CN113049251A - Bearing fault diagnosis method based on noise - Google Patents
Bearing fault diagnosis method based on noise Download PDFInfo
- Publication number
- CN113049251A CN113049251A CN202110289743.5A CN202110289743A CN113049251A CN 113049251 A CN113049251 A CN 113049251A CN 202110289743 A CN202110289743 A CN 202110289743A CN 113049251 A CN113049251 A CN 113049251A
- Authority
- CN
- China
- Prior art keywords
- signal
- data
- mean square
- root mean
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000003745 diagnosis Methods 0.000 title claims abstract description 29
- 238000012545 processing Methods 0.000 claims abstract description 31
- 238000001914 filtration Methods 0.000 claims abstract description 25
- 230000009466 transformation Effects 0.000 claims abstract description 22
- 238000001514 detection method Methods 0.000 claims abstract description 13
- 230000005236 sound signal Effects 0.000 claims abstract description 9
- 238000001228 spectrum Methods 0.000 claims description 19
- 238000012216 screening Methods 0.000 claims description 12
- 230000002159 abnormal effect Effects 0.000 claims description 8
- 238000012935 Averaging Methods 0.000 claims description 6
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 4
- 238000012544 monitoring process Methods 0.000 description 5
- 238000005096 rolling process Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 239000000284 extract Substances 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 1
- 230000009347 mechanical transmission Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
Landscapes
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- General Physics & Mathematics (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention provides a bearing fault diagnosis method based on noise, which is a non-contact detection method and comprises the following steps: 1: collecting a sound signal A0 when the equipment is operated; 2: performing DC removal processing on A0 to obtain a signal A; 3: calculating a root mean square value r0 and a sharpness s0 as initial values; 4: repeating the steps 1-2, calculating the root mean square value r and the sharpness s of the single group of signals A, judging whether r and s are increased by 50% relative to r0 and s0, if yes, performing the step 5, and if not, outputting the result that the equipment is normal; and 5: performing band-pass filtering and Fourier transformation on the signal A to obtain a signal AB, and performing hilbert transformation on the signal AB to obtain an envelope signal ABC; step 6: comparing and outputting a fault signal. The invention also provides a bearing fault diagnosis system based on noise. The invention has the advantages that: the method has the advantages that the non-contact detection is realized, the calculated amount is reduced, the map is enabled to be purer and more convenient to analyze, the bearing fault can be accurately diagnosed, and potential safety hazards caused by complex field environment are eliminated.
Description
Technical Field
The invention relates to the field of mechanical fault diagnosis, in particular to a bearing fault diagnosis method.
Background
The rolling bearing is an important component of the rotary machine, is one of the most prone to failure parts in the rotary machine, and is widely applied to the mechanical industry. Whether the rolling bearing normally runs or not has great influence on the reliability, precision, service life and other performances of the whole machine. Statistics show that nearly 30% of faults in the failure rate of the rotary machine are caused by faults of the rolling bearing, so that the condition monitoring and fault diagnosis of the rolling bearing are imperative to be researched.
The fault diagnosis method for the bearing based on the vibration signal is based on the theory that any machine consumes energy to do work externally, and simultaneously has partial energy consumed in various friction of mechanical transmission and generates normal vibration, the strength of the vibration is related to change and fault, abnormal vibration indicates that the fault tends to be serious, and the vibration characteristics caused by different faults are different; therefore, the vibration sensor can collect vibration signals and analyze the vibration characteristics of the vibration signals to detect the equipment state. Since the vibrations are generated while the machine is running, the method does not require the detection and analysis of faults in the event of a shutdown.
For example, patent application CN202011030897.4 discloses a method for diagnosing the fault vibration of a low-speed heavy-load bearing. Installing a set of vibration sensors on a low-speed heavy-duty bearing, the method comprising the steps of: acquiring vibration data of the low-speed heavy-load bearing during working in real time; identifying characteristic index information in vibration data between 1001Hz and 4000Hz and characteristic frequency in vibration data between 0Hz and 1000 Hz; establishing a vibration characteristic library of faults of the inner ring and the outer ring of the bearing, the rolling body and the retainer based on the rotation period parameters of the low-speed heavy-load bearing; comparing the actual monitoring characteristics with a characteristic library, identifying the fault type of the low-speed heavy-duty bearing, and sending fault diagnosis information to equipment operators in real time to remind the operators of shutdown maintenance; and updating the low-speed heavy-load bearing fault feature library according to the field maintenance feedback result.
However, the traditional vibration detection belongs to contact measurement, a vibration sensor must be fixedly installed on a proper measuring point (measuring position), the selection of the measuring point has a great influence on diagnosis, special talents with certain experience are needed, and the complex field environment may bring hidden dangers. And the retainer of the bearing assembly has light weight and small friction coefficient, so that the attenuation is serious when abnormal vibration is transmitted to a bearing assembly supporting upper measuring point, and the finally obtained result is inaccurate. In addition, the obtained vibration signal is interfered by factors such as environment, and the characteristic frequency of the fault is not easy to find out through analysis in a low frequency band.
Disclosure of Invention
The invention aims to solve the technical problem of how to accurately diagnose the bearing fault.
