hydroelectric generating set.
Background
The key of the state monitoring and fault diagnosis of the hydroelectric generating set is the processing of collected signals, effective fault characteristics are extracted through a signal processing means to position fault points, and the overhauling efficiency can be improved and faults can be predicted. Empirical mode decomposition is the core of the Hilbert-Huang algorithm, which carries out signal decomposition according to the time scale characteristics of data per se to form a limited number of eigenmode functions, and each decomposed eigenmode function component comprises local characteristic signals of the original signal at different time scales. The conditional constraints of the eigenmode functions include: instead, the average of the local maximum envelope and the local minimum envelope is zero, and the waveform of the signal is locally symmetric.
The rolling bearing fault diagnosis method based on EEMD and improved GSA-SOM neural network disclosed in Chinese patent literature has publication number CN112345252A, published on 2021-02-09, and decomposes a non-stationary original vibration signal into a plurality of stationary eigenmode functions (IMFs) through Empirical Mode Decomposition (EMD); then, extracting energy characteristics to obtain a characteristic vector reflecting the vibration signal; further, the weight of the neural network is optimized by adopting an improved GSA algorithm, and finally the obtained characteristic vector is input into the improved neural network for automatic fault identification. However, the hydropower monitoring signals are influenced by factors such as intermittent fault signals and noise interference, so that modal aliasing easily occurs in eigenmode function concentration: signal components of different frequencies may be contained in the same eigenmode function, while signal components of a certain frequency may also occur in different eigenmode functions; the appearance of modal aliasing weakens the characterization capability of the eigenmode function set on the time-frequency characteristics of the original signal, and influences the effectiveness of the EMD-based signal processing and feature extraction method.
Disclosure of Invention
The invention provides a hydroelectric generating set vibration signal energy characteristic identification method, aiming at overcoming the problems that in the prior art, a hydroelectric generating set vibration signal is easily influenced by factors such as intermittent fault signals, noise interference and the like, and deviation is difficult to avoid when a local extreme value of fault response is solved, so that modal aliasing is easily generated on an eigenmode function.
In order to achieve the purpose, the invention adopts the following technical scheme:
a method for recognizing the energy characteristics of a vibration signal of a hydroelectric generating set comprises the following steps:
s1, collecting an original signal to obtain an original signal of the vibration of the hydroelectric generating set;
s2, calculating a differential signal, and carrying out differential processing on the acquired original signal to obtain a differential signal;
s3, performing empirical mode decomposition, namely performing empirical mode decomposition on the signals obtained in S1 and S2, and forming an eigenmode function set by using a plurality of obtained eigenmode functions;
s4, calculating energy ratio, and calculating the total energy of the signals in the eigenmode function set and the ratio of the energy of each eigenmode function in the total energy at S3;
and S5, combining the total signal energy and the eigenmode function energy proportion in S4 to form an energy feature matrix, observing the energy feature matrix, acquiring a signal energy distribution frequency band, and extracting effective signal features.
The method adopts Empirical Mode Decomposition (EMD) to process signals, and the EMD is a self-adaptive data processing or mining method and is very suitable for processing nonlinear and non-stationary time sequences, so that the method has great advantages in the aspects of processing vibration signals of the hydroelectric generating set and extracting characteristics. The key idea of empirical mode decomposition is to decompose a complex signal with irregular frequency into the form of the sum of eigenmode functions (IMF) of a limited single frequency and residual waves, wherein each decomposed eigenmode function comprises local characteristic signals of the original signal at different time scales.
Preferably, in S1, the raw signal of the vibration of the hydroelectric generating set includes an x-axis vibration signal and a y-axis vibration signal of the hydroelectric generating set.
