CN113654782B - Mechanical equipment fault diagnosis method and system - Google Patents
Mechanical equipment fault diagnosis method and system Download PDFInfo
- Publication number
- CN113654782B CN113654782B CN202110941703.4A CN202110941703A CN113654782B CN 113654782 B CN113654782 B CN 113654782B CN 202110941703 A CN202110941703 A CN 202110941703A CN 113654782 B CN113654782 B CN 113654782B
- Authority
- CN
- China
- Prior art keywords
- noise amplitude
- noise
- optimal
- modal
- empirical mode
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
-
- 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
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/003—Machine valves
-
- 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
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Abstract
The invention discloses a method and a system for diagnosing faults of mechanical equipment, wherein the method comprises the steps of calculating modal chaos indexes corresponding to different noise amplitudes, and selecting the noise amplitude corresponding to the minimum modal chaos index as the optimal noise amplitude; calculating an optimal ensemble averaging number based on the expected decomposition error and the optimal noise amplitude value; decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after parameter configuration to obtain each mode function; the ensemble empirical mode decomposition algorithm after the parameters are configured is an ensemble empirical mode decomposition algorithm obtained by setting the amplitude of the added noise as the optimal noise amplitude and setting the ensemble average times as the optimal ensemble average times on the basis of the ensemble empirical mode decomposition algorithm; and carrying out envelope demodulation spectrum analysis on each modal function to extract the characteristic frequency of the target mechanical equipment, and further carrying out fault diagnosis on the target mechanical equipment. The invention can improve the fault diagnosis accuracy of mechanical equipment.
Description
Technical Field
The invention relates to the field of fault diagnosis of mechanical equipment, in particular to a fault diagnosis method and a fault diagnosis system of the mechanical equipment based on a general empirical mode decomposition algorithm of self-adaptive configuration optimal parameters.
Background
With the continuous development of modern industry, science and technology and economy, the industry becomes an important standard for measuring the comprehensive strength of a country. Industry has not developed stable machines, and a large number of machines have good stability in the early stages of operation, but their performance degrades over time or even fails. And the performance and safety of production are seriously influenced by the failed mechanical equipment. Therefore, the demands for reliability, safety and reduced manufacturing cost in the production process greatly promote the development of fault diagnosis technology for mechanical equipment.
Due to the diversification of the working environment of the mechanical equipment, the complexity of mechanical faults is different from the prior art, the mechanical faults often have nonlinear and non-stable characteristics, and the existing mechanical equipment fault diagnosis method is not suitable for use. Empirical mode decomposition techniques have been widely used to process nonlinear, non-stationary signals, so in the field of fault diagnosis of mechanical equipment, empirical mode decomposition techniques can be employed.
The empirical mode decomposition method is an adaptive method based on signal decomposition, and comprises the following decomposition steps: fitting local maximum (minimum) points of the signal by using a cubic spline function to obtain upper (lower) envelopes, and calculating the mean value of the upper and lower envelopes; subtracting the envelope mean value from the signal, and iterating the steps until the screened function meets the condition of the intrinsic mode function; and subtracting the selected intrinsic mode function from the signal to perform the next round of screening, and repeating the screening until the residual signal is a constant value or a monotone function, and ending the decomposition. The empirical mode decomposition EMD has problems of decomposition limitation, mode aliasing, and the like.
Aiming at the problems, Wu and the like propose an EEMD (ensemble empirical mode decomposition) method, which respectively adds N groups of white noises with the amplitude of epsilon into a signal for decomposition, wherein the addition of the white noises can change the extreme point distribution of the signal, fill high-frequency intermittence in the signal, and effectively reduce the problems of modal aliasing and the like. However, the amplitude epsilon and the group number N of the white noise need to be selected artificially, and for different signals to be decomposed, the corresponding optimal noise amplitudes may also be different, and excessively large or small noise amplitudes may cause modal aliasing to different degrees.
Obviously, the existing ensemble empirical mode decomposition technology still has the problems of decomposition limitation, mode aliasing and the like to a certain extent, and if the ensemble empirical mode decomposition technology is applied to the field of fault diagnosis of mechanical equipment, the defect of low fault diagnosis accuracy inevitably exists.
Disclosure of Invention
The invention aims to provide a mechanical equipment fault diagnosis method and system based on a general empirical mode decomposition algorithm for adaptively configuring optimal parameters.
In order to achieve the purpose, the invention provides the following scheme:
a method of fault diagnosis for a mechanical device, comprising:
determining a vibration signal of the target mechanical equipment;
calculating modal chaos indexes corresponding to different noise amplitudes, and selecting the noise amplitude corresponding to the smallest modal chaos index as the optimal noise amplitude; the modal chaos index is obtained by calculating a decomposition result of a general empirical mode decomposition algorithm;
determining an expected decomposition error, and calculating an optimal ensemble averaging number based on the expected decomposition error and the optimal noise amplitude;
decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after parameter configuration to obtain each mode function; the ensemble empirical mode decomposition algorithm after the parameters are configured is an ensemble empirical mode decomposition algorithm obtained by setting the amplitude of the added noise as the optimal noise amplitude and setting the ensemble average times as the optimal ensemble average times on the basis of the ensemble empirical mode decomposition algorithm;
and carrying out envelope demodulation spectrum analysis on each modal function obtained by decomposition to extract the characteristic frequency of the target mechanical equipment, and carrying out fault diagnosis on the target mechanical equipment based on the characteristic frequency of the target mechanical equipment.