The invention solves the technical problems through the following technical means:
a bearing fault diagnosis method based on noise is characterized in that: the method uses a sound pressure sensor which is arranged at any position near the motor and is used for non-contact detection, and comprises the following steps:
step 1: collecting a sound signal A0 when the equipment runs through a sound pressure sensor;
step 2: carrying out direct current removal processing on the acquired signal A0 to obtain a signal A;
and step 3: screening a plurality of groups of direct-current-removed signals A as standard data, and calculating root mean square values r0 and sharpness s0 of the plurality of groups of standard data as initial values;
and 4, step 4: repeating the steps 1-2, calculating the root mean square value r and the sharpness s of the single-group direct-current-removed signal A, judging whether the root mean square value r and the sharpness s are increased by 50% relative to the initial values r0 and s0, if so, performing the step 5, and if not, outputting the result that the equipment is normal;
and 5: performing band-pass filtering and Fourier transformation on the signal A to obtain a signal AB of 1.5 kHz-3 kHz, and performing hilbert transformation on the signal AB to obtain an envelope signal ABC;
step 6: and searching a bearing library according to the bearing model to find out the corresponding characteristic frequency, comparing the characteristic frequency with the envelope signal ABC, corresponding to which fault type according to which characteristic frequency, and outputting unknown faults if the characteristic frequency is not matched.
The sound pressure sensor can be installed at any position near the motor, is non-contact detection, has a better effect in the aspect of retainer fault diagnosis compared with the traditional vibration monitoring, preliminarily judges the equipment state and the initial root mean square value r0 and sharpness value s0 by detecting the change of the root mean square value and sharpness of the sound signal, ensures that the envelope spectrum obtained by the method is purer and is beneficial to fault analysis, extracts a specific high-frequency-band signal in the step 5, gives a diagnosis result by calculation, reduces the calculation amount, ensures that the spectrum is purer and more convenient to analyze, can accurately diagnose the bearing fault, and eliminates potential safety hazard caused by complex field environment.
As a further specific technical scheme, the specific steps of performing dc removal processing on the collected signal a0 in step 2 to obtain the signal a are as follows:
the average value of the group, a0-mean (a0), was subtracted from each value in the acquired signal.
As a further specific technical solution, the multiple groups of dc-removed signals a in step 3 are 10 groups.
As a further specific technical solution, the specific steps of screening the standard data a and calculating the root mean square value r0 and the sharpness s0 of the initial values in step 3 are as follows:
step 3.1: repeating the steps 1-2 for 10 times to calculate the coefficient of variation cv of the 10 groups of data, if the coefficient of variation cv is<If the coefficient of variation cv is 0.15, the 10 sets of data are considered to be stable and can be used as standard data, and the steps 3.2 to 3.3 are executed to calculate the initial value>0.15, if the data is not stable enough, the step 3.4 is required, and the step 3.2 to 3.3 are executed to calculate the initial value until the condition is met, wherein
σ is a standard value, μ is an average value,
step 3.2: root mean square value
Wherein A1 and A2 … AN are the 1 st, 2 … th and N numerical values of each sampling data; the sharpness s reflects the high frequency energy in the sound,
where N' (z) is the loudness spectrum at a Bark, g (z) is an additional coefficient, a function of the critical band;
step 3.3: repeating the step 3.2 nine times to obtain 10 groups of root mean square values r and sharpness values s, and averaging to obtain initial values r0 and s 0;
step 3.4: repeating the steps 1-2 to obtain a direct current-removed signal A, repeating the step 3.2 to obtain a root mean square value r, removing the root mean square minimum value in 10 groups of data, adding the root mean square value of the current signal to form new 10 groups of data, and calculating a variation coefficient cv;
step 3.5: and if the coefficient of variation cv is judged to be 0.15, the 10 groups of data are considered to be stable and can be used as standard data, and the steps 3.2 to 3.3 are executed to calculate an initial value, and if the coefficient of variation cv is judged to be 0.15, the data are considered to be unstable, and the steps 3.4 to 3.5 are continuously executed until the cv is satisfied to be 0.15, and the initial value is obtained.
As a further specific technical solution, the specific processing procedure of performing band-pass filtering and fourier transform on the device abnormal signal a in step 5 to obtain a signal AB of 1.5kHz to 3kHz is as follows:
generating filter coefficients F (t) with the bandwidth of 1.5 kHz-3 kHz by utilizing matlab to carry out band-pass filtering and Fourier transformation on signals, wherein the filtering equation is
Resulting in a filtered signal AB.
The invention also provides a bearing fault diagnosis system based on noise, which uses a sound pressure sensor, the sound pressure sensor is arranged at any position near the motor, and the system is used for non-contact detection and comprises the following modules:
an acquisition module: collecting a sound signal A0 when the equipment runs through a sound pressure sensor;
a direct current removing processing module: carrying out direct current removal processing on the acquired signal A0 to obtain a signal A;
an initial value calculation module: screening a plurality of groups of direct-current-removed signals A as standard data, and calculating root mean square values r0 and sharpness s0 of the plurality of groups of standard data as initial values;
a judging module: repeating the calculation processes of the acquisition module and the direct current removal processing module, calculating the root mean square value r and the sharpness s of the single-group direct current removed signal A, judging whether the root mean square value r and the sharpness s are increased by 50% relative to the initial values r0 and s0, if so, entering an envelope signal generation module, and if not, outputting a result that the equipment is normal;
an envelope signal generation module: performing band-pass filtering and Fourier transformation on the signal A to obtain a signal AB of 1.5 kHz-3 kHz, and performing hilbert transformation on the signal AB to obtain an envelope signal ABC;
a fault output module: and searching a bearing library according to the bearing model to find out the corresponding characteristic frequency, comparing the characteristic frequency with the envelope signal ABC, corresponding to which fault type according to which characteristic frequency, and outputting unknown faults if the characteristic frequency is not matched.