The original signals of the invention adopt the x-axis vibration signal and the y-axis vibration signal of the hydroelectric generating set at the same time, because the working state and the condition of the hydroelectric generating set reflected by the vibration of the x-axis and the vibration of the y-axis are not completely the same, the comprehensive performance of state monitoring and fault diagnosis of the hydroelectric generating set can be improved by selecting the vibration of the x-axis and the vibration of the y-axis to analyze simultaneously, and false detection and missed detection are avoided. The differential processing for the x-axis vibration signal and the y-axis vibration signal is as follows:
preferably, in S2, the differentiation process performed on the original signal is a second order differentiation, and a first order differential signal and a second order differential signal are obtained.
According to the method, the original signal is subjected to differential processing before empirical mode decomposition, so that the influences of intermittent fault signals and noise interference on the original signal can be removed, and modal aliasing of an eigenmode function after empirical mode decomposition is reduced. In addition, the second-order differential has stronger capability of processing the signal and removing influencing factors compared with the first-order differential, and the calculation procedure of the third-order differential is simpler.
Preferably, in S3, the acquiring the set of eigenmode functions includes:
s31, carrying out empirical mode decomposition on the original signal to obtain a plurality of eigenmode functions;
s32, combining the original signal and the eigenmode functions in S31 to form an eigenmode function set of the original signal;
s33, carrying out empirical mode decomposition on the differential signal to obtain a plurality of eigenmode functions;
s34, combining the differential signal and the eigenmode functions in S33 into a set of eigenmode functions of the differential signal.
The empirical mode decomposition step comprises a screening process and a judging process, wherein the screening process comprises the following steps: solving an extreme point, and finding out all maximum values and minimum values of the signals; fitting an envelope curve, and obtaining two smooth wave crest and wave trough fitting curves, namely an upper envelope curve and a lower envelope curve of a signal, through a maximum value set and a minimum value set and through a cubic spline interpolation method; averaging the upper envelope line and the lower envelope line to obtain an average envelope line; the mean envelope is subtracted from the original signal to obtain an intermediate signal.
The determining process determines whether the intermediate signal meets a condition of an eigenmode function (IMF): in the whole data segment, the number of extreme points and the number of zero-crossing points must be equal or the difference cannot exceed one at most; at any time, the average value of the upper envelope formed by the local maximum point and the lower envelope formed by the local minimum point is zero, that is, the upper and lower envelopes are locally symmetrical with respect to the time axis. If the condition is satisfied, the intermediate signal is the first eigenmode function IMF1 of the original signal, and if the condition is not satisfied, the screening process is continued. After reaching IMF1, subtracting IMF1 from the original signal to obtain a new original signal, and obtaining IMF2 through a screening process, and so on to complete empirical mode decomposition.
Preferably, n eigenmode functions are collected in the eigenmode function set of the signal in S4; the energy ratio of each eigenmode function is the ratio of the energy of the eigenmode function to the total energy of the signal.
Preferably, the total energy of the signal in S4 is indicative of the energy intensity of the signal; the energy content of the eigenmode functions of the signal characterizes the frequency content of the signal.
In the invention, a signal corresponds to an eigenmode function set, the eigenmode function set comprises the signal and a finite number of eigenmode functions of the signal, then the total energy of the signal and the energy of each eigenmode function are calculated, the energy of the eigenmode function is divided by the total energy of the signal to form the energy ratio of each eigenmode function, and the total energy of the signal and the energy ratio of each eigenmode function are combined into a new function set to represent the energy characteristics of the signal.
Preferably, in S4, the energy ratios of each eigenmode function in the set of eigenmode functions are sequentially added, and when the energy ratio is added to the mth eigenmode function so that the total energy ratio is greater than ninety-five percent, m is taken as the number of effective energy ratios, and the sum of the energy ratios of the m eigenmode functions is taken as the effective ratio.
The invention selects the sum of the energy ratios from the first eigenmode function to the mth eigenmode function which is more than ninety-five percent as the effective ratio, m is the number of the effective energy ratios, and the first to the mth eigenmode functions are the effective eigenmode functions, thereby avoiding the excessive increase of the number of the eigenmode functions of the signal and reducing the calculation complexity on the premise of ensuring the accuracy of feature extraction.