Optionally, the calculating of the modal chaos indexes corresponding to different noise amplitudes and selecting the noise amplitude corresponding to the smallest modal chaos index as the optimal noise amplitude specifically include:
determining a search range and a search step length of a noise amplitude coefficient;
generating a noise amplitude set according to the search range and the search step length; the ith noise amplitude in the noise amplitude set is epsilon i ＝α i ×std[x(t)]X (t) is the vibration signal of the target mechanical equipment, std [ ·]As standard deviation operator, α i Is the ith noise amplitude coefficient;
calculating modal chaos indexes corresponding to different noise amplitudes in the noise amplitude set;
and selecting the noise amplitude corresponding to the minimum modal chaos index as the optimal noise amplitude.
Optionally, the calculating modal chaos indexes corresponding to different noise amplitudes in the noise amplitude set specifically includes:
sequentially configuring each noise amplitude in the noise amplitude set to a general empirical mode decomposition algorithm to obtain a plurality of general empirical mode decomposition algorithms after noise amplitudes are configured; the number of the overall empirical mode decomposition algorithms after the noise amplitude is configured is the same as that of the noise amplitudes, and different overall empirical mode decomposition algorithms after the noise amplitudes are configured with different noise amplitudes;
decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after the configuration of the calibration noise amplitude to obtain a calibration fluctuation index set corresponding to the calibration noise amplitude; the calibration fluctuation index set comprises a segmented standard deviation fluctuation index of each mode; the calibrated noise amplitude is any noise amplitude in the noise amplitude set;
and summing the section standard deviation fluctuation indexes of each mode in the calibration fluctuation index set to obtain a mode chaos index corresponding to each noise amplitude.
Optionally, the determining an expected decomposition error, and calculating an optimal ensemble averaging time based on the expected decomposition error and the optimal noise amplitude specifically includes:
determining 1% of a standard deviation of the vibration signal as an expected decomposition error;
based on the formula
Calculating the optimal overall average times;
wherein N is opt Is the optimal ensemble average degree, ε opt Is the optimum noise amplitudeThe value, e, is the expected decomposition error.
A mechanical device fault diagnostic system, comprising:
the vibration signal determination module is used for determining a vibration signal of the target mechanical equipment;
the optimal noise amplitude determining module is used for calculating modal chaos indexes corresponding to different noise amplitudes and selecting the noise amplitude corresponding to the minimum modal chaos index as the optimal noise amplitude; the modal chaos index is obtained by calculating a decomposition result of a general empirical mode decomposition algorithm;
the optimal ensemble average times calculation module is used for determining an expected decomposition error and calculating the optimal ensemble average times based on the expected decomposition error and the optimal noise amplitude;
the modal function obtaining module is used for decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after parameters are configured to obtain each modal function; the ensemble empirical mode decomposition algorithm after the parameters are configured is an ensemble empirical mode decomposition algorithm obtained by setting the amplitude of the added noise as the optimal noise amplitude and setting the ensemble average times as the optimal ensemble average times on the basis of the ensemble empirical mode decomposition algorithm;
and the fault diagnosis module is used for carrying out envelope demodulation spectrum analysis on each modal function obtained by decomposition so as to extract the characteristic frequency of the target mechanical equipment and carrying out fault diagnosis on the target mechanical equipment based on the characteristic frequency of the target mechanical equipment.
Optionally, the optimal noise amplitude determining module specifically includes:
a search range and search step determining unit for determining a search range and a search step of the noise amplitude coefficient;
the noise amplitude set generating unit is used for generating a noise amplitude set according to the search range and the search step length; the ith noise amplitude in the noise amplitude set is epsilon i ＝α i ×std[x(t)]X (t) is the vibration signal of the target mechanical equipment, std [. cndot]As standard deviation operator, α i Is the ith noise amplitude coefficient;
the modal chaos index calculation unit is used for calculating modal chaos indexes corresponding to different noise amplitudes in the noise amplitude set;
and the optimal noise amplitude determining unit is used for selecting the noise amplitude corresponding to the minimum modal chaos index as the optimal noise amplitude.
Optionally, the modal chaos index calculating unit specifically includes:
a configuration subunit, configured to sequentially configure each noise amplitude in the noise amplitude set to a ensemble empirical mode decomposition algorithm, so as to obtain a plurality of ensemble empirical mode decomposition algorithms configured with the noise amplitudes; the number of the overall empirical mode decomposition algorithms after the noise amplitude is configured is the same as that of the noise amplitudes, and different overall empirical mode decomposition algorithms after the noise amplitudes are configured with different noise amplitudes;
the calibration fluctuation index set determining subunit is used for decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after the configuration of the calibration noise amplitude to obtain a calibration fluctuation index set corresponding to the calibration noise amplitude; the calibration fluctuation index set comprises a segmented standard deviation fluctuation index of each mode; the calibrated noise amplitude is any noise amplitude in the noise amplitude set;
and the modal chaos index calculating subunit is used for summing the segmental standard deviation fluctuation indexes of each modal in the calibration fluctuation index set to obtain a modal chaos index corresponding to each noise amplitude.