As a further specific technical scheme, the specific steps of performing dc removal processing on the collected signal a0 in the dc removal processing module to obtain the signal a are as follows:
the average value of the group, a0-mean (a0), was subtracted from each value in the acquired signal.
As a more specific technical solution, the multiple groups of dc-removed signals a in the initial value calculation module are 10 groups.
As a further specific technical scheme, the specific steps of screening the standard data a and calculating the root mean square value r0 and the sharpness s0 of the initial value in the initial value calculation module are as follows:
step 3.1: repeating the steps 1-2 for 10 times to calculate the coefficient of variation cv of the 10 groups of data, if the coefficient of variation cv is<If the coefficient of variation cv is 0.15, the 10 sets of data are considered to be stable and can be used as standard data, and the steps 3.2 to 3.3 are executed to calculate the initial value>0.15, if the data is not stable enough, the step 3.4 is required, and the step 3.2 to 3.3 are executed to calculate the initial value until the condition is met, wherein
σ is a standard value, μ is an average value,
step 3.2: root mean square value
Wherein A1 and A2 … AN are the 1 st, 2 … th and N numerical values of each sampling data; the sharpness s reflects the high frequency energy in the sound,
where N' (z) is the loudness spectrum at a Bark, g (z) is an additional coefficient, a function of the critical band;
step 3.3: repeating the step 3.2 nine times to obtain 10 groups of root mean square values r and sharpness values s, and averaging to obtain initial values r0 and s 0;
step 3.4: repeating the steps 1-2 to obtain a direct current-removed signal A, repeating the step 3.2 to obtain a root mean square value r, removing the root mean square minimum value in 10 groups of data, adding the root mean square value of the current signal to form new 10 groups of data, and calculating a variation coefficient cv;
step 3.5: and if the coefficient of variation cv is judged to be 0.15, the 10 groups of data are considered to be stable and can be used as standard data, and the steps 3.2 to 3.3 are executed to calculate an initial value, and if the coefficient of variation cv is judged to be 0.15, the data are considered to be unstable, and the steps 3.4 to 3.5 are continuously executed until the cv is satisfied to be 0.15, and the initial value is obtained.
As a further specific technical solution, a specific processing procedure of performing bandpass filtering and fourier transform on the device abnormal signal a in the judgment module to obtain a signal AB of 1.5kHz to 3kHz is as follows:
generating filter coefficients F (t) with the bandwidth of 1.5 kHz-3 kHz by utilizing matlab to carry out band-pass filtering and Fourier transformation on signals, wherein the filtering equation is
Resulting in a filtered signal AB.
The invention has the advantages that: the sound pressure sensor can be installed at any position near the motor, is non-contact detection, has a better effect in the aspect of retainer fault diagnosis compared with the traditional vibration monitoring, preliminarily judges the equipment state, the initial root mean square value r0 and the sharpness value s0 by detecting the change of the root mean square value and the sharpness of the sound signal, ensures that the envelope spectrum obtained by the method is purer and is beneficial to the analysis of the fault, extracts a specific high-frequency band (1.5 kHz-3 kHz) signal in the step 5, gives out a diagnosis result by calculation, reduces the calculated amount, ensures that the spectrum is purer and more convenient to analyze, can accurately diagnose the bearing fault, and eliminates the potential safety hazard caused by complex field environment.
Drawings
FIG. 1 is a flow chart of a noise-based bearing fault diagnosis method of the present invention;
FIG. 2a is a time domain diagram of noise analysis according to an embodiment of the present invention, and FIG. 2b is an envelope spectrum of noise analysis according to an embodiment of the present invention;
FIG. 3a is a time domain diagram of a conventional vibration signal analysis, and FIG. 3b is an envelope spectrum of a conventional vibration signal analysis;
fig. 4 is an envelope spectrum after high frequency wave processing according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention, and it is obvious that the described embodiments are some embodiments of the present invention, but not all 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 achieve the above object, according to the present invention, there is provided a noise-based bearing fault diagnosis method using a sound pressure sensor, the sound pressure sensor being mountable at an arbitrary position near a motor and being capable of non-contact detection, comprising the steps of:
step 1: collecting a sound signal A0 when the equipment runs through a sound pressure sensor;
step 2: carrying out direct current removal processing on the acquired signal A0 to obtain a signal A;
and step 3: screening 10 groups of direct-current-removed signals A as standard data, and calculating root mean square values r0 and sharpness s0 of the 10 groups of standard data as initial values;
and 4, step 4: repeating the steps 1-2, calculating the root mean square value r and the sharpness s of the single-group direct-current-removed signal A, judging whether the root mean square value r and the sharpness s are increased by 50% relative to the initial values r0 and s0, if so, performing the step 5, and if not, outputting the result that the equipment is normal;
and 5: performing band-pass filtering and Fourier transformation on the signal A to obtain a signal AB of 1.5 kHz-3 kHz, and performing hilbert transformation on the signal AB to obtain an envelope signal ABC;
step 6: and searching a bearing library according to the bearing model to find out the corresponding characteristic frequency, comparing the characteristic frequency with the envelope signal ABC, corresponding to which fault type according to which characteristic frequency, and outputting unknown faults if the characteristic frequency is not matched.