Preferably, in the energy characteristic matrix of S5, the number of rows of the energy characteristic matrix is the differential order plus one; the number of columns of the energy characteristic matrix is the number of the maximum effective energy ratio in each row plus one, and zero is filled when the number of columns in other rows is insufficient.
In the energy characteristic matrix of the invention, the first column is the total energy of the original signal and the total energy of each differential signal, so the number of rows is the differential order plus one; each row of the energy signature matrix comprises the total energy of the signal and the energy fraction of each significant eigenmode function of the signal, so the number of columns depends on the number of the largest significant energy fractions in all rows plus one.
The invention has the following beneficial effects: the method has the advantages that high-frequency components of signals are enhanced through multiple differential processing of original signals, high-frequency extreme points which are possibly submerged in the original signals are recovered, and then the original signals and the differential signals are subjected to empirical mode decomposition, so that the problem of mode confusion in the analysis of vibration signals of the hydroelectric generating set is solved, and the adaptability and the accuracy of the empirical mode decomposition are improved; the total energy of the signal and the energy ratio of each effective eigenmode function of the signal are combined into an energy characteristic matrix, and the convenience and the accuracy of signal characteristic extraction are improved.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a flowchart of detailed operations in an embodiment of the present invention;
FIG. 3 is a diagram of 4 sets of original signals and eigenmode functions according to an embodiment of the present invention;
FIG. 4 is a diagram of 4 sets of first order differential signals and eigenmode functions according to an embodiment of the present invention;
FIG. 5 is a diagram of 4 sets of second order differential signals and eigenmode functions according to an embodiment of the present invention.
Detailed Description
The invention is further described with reference to the following figures and detailed description.
As shown in fig. 1, a method for identifying energy characteristics of a vibration signal of a hydroelectric generating set includes the following steps:
s1, collecting an original signal to obtain an original signal of the vibration of the hydroelectric generating set;
s2, calculating a differential signal, and carrying out differential processing on the acquired original signal to obtain a differential signal;
s3, performing empirical mode decomposition, namely performing empirical mode decomposition on the signals obtained in S1 and S2, and forming an eigenmode function set by using a plurality of obtained eigenmode functions;
s4, calculating energy ratio, and calculating the total energy of the signals in the eigenmode function set and the ratio of the energy of each eigenmode function in the total energy at S3;
and S5, combining the total signal energy and the eigenmode function energy proportion in S4 to form an energy feature matrix, observing the energy feature matrix, acquiring a signal energy distribution frequency band, and extracting effective signal features.
At S1, the raw signals of the hydro-electric machine set vibration include an x-axis vibration signal and a y-axis vibration signal of the hydro-electric machine set.
In S2, the differentiation process performed on the original signal is a second order differentiation, and a first order differential signal and a second order differential signal are obtained.
In S3, acquiring the set of eigenmode functions includes the steps of:
s31, carrying out empirical mode decomposition on the original signal to obtain a plurality of eigenmode functions;
s32, combining the original signal and the eigenmode functions in S31 to form an eigenmode function set of the original signal;
s33, carrying out empirical mode decomposition on the differential signal to obtain a plurality of eigenmode functions;
s34, combining the differential signal and the eigenmode functions in S33 into a set of eigenmode functions of the differential signal.
In S4, n eigenmode functions are concentrated in the eigenmode functions of the signal; the energy ratio of each eigenmode function is the ratio of the energy of the eigenmode function to the total energy of the signal. The total energy of the signal in S4 characterizes the energy intensity of the signal; the energy content of the eigenmode function of a signal characterizes the frequency content of the signal.
In S4, the energy ratios of each eigenmode function in the set of eigenmode functions are sequentially added, and when the energy ratio is added to the mth eigenmode function so that the total energy ratio is greater than ninety-five percent, m is taken as the number of effective energy ratios, and the sum of the energy ratios of the m eigenmode functions is the effective ratio.