Optionally, the optimal ensemble average number calculating module specifically includes:
an expected decomposition error determination unit for determining 1% of a standard deviation of the vibration signal as an expected decomposition error;
an optimal ensemble average number calculation unit for calculating the number of times of the transmission based on the formula
Calculating an optimumThe number of ensemble averages;
wherein N is opt Is the optimum global mean number of times, ε opt Is the optimum noise amplitude and e is the desired decomposition error.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
according to the method, the modal chaos indexes corresponding to different noise amplitudes are calculated, the noise amplitude corresponding to the smallest modal chaos index is selected as the optimal noise amplitude, the optimal overall average times is calculated based on the expected decomposition error and the optimal noise amplitude, self-adaptive selection of the noise-added amplitude parameter and the overall average times parameter in the overall empirical mode decomposition algorithm is achieved, and the purpose of improving the fault diagnosis accuracy of the mechanical equipment is achieved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
FIG. 1 is a schematic flow chart of a method for diagnosing faults of mechanical equipment according to the present invention;
FIG. 2 is a schematic structural diagram of a fault diagnosis system for mechanical equipment according to the present invention;
FIG. 3 is a general flow chart of a fault diagnosis method for mechanical equipment based on a general empirical mode decomposition algorithm for adaptively configuring optimal parameters according to the present invention;
FIG. 4 is a diagram illustrating simulation signal processing results according to the present invention; FIG. 4(a) is a diagram showing the result of processing a simulation signal by a ensemble empirical mode decomposition algorithm configured with a noise amplitude of 0.2 and an ensemble average number of times of 100; FIG. 4(b) is a diagram illustrating the adaptive selection of noise amplitude results; FIG. 4(c) is a diagram showing the configuration of the optimum noise amplitude ε opt And the number of ensemble averages N opt The result schematic diagram of the simulation signal processed by the general empirical mode decomposition algorithm is shown;
FIG. 5 is a diagram illustrating the actual vibration signal processing results of the present invention; FIG. 5(a) is a diagram illustrating an actual vibration signal; FIG. 5(b) is a diagram illustrating the result of processing the actual vibration signal by the ensemble empirical mode decomposition algorithm configured with a noise amplitude of 0.2 and an ensemble averaging times of 100; FIG. 5(c) is a diagram showing the configuration of the optimum noise amplitude ε opt And the number of ensemble averages N opt The general empirical mode decomposition algorithm processes an envelope spectrum schematic diagram corresponding to a result of an actual vibration signal.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
The method based on the relative root mean square error index comprises the following steps:
the method introduces relative root mean square error RRMSE, configures noises with different amplitudes to general empirical mode decomposition, then decomposes original signals respectively, calculates RRMSE corresponding to each noise according to decomposition results, selects the noise corresponding to the largest RRMSE as the optimal selection, and the calculation formula of RRMSE is as follows:
wherein x (t) is the original signal, c max (t) is the IMF component with the largest correlation coefficient with x (t), and N is the number of sampling points of x (t).
The advantages are that: for a vibration signal having a single principal component, the method can extract the principal component from noise and low correlation components, and is advantageous for processing a vibration signal having a single principal component.
The disadvantages are as follows: the mechanical equipment working environment is diversified, mechanical fault complexity is high, mechanical vibration signals often contain multi-component signals, RRMSE can not effectively judge the performance of ensemble empirical mode decomposition from decomposition results of the multi-component signals, and the selected noise amplitude value can cause the ensemble empirical mode decomposition to have a mode aliasing problem and influence fault feature extraction.
The method based on the distribution characteristics of signal poles comprises the following steps:
the method traverses noises with different amplitudes, adds each noise into an original signal to form a mixed signal, selects the noise amplitude corresponding to the minimum std _ max and std _ min by calculating the maximum (small) value point distribution in the mixed signal, namely std _ max (std _ min), and averages the two noise amplitudes to obtain the optimal noise amplitude, wherein the calculation formulas of std _ max and std _ min are as follows:
wherein N is 1 、N 2 Respectively, the maximum value point extr _ max (x) 1 ,y 1 ) And a minimum value point extr _ min (x) 2 ,y 2 ) Number of (1), u 1 、u 2 Respectively the mean value of the abscissa interval points of adjacent maximum values and adjacent minimum value points, u 1 '、u' 2 The average values of the difference values of the vertical coordinates of the adjacent maximum values and the adjacent minimum values are respectively.
The advantages are that: the method can effectively evaluate the influence of noise with different amplitudes on the extreme point of the original signal, and selects the noise amplitude which enables the extreme point of the mixed signal to be distributed more uniformly as the optimal noise amplitude, thereby avoiding the modal aliasing phenomenon caused by high-frequency intermittence in the original signal and enabling the modal function obtained by decomposition to have more physical significance.
The disadvantages are as follows: if all the noises cannot make the original signal extreme points uniformly distributed, the noise selected by the method may not be optimal, which causes a modal aliasing problem in the decomposition result, thereby affecting the extraction accuracy of the mechanical fault features.
In order to improve the limitation existing when the optimal noise amplitude is selected by the two methods, the invention provides a mechanical equipment fault diagnosis method and system based on a general empirical mode decomposition algorithm for adaptively configuring optimal parameters, and aims to overcome the problem of mode aliasing caused by adding inappropriate noise, ensure that an intrinsic mode function obtained by decomposition has actual physical significance and realize effective diagnosis of mechanical equipment faults.
Another objective of the present invention is to provide a Ensemble Empirical Mode Decomposition (EEMD) method for adaptively configuring optimal parameters, which adaptively configures optimal noise amplitude and ensemble averaging times for different signals to be decomposed without human intervention.