Further, the step 2 of performing dc removal processing on the collected signal a0 to obtain the signal a specifically includes the following steps:
the average value of the group, a0-mean (a0), was subtracted from each value in the acquired signal.
Further, the specific steps of screening the standard data a and calculating the root mean square value r0 and the sharpness s0 of the initial values in the step 3 are as follows:
step 3.1: repeating the steps 1-2 for 10 times to calculate the coefficient of variation cv of the 10 groups of data, if the coefficient of variation cv is<If the coefficient of variation cv is 0.15, the 10 sets of data are considered to be stable and can be used as standard data, and the calculation of the initial value is performed in steps 3.2 to 3.3>0.15, if the data is not stable enough, the step 3.4 is required, and the step 3.2 to 3.3 are executed to calculate the initial value until the condition is met, wherein
σ is a standard value, μ is an average value,
step 3.2: root mean square value
Wherein A1 and A2 … AN are the 1 st, 2 … th and N numerical values of each sampling data; the sharpness s reflects the high frequency energy in the sound,
where N' (z) is the loudness spectrum at a Bark, g (z) is an additional coefficient, a function of the critical band;
step 3.3: repeating the step 3.2 nine times to obtain 10 groups of root mean square values r and sharpness values s, and averaging to obtain initial values r0 and s 0;
step 3.4: repeating the steps 1-2 to obtain a direct current-removed signal A, repeating the step 3.2 to obtain a root mean square value r, removing the root mean square minimum value in 10 groups of data, adding the root mean square value of the current signal to form new 10 groups of data, and calculating a variation coefficient cv;
step 3.5: and if the coefficient of variation cv is judged to be 0.15, the 10 groups of data are considered to be stable and can be used as standard data, and the steps 3.2 to 3.3 are executed to calculate an initial value, and if the coefficient of variation cv is judged to be 0.15, the data are considered to be unstable, and the steps 3.4 to 3.5 are continuously executed until the cv is satisfied to be 0.15, and the initial value is obtained.
Further, the specific processing procedure of performing band-pass filtering and fourier transform on the device abnormal signal a in step 5 to obtain a signal AB between 1.5kHz and 3kHz is as follows:
generating filter coefficients F (t) with the bandwidth of 1.5 kHz-3 kHz by utilizing matlab to carry out band-pass filtering and Fourier transformation on signals, wherein the filtering equation is
Resulting in a filtered signal AB.
The following table 1 shows the comparison of the noise diagnosis method of the present invention with the conventional vibration signal diagnosis method, and it can be seen from table 1 that the spectrum purity of the noise diagnosis method of the present invention is stronger, and the calculation amount can be greatly reduced, thereby greatly reducing the calculation time.
TABLE 1
| Vibration signal | The invention | |
| Purity of map | Weak (weak) | High strength |
| Calculating time | 0.97s | 0.53s |
FIG. 2a is a time domain diagram of noise analysis according to an embodiment of the present invention, and FIG. 2b is an envelope spectrum of noise analysis according to an embodiment of the present invention; fig. 3a is a time domain diagram of a conventional vibration signal analysis, and fig. 3b is an envelope spectrum of a conventional vibration signal analysis, it can be seen from a comparison between fig. 2 and fig. 3 that the method of the present invention can extract the characteristic frequency of 146Hz of the retainer, but the characteristic frequency is not extracted by the conventional vibration signal detection method, which indicates that the present application has a better effect on the retainer fault diagnosis compared with the conventional vibration monitoring. Fig. 4 is an envelope spectrum after the high-frequency wave processing according to the present invention, wherein, comparing fig. 2 with fig. 4, it can be seen that the envelope spectrum after the high-frequency filtering processing according to the present invention is purer and can obviously extract the characteristic frequency, which indicates that the envelope spectrum obtained according to the present invention is purer and is beneficial to the analysis of the fault.
Example two
Corresponding to the method of the first embodiment, the present invention further provides a noise-based bearing fault diagnosis system, which uses a sound pressure sensor, wherein the sound pressure sensor is installed at any position near a motor, and is used for non-contact detection, and the noise-based bearing fault diagnosis system includes the following modules:
an acquisition module: collecting a sound signal A0 when the equipment runs through a sound pressure sensor;
a direct current removing processing module: carrying out direct current removal processing on the acquired signal A0 to obtain a signal A;
an initial value calculation module: screening a plurality of groups of direct-current-removed signals A as standard data, and calculating root mean square values r0 and sharpness s0 of the plurality of groups of standard data as initial values;
a judging module: repeating the calculation processes of the acquisition module and the direct current removal processing module, calculating the root mean square value r and the sharpness s of the single-group direct current removed signal A, judging whether the root mean square value r and the sharpness s are increased by 50% relative to the initial values r0 and s0, if so, entering an envelope signal generation module, and if not, outputting a result that the equipment is normal;
an envelope signal generation module: performing band-pass filtering and Fourier transformation on the signal A to obtain a signal AB of 1.5 kHz-3 kHz, and performing hilbert transformation on the signal AB to obtain an envelope signal ABC;
a fault output module: and searching a bearing library according to the bearing model to find out the corresponding characteristic frequency, comparing the characteristic frequency with the envelope signal ABC, corresponding to which fault type according to which characteristic frequency, and outputting unknown faults if the characteristic frequency is not matched.
Further, the step of performing dc removal processing on the collected signal a0 in the dc removal processing module to obtain the signal a includes the following steps:
the average value of the group, a0-mean (a0), was subtracted from each value in the acquired signal.