In the energy feature matrix of S5, the number of rows of the energy feature matrix is the order of the differential plus one; the number of columns of the energy characteristic matrix is the number of the maximum effective energy ratio in each row plus one, and zero is filled when the number of columns in other rows is insufficient.
The method adopts Empirical Mode Decomposition (EMD) to process signals, and the EMD is a self-adaptive data processing or mining method and is very suitable for processing nonlinear and non-stationary time sequences, so that the method has great advantages in the aspects of processing vibration signals of the hydroelectric generating set and extracting characteristics. The key idea of empirical mode decomposition is to decompose a complex signal with irregular frequency into the form of the sum of eigenmode functions (IMF) of a limited single frequency and residual waves, wherein each decomposed eigenmode function comprises local characteristic signals of the original signal at different time scales.
The original signals of the invention adopt the x-axis vibration signal and the y-axis vibration signal of the hydroelectric generating set at the same time, because the working state and the condition of the hydroelectric generating set reflected by the vibration of the x-axis and the vibration of the y-axis are not completely the same, the comprehensive performance of state monitoring and fault diagnosis of the hydroelectric generating set can be improved by selecting the vibration of the x-axis and the vibration of the y-axis to analyze simultaneously, and false detection and missed detection are avoided. The differential processing for the x-axis vibration signal and the y-axis vibration signal is as follows:
in order to improve the modal aliasing phenomenon, the extreme point of the recovered high-frequency component can be used as an entry point. The differential operation has the characteristics of not changing the frequency composition of the original signal and being capable of linearly increasing the amplitude thereof according to the frequency of each component. Therefore, the amplitude of the high-frequency component is increased through differential operation, and then empirical mode decomposition is carried out, so that the recovery of the extreme point is more favorably realized. According to the method, the original signal is subjected to differential processing before empirical mode decomposition, so that the influences of intermittent fault signals and noise interference on the original signal can be eliminated, and modal aliasing of an eigenmode function after empirical mode decomposition is reduced. In addition, the second-order differential has stronger capability of processing the signal and removing influencing factors compared with the first-order differential, and the calculation procedure of the third-order differential is simpler. The differential signals of different orders are obtained from the original signal, and the differential signals and the original signal can be combined into a target signal set XT：
XT＝{x(t),x′(t),…,x(n)(t)}
The empirical mode decomposition step in the invention comprises a screening process and a judging process.
The screening process comprises the following steps: solving an extreme point, and finding out all maximum values and minimum values of the signals; fitting an envelope curve, and obtaining two smooth wave crest and wave trough fitting curves, namely an upper envelope curve and a lower envelope curve of a signal, through a maximum value set and a minimum value set and through a cubic spline interpolation method; averaging the upper envelope line and the lower envelope line to obtain an average envelope line; the mean envelope is subtracted from the original signal to obtain an intermediate signal.
The determining process determines whether the intermediate signal meets a condition of an eigenmode function (IMF): in the whole data segment, the number of extreme points and the number of zero-crossing points must be equal or the difference cannot exceed one at most; at any time, the average value of the upper envelope formed by the local maximum point and the lower envelope formed by the local minimum point is zero, that is, the upper and lower envelopes are locally symmetrical with respect to the time axis. If the condition is satisfied, the intermediate signal is the first eigenmode function IMF1 of the original signal, and if the condition is not satisfied, the screening process is continued. After obtaining the IMF1, subtracting the IMF1 from the original signal to obtain a new original signal, and then obtaining the IMF2 through a screening process, and so on to complete empirical mode decomposition. The expression after empirical mode decomposition is as follows:
wherein the original signal is x (t), and the residual signal is rn(t) eigenmode function is ci(t)。
For set XTPerforming empirical mode decomposition on each signal in the array, and combining eigenmode functions obtained by decomposing each order of signals to form an eigenmode function matrix MIMF：
In the matrix: first line IMF01-IMF0mThe second to n +1 th rows correspond to the first to n-th order differential signals, respectively, of the set of eigenmode functions of the original signal. The number of eigenmode functions obtained by decomposing each order of differential signals is generally different, so that the maximum number m of each order of decomposed IMF signals is selected as the length of the matrix when constructing the matrix, and 0 is supplemented in the matrix when the number of the rest decomposed IMFs is less than m.