Example one
Referring to fig. 1, a method for diagnosing a fault of a mechanical device according to this embodiment includes:
step 101: and determining a vibration signal of the target mechanical equipment.
Step 102: calculating modal chaos indexes corresponding to different noise amplitudes, and selecting the noise amplitude corresponding to the smallest modal chaos index as the optimal noise amplitude; the modal chaos index is obtained by calculating a decomposition result of a general empirical mode decomposition algorithm.
Step 103: an expected decomposition error is determined and an optimal ensemble averaging number is calculated based on the expected decomposition error and the optimal noise amplitude.
Step 104: decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after parameter configuration to obtain each mode function; the ensemble empirical mode decomposition algorithm after the parameters are configured is an ensemble empirical mode decomposition algorithm obtained by setting the amplitude of the added noise as the optimal noise amplitude and setting the ensemble average times as the optimal ensemble average times on the basis of the ensemble empirical mode decomposition algorithm.
Step 105: and carrying out envelope demodulation spectrum analysis on each modal function obtained by decomposition so as to extract the characteristic frequency of the target mechanical equipment, and carrying out fault diagnosis on the target mechanical equipment based on the characteristic frequency of the target mechanical equipment.
Wherein, step 102 specifically comprises:
and determining the search range and the search step size of the noise amplitude coefficient.
Generating a noise amplitude set according to the search range and the search step length; the ith noise amplitude in the noise amplitude set is epsilon i ＝α i ×std[x(t)]X (t) is the vibration signal of the target mechanical equipment, std [. cndot]As standard deviation operator, α i Is the ith noise amplitude coefficient.
And calculating modal chaos indexes corresponding to different noise amplitudes in the noise amplitude set.
And selecting the noise amplitude corresponding to the minimum modal chaos index as the optimal noise amplitude.
Further, the calculating modal chaos indicators corresponding to different noise amplitudes in the noise amplitude set specifically includes:
sequentially configuring each noise amplitude in the noise amplitude set to a general empirical mode decomposition algorithm to obtain a plurality of general empirical mode decomposition algorithms after noise amplitudes are configured; the number of the overall empirical mode decomposition algorithms after the configuration of the noise amplitude is the same as that of the noise amplitude, and different overall empirical mode decomposition algorithms after the configuration of the noise amplitude are configured with different noise amplitudes.
Decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after the configuration of the calibration noise amplitude to obtain a calibration fluctuation index set corresponding to the calibration noise amplitude; the calibration fluctuation index set comprises a segmented standard deviation fluctuation index of each mode; the calibrated noise amplitude is any noise amplitude in the noise amplitude set.
And summing the section standard deviation fluctuation indexes of each mode in the calibration fluctuation index set to obtain a mode chaos index corresponding to each noise amplitude.
determining 1% of a standard deviation of the vibration signal as an expected decomposition error.
Based on the formula
And calculating the optimal ensemble average times.
Wherein N is opt Is the optimum global mean number of times, ε opt Is the optimum noise amplitude and e is the desired decomposition error.
The core innovation points of the embodiment of the invention are as follows:
first, a set of noise amplitudes is generated
The method of (1).
Secondly, the provided segment standard deviation fluctuation index and the modal chaos index are key indexes for evaluating the decomposition result, and the two key indexes ensure that the selected noise amplitude is optimal.
Thirdly, an expected decomposition error is set, and the determined optimal added noise amplitude is combined, and the overall average times are selected in a self-adaptive mode through a probability statistical rule, so that the minimum decomposition error and the minimum calculation cost are guaranteed.
In view of the above core innovation points, compared with the prior art, the method and the device provided by the invention overcome the problems of overlarge decomposition error and calculation cost caused by artificial selection of the times of the set, also overcome the influence on different signals to be decomposed when the noise amplitude is artificially selected, and furthest inhibit the mode aliasing problem of EEMD, thereby ensuring that the intrinsic mode function obtained by decomposition has actual physical significance and realizing effective diagnosis of mechanical equipment faults.
Example two
Referring to fig. 2, a mechanical equipment fault diagnosis system includes:
a vibration signal determination module 201, configured to determine a vibration signal of the target mechanical device;
the optimal noise amplitude determining module 202 is configured to calculate modal chaos indexes corresponding to different noise amplitudes, and select a noise amplitude corresponding to a smallest modal chaos index as an optimal noise amplitude; the modal chaos index is obtained by calculating a decomposition result of a general empirical mode decomposition algorithm;
an optimal ensemble averaging times calculation module 203, configured to determine an expected decomposition error, and calculate an optimal ensemble averaging times based on the expected decomposition error and the optimal noise amplitude;
a mode function obtaining module 204, configured to decompose the vibration signal by using a general empirical mode decomposition algorithm after parameter configuration to obtain each mode function; the ensemble empirical mode decomposition algorithm after the parameters are configured is an ensemble empirical mode decomposition algorithm obtained by setting the amplitude of the added noise as the optimal noise amplitude and setting the ensemble average times as the optimal ensemble average times on the basis of the ensemble empirical mode decomposition algorithm;
the fault diagnosis module 205 is configured to perform envelope demodulation spectrum analysis on each of the decomposed modal functions to extract a characteristic frequency of a target mechanical device, and perform fault diagnosis on the target mechanical device based on the characteristic frequency of the target mechanical device.