Furthermore, the plurality of groups of signals A after the direct current is removed in the initial value calculation module are 10 groups.
Further, the specific steps of screening the standard data a and calculating the root mean square value r0 and the sharpness s0 of the initial value in the initial value calculation module are as follows:
step 3.1: repeating the steps 1-2 for 10 times to calculate the coefficient of variation cv of the 10 groups of data, if the coefficient of variation cv is<If the coefficient of variation is 0.15, the 10 sets of data are considered to be stable and can be used as standard data, and the calculation of the initial value is performed in steps 3.2-3.3cv>0.15, if the data is not stable enough, the step 3.4 is required, and the step 3.2 to 3.3 are executed to calculate the initial value until the condition is met, wherein
σ is a standard value, μ is an average value,
step 3.2: root mean square value
Wherein A1 and A2 … AN are the 1 st, 2 … th and N numerical values of each sampling data; the sharpness s reflects the high frequency energy in the sound,
where N' (z) is the loudness spectrum at a Bark, g (z) is an additional coefficient, a function of the critical band;
step 3.3: repeating the step 3.2 nine times to obtain 10 groups of root mean square values r and sharpness values s, and averaging to obtain initial values r0 and s 0;
step 3.4: repeating the steps 1-2 to obtain a direct current-removed signal A, repeating the step 3.2 to obtain a root mean square value r, removing the root mean square minimum value in 10 groups of data, adding the root mean square value of the current signal to form new 10 groups of data, and calculating a variation coefficient cv;
step 3.5: and if the coefficient of variation cv is judged to be 0.15, the 10 groups of data are considered to be stable and can be used as standard data, and the steps 3.2 to 3.3 are executed to calculate an initial value, and if the coefficient of variation cv is judged to be 0.15, the data are considered to be unstable, and the steps 3.4 to 3.5 are continuously executed until the cv is satisfied to be 0.15, and the initial value is obtained.
Further, the specific processing procedure of performing band-pass filtering and fourier transform on the device abnormal signal a in the judgment module to obtain a signal AB of 1.5kHz to 3kHz is as follows:
generating filter coefficients F (t) with the bandwidth of 1.5 kHz-3 kHz by utilizing matlab to carry out band-pass filtering and Fourier transformation on signals, wherein the filtering equation is
Resulting in a filtered signal AB.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A bearing fault diagnosis method based on noise is characterized in that: the method uses a sound pressure sensor which is arranged at any position near the motor and is used for non-contact detection, and comprises the following steps:
step 1: collecting a sound signal A0 when the equipment runs through a sound pressure sensor;
step 2: carrying out direct current removal processing on the acquired signal A0 to obtain a signal A;
and step 3: screening a plurality of groups of direct-current-removed signals A as standard data, and calculating root mean square values r0 and sharpness s0 of the plurality of groups of standard data as initial values;
and 4, step 4: repeating the steps 1-2, calculating the root mean square value r and the sharpness s of the single-group direct-current-removed signal A, judging whether the root mean square value r and the sharpness s are increased by 50% relative to the initial values r0 and s0, if so, performing the step 5, and if not, outputting the result that the equipment is normal;
and 5: performing band-pass filtering and Fourier transformation on the signal A to obtain a signal AB of 1.5 kHz-3 kHz, and performing hilbert transformation on the signal AB to obtain an envelope signal ABC;
step 6: and searching a bearing library according to the bearing model to find out the corresponding characteristic frequency, comparing the characteristic frequency with the envelope signal ABC, corresponding to which fault type according to which characteristic frequency, and outputting unknown faults if the characteristic frequency is not matched.
2. A noise-based bearing fault diagnosis method according to claim 1, characterized in that: the specific steps of performing dc removal processing on the collected signal a0 in the step 2 to obtain the signal a are as follows:
the average value of the group, a0-mean (a0), was subtracted from each value in the acquired signal.
3. A noise-based bearing fault diagnosis method according to claim 1, characterized in that: in step 3, the multiple groups of direct current removed signals A are 10 groups.
4. A noise-based bearing fault diagnosis method according to claim 3, wherein: the specific steps of screening the standard data A and calculating the root mean square value r0 and the sharpness s0 of the initial values in the step 3 are as follows:
step 3.1: repeating the steps 1-2 for 10 times to calculate the coefficient of variation cv of the 10 groups of data, if the coefficient of variation cv is<If the coefficient of variation cv is 0.15, the 10 sets of data are considered to be stable and can be used as standard data, and the steps 3.2 to 3.3 are executed to calculate the initial value>0.15, if the data is not stable enough, the step 3.4 is required, and the step 3.2 to 3.3 are executed to calculate the initial value until the condition is met, wherein
σ is a standard value, μ is an average value,
step 3.2: root mean square value
Wherein A1 and A2 … AN are the 1 st, 2 … th and N numerical values of each sampling data; the sharpness s reflects the high frequency energy in the sound,
where N' (z) is the loudness spectrum at a Bark, g (z) is an additional coefficient, a function of the critical band;
step 3.3: repeating the step 3.2 nine times to obtain 10 groups of root mean square values r and sharpness values s, and averaging to obtain initial values r0 and s 0;
step 3.4: repeating the steps 1-2 to obtain a direct current-removed signal A, repeating the step 3.2 to obtain a root mean square value r, removing the root mean square minimum value in 10 groups of data, adding the root mean square value of the current signal to form new 10 groups of data, and calculating a variation coefficient cv;
step 3.5: and if the coefficient of variation cv is judged to be 0.15, the 10 groups of data are considered to be stable and can be used as standard data, and the steps 3.2 to 3.3 are executed to calculate an initial value, and if the coefficient of variation cv is judged to be 0.15, the data are considered to be unstable, and the steps 3.4 to 3.5 are continuously executed until the cv is satisfied to be 0.15, and the initial value is obtained.