In the invention, a signal corresponds to an eigenmode function set, the eigenmode function set comprises the signal and a finite number of eigenmode functions of the signal, then the total energy of the signal and the energy of each eigenmode function are calculated, the energy of the eigenmode function is divided by the total energy of the signal to form the energy ratio of each eigenmode function, and the total energy of the signal and the energy ratio of each eigenmode function are combined into a new function set to represent the energy characteristics of the signal. Energy E of each eigenmode functioniComprises the following steps:
the total energy ET of the signal and the energy ratio EP of the respective eigenmode functions can be expressed as:
the invention selects the sum of the energy ratios from the first eigenmode function to the mth eigenmode function which is more than ninety-five percent as the effective ratio, m is the number of the effective energy ratios, and the first to the mth eigenmode functions are the effective eigenmode functions, thereby avoiding the excessive increase of the number of the eigenmode functions of the signal and reducing the calculation complexity on the premise of ensuring the accuracy of feature extraction.
In the energy characteristic matrix of the invention, the first column is the total energy of the original signal and the total energy of each differential signal, so the number of rows is the differential order plus one; each row of the energy signature matrix includes the total energy of the signal and the energy of each significant eigenmode function of the signalThe fraction, and therefore the number of columns, depends on the maximum active energy fraction m in all rows plus one. Each row of the energy feature matrix is an energy distribution feature vector FV ═ ET, EP1,…,EPm]。
Fig. 2 is a flowchart of a method for identifying energy characteristics of a vibration signal of a hydroelectric generating set in this embodiment. In the figure, a represents the sum of energy ratios, and B represents the differential order, and theoretically, in the empirical mode decomposition operation process of the differential processing of the present invention, the higher the selected differential order is, the more energy characteristic information can be obtained. On the other hand, the differential operation with an excessively high order number in the actual operation causes a serious distortion of the signal, thereby destroying the original characteristics of the signal. Therefore, the highest order of differentiation B should be set according to the complexity of the target signal, in this case the second order differentiation is chosen. It should be noted that while the multiple differentiation amplifies the weak fault signal, it also amplifies the noise, thereby affecting the processing of the signal and the feature extraction, however, in this embodiment, the signal is decomposed by directly applying the empirical mode decomposition after the multiple differentiation operation, and the noise signal is usually higher in frequency, and can be effectively filtered by the empirical mode decomposition.
The vibration signal for setting the hydroelectric generating set is formed by linear coupling of the following formula:
in the formula: e.g. of the type1(t) represents a vibration signal at a fundamental frequency of 60 Hz; e.g. of the type2(t) represents a frequency-2 multiplied signal having a modulation characteristic; e.g. of the type3(t) and e4(t) represents 4-frequency multiplied and 6-frequency multiplied sinusoidal signals, respectively; e.g. of the type5(t) represents a noise signal.
Referring to a hydroelectric generating set fault set, when a mass unbalance fault occurs, a vibration signal contains a 60Hz base frequency signal, when an misalignment fault occurs, the vibration signal contains a 2-time frequency multiplication signal, and when a rubbing fault occurs, the vibration signal contains 4-time and 6-time frequency multiplication low-frequency component signals. By combining the above equations differently, the following four sets of test signals can be formed:
x1, x2, x3 and x4 are four sets of original signals (original) that need to be differentiated. In the formula: the value range of A is [0.3, 0.4], and represents the coefficient of 2 frequency multiplication components; the value range of B is [0.28, 0.32], which represents the coefficient of 4 frequency multiplication components; the value range of C is [0.09, 0.11], which represents the coefficient of 6 frequency multiplication components.