The optimal noise amplitude determining module 202 specifically includes:
and the searching range and searching step determining unit is used for determining the searching range and the searching step of the noise amplitude coefficient.
The noise amplitude set generating unit is used for generating a noise amplitude set according to the search range and the search step length; the ith noise amplitude in the noise amplitude set is epsilon i ＝α i ×std[x(t)]X (t) is the vibration signal of the target mechanical equipment, std [. cndot]As standard deviation operator, α i Is the ith noise amplitude coefficient.
And the modal chaos index calculation unit is used for calculating modal chaos indexes corresponding to different noise amplitudes in the noise amplitude set.
And the optimal noise amplitude determining unit is used for selecting the noise amplitude corresponding to the minimum modal chaos index as the optimal noise amplitude.
Further, the modal chaos index calculating unit specifically includes:
a configuration subunit, configured to sequentially configure each noise amplitude in the noise amplitude set to a ensemble empirical mode decomposition algorithm, so as to obtain a plurality of ensemble empirical mode decomposition algorithms configured with the noise amplitudes; the number of the overall empirical mode decomposition algorithms after the configuration of the noise amplitude is the same as that of the noise amplitude, and different overall empirical mode decomposition algorithms after the configuration of the noise amplitude are configured with different noise amplitudes.
The calibration fluctuation index set determining subunit is used for decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after the configuration of the calibration noise amplitude to obtain a calibration fluctuation index set corresponding to the calibration noise amplitude; the calibration fluctuation index set comprises a segmented standard deviation fluctuation index of each mode; the calibrated noise amplitude is any noise amplitude in the noise amplitude set.
And the modal chaos index calculating subunit is used for summing the segmental standard deviation fluctuation indexes of each modal in the calibration fluctuation index set to obtain a modal chaos index corresponding to each noise amplitude.
The optimal ensemble average times calculation module 203 specifically includes:
an expected decomposition error determination unit for determining 1% of a standard deviation of the vibration signal as an expected decomposition error.
An optimal ensemble average number calculation unit for calculating the number of times of the transmission based on the formula
And calculating the optimal ensemble average times.
Wherein N is opt Is the optimum global mean number of times, ε opt Is the optimum noise amplitude and e is the desired decomposition error.
EXAMPLE III
The embodiment discloses a mechanical equipment fault diagnosis method of an EEMD algorithm for adaptively configuring optimal parameters. By utilizing the method, the optimal parameters of the ensemble empirical mode decomposition algorithm aiming at different signals to be decomposed can be determined in a self-adaptive manner, the mode aliasing problem generated when improper noise is added is solved, and the problems of overlarge decomposition errors and calculation cost generated by different ensemble average times are solved, so that the intrinsic mode function obtained by decomposition has actual physical significance, and the effective diagnosis of the mechanical equipment fault is realized.
Referring to fig. 3, the method for diagnosing a fault of a mechanical device based on a ensemble empirical mode decomposition algorithm for adaptively configuring optimal parameters according to the present embodiment includes:
step 1: the signal to be decomposed x (t) is determined.
Step 2: and setting the search range and the step length of the noise amplitude coefficient.
Wherein the ith noise amplitude is epsilon i ＝α i ×std[x(t)]Coefficient of i The search range of (a) is 0.02 to 0.5, the step length value is 0.02, and the coefficient alpha i Is determined from the lower limit to the upper limit of the search range, and takes a value every other step, and takes 0.01 separately.
Generating a noise amplitude set according to the search range and the step length
Comprises the following steps:
{ε 1 ＝0.01std[x(t)],ε 2 ＝0.02std[x(t)],ε 3 ＝0.04std[x(t)],...,ε 26 ＝0.5std[x(t)]}。
and step 3: calculating modal chaos indexes obtained by overall empirical mode decomposition algorithm decomposition corresponding to different noise amplitudes, and selecting the noise amplitude corresponding to the minimum modal chaos index as the optimal noise amplitude, wherein the steps are as follows:
step 3-1, sequentially collecting the noise amplitude values
Is configured to the ensemble empirical mode decomposition algorithm to decompose the signal to be decomposed x (t) 26 times.
Step 3-2, calculating the ith noise amplitude epsilon i And correspondingly decomposing the data by a general empirical mode decomposition algorithm to obtain a sectional standard deviation fluctuation index of each mode, and summing the sectional standard deviation fluctuation indexes of each mode to obtain a mode chaos index corresponding to the ith noise amplitude. The calculation formula is as follows:
among them, SSDVI k Is the fractional standard deviation fluctuation index of the kth mode, std [. cndot.]Is a standard deviation operator, and the value of S is 10, which means that the modality is divided into 10 sections.
Is the nth signal in the kth mode, n is 1,2 i The model chaos index corresponding to the ith noise amplitude value is m, and the model order is obtained by decomposition.
3-3, selecting the noise amplitude epsilon corresponding to the minimum modal chaos index i As the optimum noise amplitude epsilon opt Namely:
and 4, step 4: and setting an expected decomposition error, and calculating the overall average times through a statistical rule on the basis of the expected decomposition error and the optimal noise amplitude.
Setting the expected decomposition error to be 1% of the standard deviation of the signal x (t) to be decomposed, and calculating the formula as follows:
e＝0.01×std[x(t)]。
based on the expected decomposition error and the optimal noise amplitude, the overall average times are selected in a self-adaptive mode through a probability statistical rule, and the calculation formula is as follows:
wherein N is opt Is the optimal ensemble averaging times.