5. A noise-based bearing fault diagnosis method according to claim 1, characterized in that: the specific processing procedure of performing band-pass filtering and Fourier transformation on the equipment abnormal signal A in the step 5 to obtain a signal AB of 1.5 kHz-3 kHz is as follows:
generating filter coefficients F (t) with the bandwidth of 1.5 kHz-3 kHz by utilizing matlab to carry out band-pass filtering and Fourier transformation on signals, wherein the filtering equation is
Resulting in a filtered signal AB.
6. A noise-based bearing fault diagnostic system, characterized by: the sound pressure sensor is installed at any position near the motor, is used for non-contact detection and comprises the following modules:
an acquisition module: collecting a sound signal A0 when the equipment runs through a sound pressure sensor;
a direct current removing processing module: carrying out direct current removal processing on the acquired signal A0 to obtain a signal A;
an initial value calculation module: screening a plurality of groups of direct-current-removed signals A as standard data, and calculating root mean square values r0 and sharpness s0 of the plurality of groups of standard data as initial values;
a judging module: repeating the calculation processes of the acquisition module and the direct current removal processing module, calculating the root mean square value r and the sharpness s of the single-group direct current removed signal A, judging whether the root mean square value r and the sharpness s are increased by 50% relative to the initial values r0 and s0, if so, entering an envelope signal generation module, and if not, outputting a result that the equipment is normal;
an envelope signal generation module: performing band-pass filtering and Fourier transformation on the signal A to obtain a signal AB of 1.5 kHz-3 kHz, and performing hilbert transformation on the signal AB to obtain an envelope signal ABC;
a fault output module: and searching a bearing library according to the bearing model to find out the corresponding characteristic frequency, comparing the characteristic frequency with the envelope signal ABC, corresponding to which fault type according to which characteristic frequency, and outputting unknown faults if the characteristic frequency is not matched.
7. A noise-based bearing fault diagnostic system as defined in claim 6, wherein: the specific steps of performing dc removal processing on the collected signal a0 in the dc removal processing module to obtain the signal a are as follows:
the average value of the group, a0-mean (a0), was subtracted from each value in the acquired signal.
8. A noise-based bearing fault diagnostic system as defined in claim 6, wherein: the initial value calculation module has 10 sets of signals A after direct current removal.
9. A noise-based bearing fault diagnostic system as defined in claim 8, wherein: the specific steps of screening standard data A and calculating an initial value root mean square value r0 and sharpness s0 in the initial value calculation module are as follows:
step 3.1: repeating the steps 1-2 for 10 times to calculate the coefficient of variation cv of the 10 groups of data, if the coefficient of variation cv is<If the coefficient of variation cv is 0.15, the 10 sets of data are considered to be stable and can be used as standard data, and the steps 3.2 to 3.3 are executed to calculate the initial value>0.15, if the data is not stable enough, the step 3.4 is required to be carried out, and the step 3.2-3.3 are carried out again until the conditions are met to calculate the initial valueValue of wherein
σ is a standard value, μ is an average value,
step 3.2: root mean square value
Wherein A1 and A2 … AN are the 1 st, 2 … th and N numerical values of each sampling data; the sharpness s reflects the high frequency energy in the sound,
where N' (z) is the loudness spectrum at a Bark, g (z) is an additional coefficient, a function of the critical band;
step 3.3: repeating the step 3.2 nine times to obtain 10 groups of root mean square values r and sharpness values s, and averaging to obtain initial values r0 and s 0;
step 3.4: repeating the steps 1-2 to obtain a direct current-removed signal A, repeating the step 3.2 to obtain a root mean square value r, removing the root mean square minimum value in 10 groups of data, adding the root mean square value of the current signal to form new 10 groups of data, and calculating a variation coefficient cv;
step 3.5: and if the coefficient of variation cv is judged to be 0.15, the 10 groups of data are considered to be stable and can be used as standard data, and the steps 3.2 to 3.3 are executed to calculate an initial value, and if the coefficient of variation cv is judged to be 0.15, the data are considered to be unstable, and the steps 3.4 to 3.5 are continuously executed until the cv is satisfied to be 0.15, and the initial value is obtained.
10. A noise-based bearing fault diagnostic system as defined in claim 6, wherein: the specific processing process of performing band-pass filtering and Fourier transformation on the equipment abnormal signal A in the judgment module to obtain a signal AB between 1.5kHz and 3kHz is as follows:
generating filter coefficient F (t) with bandwidth of 1.5 kHz-3 kHz by utilizing matlab to perform band-pass filtering and Fourier transformation on signalsThe filter equation is
Resulting in a filtered signal AB.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110289743.5A CN113049251B (en) | 2021-03-16 | Bearing fault diagnosis method based on noise |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110289743.5A CN113049251B (en) | 2021-03-16 | Bearing fault diagnosis method based on noise |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113049251A true CN113049251A (en) | 2021-06-29 |
| CN113049251B CN113049251B (en) | 2024-09-24 |
Family
ID=
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113702041A (en) * | 2021-09-18 | 2021-11-26 | 华能威宁风力发电有限公司 | Bearing fault diagnosis method |
Citations (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11271181A (en) * | 1998-01-22 | 1999-10-05 | Nippon Steel Corp | Method and device for diagnosing failure in rolling bearing |
| TW462900B (en) * | 2000-10-11 | 2001-11-11 | China Steel Corp | Vibration diagnosis method for abnormality or malfunctions of roller of hot rolling machine for steel strip |
| JP2005300517A (en) * | 2004-03-17 | 2005-10-27 | Nippon Steel Corp | Diagnostic method and device for abnormality of roller bearing |
| CA2687785A1 (en) * | 2008-12-04 | 2010-06-04 | University Of Ottawa | Parameter independent detection of rotating machinery faults |
| KR101290928B1 (en) * | 2012-04-26 | 2013-07-29 | 한양대학교 산학협력단 | Equipment and method for diagnosing malfunction using sound quality parameter |
| US20130297257A1 (en) * | 2012-05-04 | 2013-11-07 | Jun Yang | Energy Separation System, Method, and Program Product for Detecting Faulty Bearings |
| CN105588720A (en) * | 2015-12-15 | 2016-05-18 | 广州大学 | Fault diagnosis device and method for antifriction bearing based on analysis on morphological component of acoustic signal |
| JP2016218502A (en) * | 2015-05-14 | 2016-12-22 | 株式会社リコー | Image processing device, control method, and program |
| CN106295683A (en) * | 2016-08-01 | 2017-01-04 | 上海理工大学 | A kind of outlier detection method of time series data based on sharpness |
| CN106568501A (en) * | 2016-10-25 | 2017-04-19 | 浙江工业大学 | Low-noise product sound quality objective parameter near-field detection method |
| US20180115226A1 (en) * | 2016-10-25 | 2018-04-26 | General Electric Company | Method and system for monitoring motor bearing condition |
| CN108444704A (en) * | 2018-03-30 | 2018-08-24 | 华中科技大学 | A kind of Fault Diagnosis of Rolling Element Bearings method |
| CN109406147A (en) * | 2018-10-29 | 2019-03-01 | 安徽大学 | Train bearing rail side acoustic diagnosis method under variable speed working condition |
| CN109855874A (en) * | 2018-12-13 | 2019-06-07 | 安徽大学 | A kind of accidental resonance filter of sound ancillary vibration small-signal enhancing detection |
| CN110487549A (en) * | 2019-09-10 | 2019-11-22 | 哈工大机器人（山东）智能装备研究院 | Bearing fault recognition methods, device, computer equipment and storage medium |
| CN211783846U (en) * | 2020-03-19 | 2020-10-27 | 上海大众动力总成有限公司 | Portable engine abnormal sound identification and positioning system |
| CN112052796A (en) * | 2020-09-07 | 2020-12-08 | 电子科技大学 | Permanent magnet synchronous motor fault diagnosis method based on deep learning |
| CN112464151A (en) * | 2020-10-27 | 2021-03-09 | 国电电力湖南新能源开发有限公司 | Wind turbine generator yaw system abnormal sound diagnosis method based on acoustic diagnosis |
Patent Citations (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11271181A (en) * | 1998-01-22 | 1999-10-05 | Nippon Steel Corp | Method and device for diagnosing failure in rolling bearing |
| TW462900B (en) * | 2000-10-11 | 2001-11-11 | China Steel Corp | Vibration diagnosis method for abnormality or malfunctions of roller of hot rolling machine for steel strip |
| JP2005300517A (en) * | 2004-03-17 | 2005-10-27 | Nippon Steel Corp | Diagnostic method and device for abnormality of roller bearing |
| CA2687785A1 (en) * | 2008-12-04 | 2010-06-04 | University Of Ottawa | Parameter independent detection of rotating machinery faults |
| KR101290928B1 (en) * | 2012-04-26 | 2013-07-29 | 한양대학교 산학협력단 | Equipment and method for diagnosing malfunction using sound quality parameter |
| US20130297257A1 (en) * | 2012-05-04 | 2013-11-07 | Jun Yang | Energy Separation System, Method, and Program Product for Detecting Faulty Bearings |
| JP2016218502A (en) * | 2015-05-14 | 2016-12-22 | 株式会社リコー | Image processing device, control method, and program |
| CN105588720A (en) * | 2015-12-15 | 2016-05-18 | 广州大学 | Fault diagnosis device and method for antifriction bearing based on analysis on morphological component of acoustic signal |
| CN106295683A (en) * | 2016-08-01 | 2017-01-04 | 上海理工大学 | A kind of outlier detection method of time series data based on sharpness |
| CN106568501A (en) * | 2016-10-25 | 2017-04-19 | 浙江工业大学 | Low-noise product sound quality objective parameter near-field detection method |
| US20180115226A1 (en) * | 2016-10-25 | 2018-04-26 | General Electric Company | Method and system for monitoring motor bearing condition |
| CN108444704A (en) * | 2018-03-30 | 2018-08-24 | 华中科技大学 | A kind of Fault Diagnosis of Rolling Element Bearings method |
| CN109406147A (en) * | 2018-10-29 | 2019-03-01 | 安徽大学 | Train bearing rail side acoustic diagnosis method under variable speed working condition |
| CN109855874A (en) * | 2018-12-13 | 2019-06-07 | 安徽大学 | A kind of accidental resonance filter of sound ancillary vibration small-signal enhancing detection |
| CN110487549A (en) * | 2019-09-10 | 2019-11-22 | 哈工大机器人（山东）智能装备研究院 | Bearing fault recognition methods, device, computer equipment and storage medium |
| CN211783846U (en) * | 2020-03-19 | 2020-10-27 | 上海大众动力总成有限公司 | Portable engine abnormal sound identification and positioning system |
| CN112052796A (en) * | 2020-09-07 | 2020-12-08 | 电子科技大学 | Permanent magnet synchronous motor fault diagnosis method based on deep learning |
| CN112464151A (en) * | 2020-10-27 | 2021-03-09 | 国电电力湖南新能源开发有限公司 | Wind turbine generator yaw system abnormal sound diagnosis method based on acoustic diagnosis |
Non-Patent Citations (9)
| Title |
|---|
| 伊海铭;侯磊;高朋;韩清扬;: "含中介轴承内圈局部缺陷的双转子系统的组合共振特性", 航空动力学报, no. 09, 15 September 2020 (2020-09-15) * |
| 傅勤毅, 王峰林, 李明章: "滚动轴承故障特征信号的提取", 轴承, no. 03, 5 March 1997 (1997-03-05) * |
| 吕琛, 王桂增, 叶昊, 黄晓中: "基于噪声小波包络谱的主轴承磨损故障诊断", 中南大学学报(自然科学版), no. 04, 30 August 2003 (2003-08-30) * |
| 周传迪, 董健等: "基于解卷积降噪和稀疏分解的风机轴承故障特征提取", 第三十九届中国控制会议论文集（7）, 27 July 2020 (2020-07-27) * |
| 周国宪;伍星;刘韬;: "基于cRIO的滚动轴承故障诊断系统设计", 昆明理工大学学报(自然科学版), no. 04, 15 August 2017 (2017-08-15) * |
| 易子馗;谭建平;: "一种基于心理声学品质和调制频率的车窗电机检测方法", 长安大学学报(自然科学版), no. 02, 15 March 2018 (2018-03-15) * |
| 李辉;郑海起;唐力伟;: "声测法和经验模态分解在轴承故障诊断中的应用", 中国电机工程学报, no. 15, 15 August 2006 (2006-08-15) * |
| 王俊: "基于多尺度非线性状态特征增强的旋转机械故障诊断方法研究", 中国博士学位论文全文数据库 工程科技Ⅱ辑, no. 2015, 16 August 2015 (2015-08-16) * |
| 肖淙文;王岩松;石磊;: "基于LS-SVM算法的加速车内噪声品质评价模型", 机械科学与技术, no. 01, 15 January 2015 (2015-01-15) * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113702041A (en) * | 2021-09-18 | 2021-11-26 | 华能威宁风力发电有限公司 | Bearing fault diagnosis method |
| CN113702041B (en) * | 2021-09-18 | 2024-04-23 | 华能威宁风力发电有限公司 | Bearing fault diagnosis method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110160791B (en) | System and method for diagnosing faults of induction motor bearing based on wavelet-spectral kurtosis | |
| US11441940B2 (en) | Condition monitoring apparatus, condition monitoring system, and condition monitoring method | |
| CN101726413B (en) | Method of fault diagnosis on ball socketed bearing of steel-making converter by comprehensive analysis | |
| CN109883703B (en) | Fan bearing health monitoring and diagnosing method based on vibration signal coherent cepstrum analysis | |
| CN102840907B (en) | Rolling bearing vibration signal characteristic extracting and analyzing method under early fault state | |
| JPH1026580A (en) | Method and device for diagnosing speed-change-type rotary mechanical equipment | |
| EP3788328B1 (en) | System and process for pattern matching bearing vibration diagnostics | |
| CN116304848B (en) | Rolling bearing fault diagnosis system and method | |
| EP4234927A2 (en) | State monitoring device, state monitoring system, and state monitoring method | |
| CN111060302A (en) | Multi-parameter equipment state comprehensive monitoring and diagnosing device and method | |
| CN117836599A (en) | Method for detecting bearing defects in a rotating system and monitoring system for implementing said method | |
| Thanagasundram et al. | A fault detection tool using analysis from an autoregressive model pole trajectory | |
| CN116644280A (en) | High-frequency vibration diagnosis method for failure of main bearing of engine | |
| CN113049251A (en) | Bearing fault diagnosis method based on noise | |
| CN113049251B (en) | Bearing fault diagnosis method based on noise | |
| Tranter | The fundamentals of, and the application of computers to, condition monitoring and predictive maintenance | |
| CN115235768A (en) | Rolling bearing fault diagnosis method and system | |
| CN107436244B (en) | Equipment fault alarm method based on frequency segmentation vibration data acquisition | |
| CN110057587A (en) | A kind of nuclear power pump bearing intelligent failure diagnosis method and system | |
| CN115683580A (en) | Rotating machinery fault diagnosis method based on phase domain adaptive filtering | |
| WO2022152336A1 (en) | A method for monitoring turbine blade vibration | |
| Thanagasundram et al. | Autoregressive based diagnostics scheme for detection of bearing faults | |
| KR101482511B1 (en) | Diagnosis System and Method of Bearing Defect by Phase Lag and Data Dispersion Shape Factor | |
| CN112525337A (en) | Method for preprocessing vibration monitoring data of mechanical press | |
| CN113567123B (en) | Automatic diagnosis method for impact faults of rotary machinery |
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 |