The four sets of original signals x1, x2, x3 and x4 are subjected to second order differentiation as shown in the following formula:
four sets of first order differential signals (1-ODSs) and four sets of second order differential signals (2-ODSs) are obtained in total.
Empirical mode decomposition is performed on four groups of original signals x1, x2, x3 and x4 to obtain a schematic diagram of the original signals and eigenmode functions as shown in fig. 3, wherein each column represents a group of original signals and its eigenmode function respectively. After performing empirical mode decomposition on the four groups of first-order differential signals, a schematic diagram of the first-order differential signals and eigenmode functions is obtained as shown in fig. 4, wherein each column in the diagram represents one group of first-order differential signals and its eigenmode function. The four sets of second order differential signals are subjected to empirical mode decomposition to obtain a second order differential signal and an eigenmode function diagram as shown in fig. 5, wherein each column in the diagram represents one set of two-pole differential signals and an eigenmode function thereof.
And calculating the total energy of the original signal of x1 and the energy ratio of each eigenmode function, adding the energy ratios of the first m eigenmode functions, when the sum of the energy ratios is more than ninety-five percent, selecting m as the number of effective energy ratios, and the sum of the energy ratios of the m eigenmode functions is the effective ratio, so that the energy distribution characteristic of the original signal of x1 can be obtained for some times. The calculation method of the energy distribution characteristics of the four sets of original signals, the four sets of first order differential signals, and the four sets of second order differential signals in the present embodiment is the same as the original signal energy distribution characteristics of x 1.
The energy distribution characteristics of the four sets of original signals, the four sets of first order differential signals and the four sets of second order differential signals are listed and combined as shown in the following table:
in the table, OS represents the original signal, 1-ODS represents the first order differential signal, 2-ODS represents the second order differential signal, ET represents the total energy of the signal, and EPi represents the energy ratio of the ith eigenmode function. By comparing the energy distribution characteristic matrixes of different signals, it can be found that the energy characteristic vectors corresponding to the eigenmode functions of the original signals of the four groups of signals are very close, and the energy characteristic vectors corresponding to the eigenmode functions of differential signals of each order have obvious difference, which is enough to reflect the different energy distribution characteristics of the original signals of the four groups of signals.
Energy distribution feature vector FV ═ ET, EP1,…,EPm]The energy characteristics of the signal are described in terms of macroscopic and microscopic aspects, respectively. In general, the number of eigenmode function components in the empirical mode decomposition results obtained from different signals is likely to be different, and in order to ensure that the energy distribution feature vectors have the same dimension, only a fixed number of eigenmode functions can be selected to calculate FV. The dominant energy of the signal is concentrated in the first few eigenmode function components, and the energy contained in the next-ranked eigenmode functions is usually lower. Thus, in the present example, for a single signal, the energy sum of the first m eigenmode function components, which account for more than ninety-five percent of the total energy, is selected, while for multi-signal analysis the largest selected number l of IMFs, which is all signals, is selected from the selected number m of eigenmode functions for each signal, the feature vector is FV ═ ET, EP1,…,EPl]. When the total number of eigenmode functions of the individual signals is less than l, the missing part of the vector is performed with 0And (4) supplementing.
By adopting the method, when the vibration signal of the hydroelectric generating set is analyzed by using the empirical mode decomposition method, mode confusion can be effectively solved, the energy proportion of each eigenmode function can be obviously distinguished by the energy distribution characteristic matrix, and the aim of accurately and quickly extracting the characteristic can be fulfilled.
The above embodiments are described in detail for the purpose of illustration and understanding, and no unnecessary limitations are to be understood therefrom, and any modifications, equivalents, and improvements made within the spirit and principle of the present invention should be included therein.
Claims (8)
1. A method for recognizing the energy characteristics of a vibration signal of a hydroelectric generating set is characterized by comprising the following steps of:
s1, collecting an original signal to obtain an original signal of the vibration of the hydroelectric generating set;
s2, calculating a differential signal, and carrying out differential processing on the acquired original signal to obtain a differential signal;
s3, performing empirical mode decomposition, namely performing empirical mode decomposition on the signals obtained in S1 and S2, and forming an eigenmode function set by using a plurality of obtained eigenmode functions;
s4, calculating energy ratio, and calculating the total energy of the signals in the eigenmode function set and the ratio of the energy of each eigenmode function in the total energy at S3;
and S5, combining the total signal energy and the eigenmode function energy proportion in S4 to form an energy feature matrix, observing the energy feature matrix, acquiring a signal energy distribution frequency band, and extracting effective signal features.
2. The method for identifying the energy characteristics of the vibration signal of the hydroelectric generating set according to claim 1, wherein in the step S1, the original signal of the vibration of the hydroelectric generating set comprises an x-axis vibration signal and a y-axis vibration signal of the hydroelectric generating set.
3. The method for identifying the energy characteristics of the vibration signals of the hydroelectric generating set according to claim 1 or 2, wherein in the step S2, the differentiation processing performed on the original signals is second order differentiation, which is as follows:
a first order differential signal and a second order differential signal are obtained, respectively.
4. The method for identifying the energy characteristics of the vibration signals of the hydroelectric generating set according to claim 1, wherein in the step S3, the obtaining of the eigenmode function set comprises the following steps:
s31, carrying out empirical mode decomposition on the original signal to obtain a plurality of eigenmode functions;
s32, combining the original signal and the eigenmode functions in S31 to form an eigenmode function set of the original signal;
s33, carrying out empirical mode decomposition on the differential signal to obtain a plurality of eigenmode functions;
s34, combining the differential signal and the eigenmode functions in S33 into a set of eigenmode functions of the differential signal.
5. The method for identifying the energy characteristics of the vibration signals of the hydroelectric generating set according to claim 1, wherein n eigenmode functions are concentrated in the eigenmode functions of the signals in S4; the energy ratio of each eigenmode function is the ratio of the energy of the eigenmode function to the total energy of the signal.
6. The method for identifying the energy characteristics of the vibration signals of the hydroelectric generating set according to claim 1 or 5, wherein the total energy of the signals in the step S4 represents the energy intensity of the signals; the energy ratio of the eigenmode functions of the signal characterizes the frequency composition of the signal; the combination of the total energy ET of a signal and the energy ratio EP of the eigenmode function forms the energy feature vector FV [ ET, EP ] of said signal1,…,EPm]。
7. The method for identifying the energy characteristics of the vibration signals of the hydroelectric generating set according to claim 1, wherein in step S4, the energy ratios of each eigenmode function in the eigenmode function set are sequentially added, and when the energy ratio is added to the mth eigenmode function so that the total energy ratio is greater than ninety-five percent, m is taken as the number of the effective energy ratios, and the sum of the energy ratios of the m eigenmode functions is the effective ratio.
8. The method for identifying the energy characteristics of the vibration signals of the hydroelectric generating set according to claim 7, wherein in the energy characteristic matrix of S5, the number of rows of the energy characteristic matrix is the differential order plus one; the number of columns of the energy characteristic matrix is the number of the maximum effective energy ratio in each row plus one, and zero is filled when the number of columns in other rows is insufficient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110789816.7A CN113688669A (en) | 2021-07-13 | 2021-07-13 | Hydroelectric generating set vibration signal energy characteristic identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110789816.7A CN113688669A (en) | 2021-07-13 | 2021-07-13 | Hydroelectric generating set vibration signal energy characteristic identification method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113688669A true CN113688669A (en) | 2021-11-23 |
Family
ID=78577399
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110789816.7A Pending CN113688669A (en) | 2021-07-13 | 2021-07-13 | Hydroelectric generating set vibration signal energy characteristic identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113688669A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115062653A (en) * | 2022-06-09 | 2022-09-16 | 山东龙源电力工程有限公司 | Analysis maintenance system based on steam turbine of thermal power plant |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103049798A (en) * | 2012-12-05 | 2013-04-17 | 浙江大学城市学院 | Short-period electric generation power forecasting method applied to photovoltaic electric generation system |
| CN106895906A (en) * | 2017-03-23 | 2017-06-27 | 西安理工大学 | A kind of feature extracting method of vibration of hydrogenerator set failure |
| CN107423671A (en) * | 2017-05-05 | 2017-12-01 | 山西财经大学 | A kind of noise-reduction method of the empirical mode decomposition high-frequency data based on component vector |
| CN109472288A (en) * | 2018-10-08 | 2019-03-15 | 华中科技大学 | A kind of pump-storage generator vibration hybrid feature extraction and classification method |
-
2021
- 2021-07-13 CN CN202110789816.7A patent/CN113688669A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103049798A (en) * | 2012-12-05 | 2013-04-17 | 浙江大学城市学院 | Short-period electric generation power forecasting method applied to photovoltaic electric generation system |
| CN106895906A (en) * | 2017-03-23 | 2017-06-27 | 西安理工大学 | A kind of feature extracting method of vibration of hydrogenerator set failure |
| CN107423671A (en) * | 2017-05-05 | 2017-12-01 | 山西财经大学 | A kind of noise-reduction method of the empirical mode decomposition high-frequency data based on component vector |
| CN109472288A (en) * | 2018-10-08 | 2019-03-15 | 华中科技大学 | A kind of pump-storage generator vibration hybrid feature extraction and classification method |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115062653A (en) * | 2022-06-09 | 2022-09-16 | 山东龙源电力工程有限公司 | Analysis maintenance system based on steam turbine of thermal power plant |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sun et al. | Fault diagnosis of rolling bearing based on empirical mode decomposition and improved manhattan distance in symmetrized dot pattern image | |
| Dong et al. | Non-iterative denoising algorithm for mechanical vibration signal using spectral graph wavelet transform and detrended fluctuation analysis | |
| CN112465264A (en) | Food safety risk grade prediction method and device and electronic equipment | |
| CN111079710A (en) | Multilayer noise reduction method based on improved CEEMD rolling bearing signal | |
| CN113297987B (en) | Variational modal decomposition signal noise reduction method based on dual-objective function optimization | |
| CN113688669A (en) | Hydroelectric generating set vibration signal energy characteristic identification method | |
| CN103119848B (en) | Data processing method and device | |
| CN113269169A (en) | Bearing fault detection method and device | |
| CN113390631A (en) | Fault diagnosis method for gearbox of diesel engine | |
| CN113361205A (en) | Bearing vibration signal fault diagnosis optimization design method based on genetic algorithm and CEEMD | |
| CN114358093A (en) | Method and equipment for detecting partial discharge in power equipment | |
| CN108694355B (en) | Signal processing method and device | |
| CN112098093A (en) | Bearing fault feature identification method and system | |
| CN116306767A (en) | Gear fault diagnosis method for improving VMD of whale | |
| CN115293188A (en) | Fault diagnosis method and device for reciprocating mechanical equipment | |
| CN112817056B (en) | Magnetotelluric signal denoising method and system | |
| CN113627539A (en) | Rolling bearing composite fault diagnosis method and system, storage medium and computing equipment | |
| CN103020907B (en) | Based on the DSPI striped filtering system of two-dimensional ensemble empirical mode decomposition | |
| CN115200797B (en) | Leakage detection system for zero leakage valve | |
| CN102170519A (en) | High-efficient image denoising algorithm and hardware realization apparatus of algorithm | |
| CN112729531B (en) | Fault studying and judging method and system for distribution transformer equipment | |
| CN111803060B (en) | Electrocardio artifact signal removing method and device | |
| CN113229826A (en) | QRS wave detection method and device and electronic equipment | |
| CN111345806B (en) | Cardiac cycle detection method and device | |
| CN116671867B (en) | Sleep quality evaluation method and system for underwater operators |
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 |