And 5: using configured optimal parameters (optimal noise amplitude ε) opt Corresponding noise and ensemble averaging times N opt ) The ensemble empirical mode decomposition algorithm of (2) decomposes the signal to be decomposed.
Step 6: and carrying out envelope demodulation spectrum analysis on each modal function obtained by decomposition so as to extract fault characteristic frequency.
Example four
In order to verify the effectiveness of the technical scheme provided by the invention, a group of complex simulation signals is firstly analyzed: x (t) ═ sin (2 π × 2 × t) + cos (2 π × 5 × t) + c3+0.2 × w (t).
Where c3 is a modulated signal: c3 ═ 1+ sin (2 π × 0.5 × t)) × cos (2 π × 10 × t), and w (t) is a white gaussian noise signal
Data length 900, sampling frequency 100 Hz.
Decomposing the simulation signal by using a general empirical mode decomposition algorithm, and configuring parameters: the added noise amplitude is 0.2, the ensemble average degree is 100, and the decomposition result is shown in fig. 4 (a). The overall empirical mode decomposition algorithm can be carried out according to the prior art of 1-41 in Wu Z, Huang N E, Ensemble empirical mode decomposition, a noise-associated data analysis method [ J ]. Advances in Adaptive data analysis,2009,1 (1).
Setting the search range of the noise amplitude coefficient to be 0.02-0.5, independently taking 0.01, taking the step length value to be 0.02, and generating a noise amplitude set
Comprises the following steps:
{ε 1 ＝0.01std[x(t)],ε 2 ＝0.02std[x(t)],ε 3 ＝0.04std[x(t)],...,ε 26 ＝0.5std[x(t)]}。
FIG. 4(b) shows a set of noise amplitudes
Selecting the noise amplitude corresponding to the minimum modal chaos index as the added optimal noise amplitude, namely epsilon opt ＝ε 17 ＝0.32std[x(t)]The ensemble average number is calculated again as 0.4330:
FIG. 4(c) shows the EEMD decomposition results with the above two parameters. As can be seen from the figure, the EEMD configured with the adaptive parameters has better performance of extracting sine, cosine and modulation signals, and frequency aliasing hardly exists among all modes; and the decomposition result configured with the conventional parameter EEMD has a serious modal aliasing phenomenon, so the technical scheme provided by the invention can effectively inhibit the modal aliasing problem of the EEMD.
The embodiment of the invention adopts the mechanical vibration signal of the high-pressure diaphragm pump station in the wear breakdown state of the one-way valve as experimental data and carries out analysis and processing. Wherein the high-pressure diaphragm pump is a TZPM series three-cylinder crankshaft driving piston type diaphragm pump, and the highest working pressure is 24.44 MPa. Fig. 5(a) shows mechanical vibration signals collected in a wear breakdown state of a one-way valve of a high-pressure diaphragm pump station, wherein the sampling frequency is 2560Hz, the number of sampling points is 10240, the frequency of the one-way valve in normal operation is 0.5Hz to 0.517Hz, and the fundamental frequency of the one-way valve in operation is further calculated to be 1Hz to 1.034 Hz.
Decomposing vibration information by using a general empirical mode decomposition algorithm, and configuring parameters: the added noise amplitude is 0.2, the ensemble averaging times is 100, and the decomposition result is shown in fig. 5 (b).
The illustrated EEMD algorithm using adaptive configuration parameters decomposes the actual vibration signal, and the envelope demodulation spectrum corresponding to the decomposition result (each mode function) is shown in fig. 5 (c).
One-way valveThe normal frequency of the one-way valve is 1 Hz-1.034 Hz, and the approximate frequency f of the normal frequency range of the one-way valve is determined by the envelope demodulation spectrums of the IMF4 and the IMF5 r2 Characterized by 0.9375 Hz. Meanwhile, the envelope demodulation spectrum of IMF2 obtains f r1 0.625Hz and other frequency multiplication components (2 f) r1 、3f r1 、4f r1 、5f r1 、6f r1 、7f r1 、9f r1 ) Of 8f, furthermore r1 The frequency multiplication components highlighted by the envelope demodulation spectra of IMFs 4, IMFs 5, characterized in IMF4, are also labeled in fig. 5. These frequency characteristics become the dominant frequency of the check valve fault signal, indicating that the check valve must now fail. Therefore, the technical scheme provided by the invention has a good effect on detecting the fault characteristic frequency information of the one-way valve and has strong self-adaptability.
Experiments on simulation signals and actual signals show that the EEMD algorithm for adaptively configuring the optimal parameters can reduce the modal aliasing problem to the maximum extent, improve the decomposition performance of the method and enable the obtained inherent modal function to have actual physical significance. In addition, the method provided by the invention has the characteristics of better decomposition result, no need of manual intervention and the like, and can better realize fault diagnosis of mechanical equipment.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (4)
1. A method of diagnosing a fault in a mechanical device, comprising:
determining a vibration signal of the target mechanical equipment;
calculating modal chaos indexes corresponding to different noise amplitudes, and selecting the noise amplitude corresponding to the smallest modal chaos index as the optimal noise amplitude; the modal chaos index is obtained by calculating a decomposition result of a general empirical mode decomposition algorithm;
determining an expected decomposition error, and calculating an optimal ensemble averaging number based on the expected decomposition error and the optimal noise amplitude;
decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after parameter configuration to obtain each mode function; the ensemble empirical mode decomposition algorithm after the parameters are configured is an ensemble empirical mode decomposition algorithm obtained by setting the amplitude of the added noise as the optimal noise amplitude and setting the ensemble average times as the optimal ensemble average times on the basis of the ensemble empirical mode decomposition algorithm;
carrying out envelope demodulation spectrum analysis on each modal function obtained by decomposition to extract the characteristic frequency of the target mechanical equipment, and carrying out fault diagnosis on the target mechanical equipment based on the characteristic frequency of the target mechanical equipment;
the calculating of the modal chaos indexes corresponding to different noise amplitudes and selecting the noise amplitude corresponding to the smallest modal chaos index as the optimal noise amplitude specifically include:
determining a search range and a search step length of a noise amplitude coefficient;
generating a noise amplitude set according to the search range and the search step length; the ith noise amplitude in the noise amplitude set is epsilon i ＝α i ×std[x(t)]X (t) is the vibration signal of the target mechanical equipment, std [. cndot]As standard deviation operator, α i Is the ith noise amplitude coefficient;
calculating modal chaos indexes corresponding to different noise amplitudes in the noise amplitude set;
selecting the noise amplitude corresponding to the minimum modal chaos index as the optimal noise amplitude;
the calculating of the modal chaos indexes corresponding to different noise amplitudes in the noise amplitude set specifically includes:
sequentially configuring each noise amplitude in the noise amplitude set to a general empirical mode decomposition algorithm to obtain a plurality of general empirical mode decomposition algorithms after noise amplitudes are configured; the number of the overall empirical mode decomposition algorithms after the noise amplitude is configured is the same as that of the noise amplitudes, and different overall empirical mode decomposition algorithms after the noise amplitudes are configured with different noise amplitudes;
decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after the configuration of the calibrated noise amplitude to obtain a calibrated fluctuation index set corresponding to the calibrated noise amplitude; the calibration fluctuation index set comprises a segmented standard deviation fluctuation index of each mode; the calibrated noise amplitude is any noise amplitude in the noise amplitude set;
summing the segmental standard deviation fluctuation indexes of each mode in the calibration fluctuation index set to obtain a mode chaos index corresponding to each noise amplitude;
the calculation formula of the piecewise standard deviation fluctuation index of the kth modality is as follows:
among them, SSDVI k Is the fractional standard deviation fluctuation index of the kth mode, std [. cndot.]Is the standard deviation operator, and the standard deviation operator,
is the nth signal in the kth mode, n being 1, 2.
2. The method according to claim 1, wherein the determining an expected decomposition error and calculating an optimal ensemble averaging number based on the expected decomposition error and the optimal noise amplitude specifically comprises:
determining 1% of a standard deviation of the vibration signal as an expected decomposition error;
based on the formula
Calculating the optimal overall average times;
wherein N is opt Is the optimum global mean number of times, ε opt Is the optimum noise amplitude and e is the desired decomposition error.
3. A mechanical device fault diagnostic system, comprising:
the vibration signal determination module is used for determining a vibration signal of the target mechanical equipment;
the optimal noise amplitude determining module is used for calculating modal chaos indexes corresponding to different noise amplitudes and selecting the noise amplitude corresponding to the minimum modal chaos index as the optimal noise amplitude; the modal chaos index is obtained by calculating a decomposition result of a general empirical mode decomposition algorithm;
the optimal ensemble average times calculation module is used for determining an expected decomposition error and calculating the optimal ensemble average times based on the expected decomposition error and the optimal noise amplitude;
the modal function obtaining module is used for decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after parameters are configured to obtain each modal function; the ensemble empirical mode decomposition algorithm after the parameters are configured is an ensemble empirical mode decomposition algorithm obtained by setting the amplitude of the added noise as the optimal noise amplitude and setting the ensemble average times as the optimal ensemble average times on the basis of the ensemble empirical mode decomposition algorithm;
the fault diagnosis module is used for carrying out envelope demodulation spectrum analysis on each modal function obtained by decomposition so as to extract the characteristic frequency of the target mechanical equipment and carrying out fault diagnosis on the target mechanical equipment based on the characteristic frequency of the target mechanical equipment;
the optimal noise amplitude determination module specifically includes:
a search range and search step determining unit for determining a search range and a search step of the noise amplitude coefficient;
the noise amplitude set generating unit is used for generating a noise amplitude set according to the search range and the search step length; the ith noise amplitude in the noise amplitude set is epsilon i ＝α i ×std[x(t)]X (t) is the vibration signal of the target mechanical equipment, std [. cndot]As standard deviation operator, α i Is the ith noise amplitude coefficient;
the modal chaos index calculation unit is used for calculating modal chaos indexes corresponding to different noise amplitudes in the noise amplitude set;
the optimal noise amplitude determining unit is used for selecting the noise amplitude corresponding to the minimum modal chaos index as the optimal noise amplitude;
the modal chaos index calculation unit specifically comprises:
a configuration subunit, configured to sequentially configure each noise amplitude in the noise amplitude set to a ensemble empirical mode decomposition algorithm, so as to obtain a plurality of ensemble empirical mode decomposition algorithms configured with the noise amplitudes; the number of the overall empirical mode decomposition algorithms after the noise amplitude is configured is the same as that of the noise amplitudes, and different overall empirical mode decomposition algorithms after the noise amplitudes are configured with different noise amplitudes;
the calibration fluctuation index set determining subunit is used for decomposing the vibration signal by adopting a general empirical mode decomposition algorithm after the configuration of the calibration noise amplitude to obtain a calibration fluctuation index set corresponding to the calibration noise amplitude; the calibration fluctuation index set comprises a segmented standard deviation fluctuation index of each mode; the calibrated noise amplitude is any noise amplitude in the noise amplitude set;
the modal chaos index calculation subunit is used for summing the piecewise standard deviation fluctuation indexes of each modal in the calibration fluctuation index set to obtain a modal chaos index corresponding to each noise amplitude;
the calculation formula of the piecewise standard deviation fluctuation index of the kth modality is as follows:
among them, SSDVI k Is the fractional standard deviation fluctuation index of the kth mode, std [. cndot.]Is the standard deviation operator, and the standard deviation operator,
is the nth signal in the kth mode, n being 1, 2.
4. The mechanical equipment fault diagnosis system according to claim 3, wherein the optimal ensemble average times calculation module specifically includes:
an expected decomposition error determination unit for determining 1% of a standard deviation of the vibration signal as an expected decomposition error;
an optimal ensemble average number calculation unit for calculating the number of times of the transmission based on the formula
Calculating the optimal overall average times;
wherein N is opt Is the optimal ensemble average degree, ε opt Is the optimum noise amplitude and e is the desired decomposition error.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110941703.4A CN113654782B (en) | 2021-08-17 | 2021-08-17 | Mechanical equipment fault diagnosis method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110941703.4A CN113654782B (en) | 2021-08-17 | 2021-08-17 | Mechanical equipment fault diagnosis method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113654782A CN113654782A (en) | 2021-11-16 |
| CN113654782B true CN113654782B (en) | 2022-08-09 |
Family
ID=78480453
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110941703.4A Active CN113654782B (en) | 2021-08-17 | 2021-08-17 | Mechanical equipment fault diagnosis method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113654782B (en) |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102778357B (en) * | 2012-08-15 | 2014-10-15 | 重庆大学 | Mechanical failure feature extracting method based on optimal parameter ensemble empirical mode decomposition (EEMD) |
| CN110514441A (en) * | 2019-08-28 | 2019-11-29 | 湘潭大学 | A kind of Fault Diagnosis of Roller Bearings based on vibration signal denoising and Envelope Analysis |
| CN110503075B (en) * | 2019-08-29 | 2022-03-15 | 山东大学 | Planetary gearbox fault diagnosis method and system based on PSO-EEMD method |
-
2021
- 2021-08-17 CN CN202110941703.4A patent/CN113654782B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113654782A (en) | 2021-11-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102809493B (en) | Abnormal sound diagnosis device | |
| CN105424359B (en) | A kind of gear and bearing mixed fault feature extracting method based on sparse decomposition | |
| CN104483127A (en) | Method for extracting weak fault characteristic information of planetary gear | |
| CN105651376A (en) | Mechanical equipment offline detection system vibration signal spectrum analysis and alarm method | |
| CN105865794B (en) | Engine fire fault diagnosis method based on short time discrete Fourier transform and principal component analysis | |
| CN110160765A (en) | A kind of shock characteristic recognition methods and system based on sound or vibration signal | |
| CN108151869B (en) | Mechanical vibration characteristic index extraction method, system and device | |
| CN104849590B (en) | The lower Weak pulse signal detection method of a kind of mixed noise interference | |
| CN110426191B (en) | Fault diagnosis method for anti-interference rotating machine | |
| CN109029999B (en) | Rolling bearing fault diagnosis method based on enhanced modulation bispectrum analysis | |
| CN111898556A (en) | Method for extracting transient impact characteristics of bearings based on CK index consistency | |
| CN116873156B (en) | Intelligent monitoring method for power abnormality of natural gas ship based on big data | |
| CN109404285A (en) | The algorithm enhancing self-adaptive band-pass filter method that leapfrogs is shuffled in a kind of improvement of screw compressor fault diagnosis | |
| CN112287780A (en) | Spectral kurtosis algorithm-based mechanical equipment fault diagnosis method and system and readable storage medium | |
| CN112487882A (en) | Method for generating non-sparse index-guided enhanced envelope spectrum based on spectrum coherence | |
| CN113654782B (en) | Mechanical equipment fault diagnosis method and system | |
| CN109596354B (en) | Band-pass filtering method based on self-adaptive resonance frequency band identification | |
| Wang et al. | A novel optimal demodulation frequency band extraction method of fault bearing based on power spectrum screening combination-gram | |
| CN111623968B (en) | Fault feature extraction method based on adaptive morphological filtering | |
| CN116961799A (en) | Signal interference detection method based on time-frequency domain distribution characteristics | |
| Wang et al. | Fast spectral correlation based on sparse representation self-learning dictionary and its application in fault diagnosis of rotating machinery | |
| Xu et al. | A novel method for extracting maximum kurtosis component and its applications in rolling bearing fault diagnosis | |
| CN110529746B (en) | Method, device and equipment for detecting pipeline leakage | |
| Zheng et al. | Zero-Phase Filter-Based Adaptive Fourier Decomposition and Its Application to Fault Diagnosis of Rolling Bearing | |
| CN112883329B (en) | Mixed double-peak non-Gaussian response amplitude probability distribution model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |