CN117972638B - Multi-source multi-feature fusion diagnosis method for main bearing faults of aero-engine - Google Patents
Multi-source multi-feature fusion diagnosis method for main bearing faults of aero-engine Download PDFInfo
- Publication number
- CN117972638B CN117972638B CN202410384395.3A CN202410384395A CN117972638B CN 117972638 B CN117972638 B CN 117972638B CN 202410384395 A CN202410384395 A CN 202410384395A CN 117972638 B CN117972638 B CN 117972638B
- Authority
- CN
- China
- Prior art keywords
- main bearing
- frequency
- fault
- characteristic quantity
- band
- 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
- 230000004927 fusion Effects 0.000 title claims abstract description 70
- 238000003745 diagnosis Methods 0.000 title claims abstract description 52
- 238000000034 method Methods 0.000 title claims abstract description 43
- 239000002184 metal Substances 0.000 claims abstract description 56
- 239000010687 lubricating oil Substances 0.000 claims abstract description 53
- 239000002245 particle Substances 0.000 claims description 49
- 230000005294 ferromagnetic effect Effects 0.000 claims description 47
- 239000003921 oil Substances 0.000 claims description 17
- 230000005012 migration Effects 0.000 claims description 16
- 238000013508 migration Methods 0.000 claims description 16
- 238000001228 spectrum Methods 0.000 claims description 14
- 238000012544 monitoring process Methods 0.000 claims description 12
- 230000001133 acceleration Effects 0.000 claims description 11
- 238000004458 analytical method Methods 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 9
- 230000009467 reduction Effects 0.000 claims description 8
- 238000000354 decomposition reaction Methods 0.000 claims description 7
- 238000001514 detection method Methods 0.000 claims description 7
- 230000007704 transition Effects 0.000 claims description 7
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 5
- 230000005856 abnormality Effects 0.000 claims description 3
- 238000012067 mathematical method Methods 0.000 claims description 3
- 239000002923 metal particle Substances 0.000 claims description 3
- 238000010183 spectrum analysis Methods 0.000 claims description 3
- 230000008901 benefit Effects 0.000 abstract description 3
- 238000005299 abrasion Methods 0.000 description 6
- 238000004901 spalling Methods 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 230000001186 cumulative effect Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 239000000428 dust Substances 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 230000003595 spectral effect Effects 0.000 description 3
- 208000024891 symptom Diseases 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000005096 rolling process Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013524 data verification Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000002405 diagnostic procedure Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000001050 lubricating effect Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Landscapes
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a multisource multi-feature fusion diagnosis method for main bearing faults of an aeroengine. The result shows that compared with the means such as the onboard vibration value, the lubricating oil metal chip annunciator and the like, the fusion diagnosis method adopted by the invention has the advantages that the diagnosis accuracy reaches more than 95 percent, the main bearing fault is discovered at least 4.5 hours in advance, and the accuracy of main bearing fault diagnosis and the early warning and early warning of failure are greatly improved.
Description
Technical Field
The invention belongs to the field of aeroengine fault diagnosis, and particularly relates to an airborne multisource multi-feature fusion diagnosis method for main bearing faults of an aeroengine.
Background
The main bearing is a main support and a key piece of a modern aeroengine, and with the development of advanced aeroengine technology, a mechanical system is required to work at high temperature, high speed, high load, light weight, drastic state change, compact space limitation, long service life and high reliability, the quality and performance of the main bearing directly influence the reliability, safety and service life of the engine, and the main bearing failure is high and extremely harmful when the aeroengine main bearing works under severe working conditions such as high temperature, high speed, large load change interval, severe state change and the like. The main bearing can directly influence the use safety of the aeroengine once failing, the rotor system vibration can be increased, the rotor and stator can be rubbed, the transmission fails, and even catastrophic accidents can be caused when serious. Is acknowledged as one of the leading reasons for the air parking and the advance exchange of military aviation engines at home and abroad.
The fault diagnosis of the main bearing is a major technical problem of the most challenge of the aeroengine, the high-frequency vibration and the lubricating oil metal scraps are two key core key technologies which are mainly developed in recent years, and with the deep development of fault research and verification, the related technologies are gradually mature and enter model development successively. The high-frequency vibration has direct contribution to effectively capturing early faults of the main bearing, but is sometimes too sensitive due to the fact that vibration analysis influences many factors and large interference, so that false alarms occur; the lubricating oil metal chip monitoring method can measure ferromagnetic metal chips and nonferromagnetic metal chips in a lubricating oil pipeline on line in real time, is also an emerging means for monitoring the state of a transmission lubricating part, and is also frequently subjected to false alarm due to sensitivity to service environments such as vibration, high temperature and the like.
Therefore, in order to effectively solve the major challenges of main bearing diagnosis, it is very necessary to integrate the technical advantages of two new means, fully utilize the characteristics of high-frequency vibration that is sensitive in early and middle stages and that is sensitive in middle and later stages of lubricating oil metal scraps, and reduce the false alarm rate through fusion diagnosis and ensure that the system has ideal diagnosis accuracy.
Disclosure of Invention
Aiming at the defects related to the background technology, the invention provides a multisource multi-feature fusion diagnosis method for main bearing faults of an aeroengine.
The invention adopts the following technical scheme for solving the technical problems:
a multisource multi-feature fusion diagnosis method for main bearing faults of an aeroengine comprises the following steps:
Constructing a main bearing dimensionless fault frequency characteristic quantity F dB by using a frequency band envelope spectrum method based on a high-frequency vibration acceleration sensor installed on an aircraft, constructing a frequency band migration energy characteristic quantity FBE 5 by using a frequency band energy migration method, and carrying out fuzzy fusion on two homologous vibration characteristics of the main bearing dimensionless fault frequency characteristic quantity F dB and the frequency band migration energy characteristic quantity FBE 5;
Constructing a total accumulated mass characteristic quantity F WMass of ferromagnetic particles of the lubricating oil metal chips and a generation rate characteristic quantity F VMass of the ferromagnetic particles of the lubricating oil metal chips based on a lubricating oil metal chip sensor installed on an aircraft, and carrying out fuzzy fusion on two homologous characteristics of the total accumulated mass characteristic quantity F WMass of the ferromagnetic particles of the lubricating oil metal chips and the generation rate characteristic quantity F VMass of the ferromagnetic particles of the lubricating oil metal chips;
Finally, carrying out multi-source fuzzy fusion diagnosis on the main bearing dimensionless fault frequency characteristic quantity F dB, the frequency band migration energy characteristic quantity FBE 5, the total accumulated mass characteristic quantity F WMass of the ferromagnetic particles of the lubricating oil metal chips and the ferromagnetic particle generation rate characteristic quantity F VMass of the lubricating oil metal chips, and realizing accurate main bearing fault diagnosis and high-early-stage failure early warning.
Further, according to the collected high-frequency vibration acceleration signals, vibration acceleration stable signals when the rotating speed (namely N2) of the high-voltage rotor is greater than 95% of the rated maximum rotating speed are selected, 6 band-pass filters are respectively adopted to process the signals to obtain 6 different frequency bands, and envelope analysis, normalized autocorrelation noise reduction and spectrum analysis are respectively carried out to obtain 6 frequency band envelope spectrums; setting the variation range of the empirical contact angle of the main bearing in the state that the rotating speed of the high-pressure rotor of the engine is greater than 95% of the rated maximum rotating speed to be 20-45 degrees, obtaining the variation range of fault characteristic frequency to be F 01 -F 02, searching the maximum peak value of fault frequency in the variation range of the envelope spectrum characteristic frequency of each frequency band, calculating the dimensionless characteristic ratio of the maximum peak value to the average value of the energy of the frequency band, selecting the maximum value of the ratio in 6 frequency bands, and calculating the decibel (db) value of the maximum value to obtain the dimensionless fault frequency characteristic quantity F dB of the main bearing.
Further, according to the acquired high-frequency vibration acceleration signals, 6 band-pass filters are respectively adopted to process the acquired high-frequency vibration acceleration signals to obtain 6 signals in different frequency bands, an autocorrelation method is applied to noise reduction of the frequency band decomposition signals, the noise-reduced frequency band decomposition signals are obtained, the effective value of the frequency band decomposition signals is calculated, and the migration energy characteristic quantity of each frequency band is obtained: FBE 1、FBE2、FBE3、FBE4、FBE5、FBE6; the migration energy feature quantity FBE 5 of the 5 th frequency band is selected as the main bearing fault monitoring feature.
Further, the frequency band ranges of the 6 band pass filters are specifically band 1:25000-50000Hz; band 2:12500-25000Hz; band 3:6250-12500Hz; band 4:3125-6250Hz; band 5:1560-3125Hz; band 6:10-1560Hz.
Further, accumulating the total mass of ferromagnetic metal particles above 150 microns detected by an oil supply circuit lubricating oil metal chip sensor in each engine operation to obtain a total accumulated mass characteristic quantity F WMass of ferromagnetic particles of lubricating oil metal chips; and calculating the ratio of the total mass of the ferromagnetic particles with the diameter of more than 150 microns to the current working time of the engine in the current engine working process to obtain the characteristic quantity F VMass of the ferromagnetic particle generation rate of the lubricating oil metal scraps.
Further, in order to perform fusion diagnosis, it is first necessary to perform blurring processing on vibration and oil metal chip detection data, that is, uniformly processing the data to a number between 0 and 1 according to a warning limit T 1 and an anomaly limit T 2; for this purpose, a fuzzy membership function is constructed by using a fuzzy mathematical method, and four feature quantities are set: the main bearing dimensionless fault frequency characteristic quantity F dB, the migration energy characteristic quantity FBE 5 of the frequency band 5, the total accumulated mass characteristic quantity F WMass of the ferromagnetic particles of the lubricating oil metal chips, the relation R 11、R12、R13、R14 of the ferromagnetic particle generation rate characteristic quantity F VMass of the lubricating oil metal chips and the main bearing fault is 1, and the vibration fuzzy fusion characteristic and the lubricating oil metal chip fuzzy fusion characteristic are determined according to the fuzzy membership functions and the correlation coefficients of the four characteristic quantities.
Further, each of the warning limit T 1 and the anomaly limit T 2 is empirically set to: main bearing dimensionless fault frequency characteristic quantity F dB: warning limit: 15. abnormal limit: 20, a step of; the transition energy characteristic quantity FBE 5 of the 5 th frequency band: warning limit 0.05, anomaly limit: 0.1, a ferromagnetic particle total accumulated mass characteristic quantity F WMass of lubricating oil metal scraps: warning limit: 5mg, limit of abnormality: 50mg; ferromagnetic particle generation rate characteristic quantity F VMass of lubricating oil metal chip: warning limit: 0.2/min, anomaly limit: 0.5mg/min.
And further, carrying out multisource fuzzy reasoning fusion diagnosis on the vibration fuzzy fusion characteristic and the lubricating oil metal chip fuzzy fusion characteristic, setting a fault detection threshold value, and judging the fault when the fault detection threshold value is reached.
Further, the band-pass filter is optimally designed by comprehensively considering fault impact and cycle characteristics.
Compared with the prior art, the technical scheme provided by the invention has the following technical effects:
1) Compared with the traditional kurtosis, effective value and other monitoring indexes, the monitoring result obtained by the method provided by the invention has the advantages that the diagnosis and analysis of faults are more accurate and timely, various diagnosis indexes are fused, the diagnosis result is more effective, and early warning can be given out at the early stage of the faults;
2) The diagnosis method adopted by the invention can be used for fusing the evolution characteristics of different types of fault envelope spectrums, accurately analyzing faults and ensuring that test data and calculation data are very consistent.
In conclusion, the method can realize early weak fault warning of the rolling bearing, and has important significance for effectively implementing rolling bearing state monitoring, fault diagnosis and health management.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is an initial spalling failure physical diagram of the fulcrum bearing outer ring No. 3;
Fig. 3 is an engine on-line vibration outer ring failure frequency characteristic quantity F dB;
fig. 4 is a band shift energy characteristic quantity FBE 5;
FIG. 5 is a graph showing the total cumulative mass characteristic F WMass of ferromagnetic particles of an online lubricating oil metal chip;
FIG. 6 is an on-line ferromagnetic slip metal chip particle generation rate characteristic quantity F VMass;
FIG. 7 is a membership of the number 3 pivot main bearing outer race fault frequency feature F dB to the number 3 pivot main bearing outer race fault F 3;
FIG. 8 is a membership of band transition energy feature quantity FBE 5 to fulcrum No. 3 main bearing outer race fault F 3;
FIG. 9 is a vibration quantity fusion diagnostic result Fv;
FIG. 10 is a chart of cumulative abrasive particle mass versus number 3 pivot main bearing outer race failure F 3;
FIG. 11 is a chart of cumulative abrasive particle velocity versus number 3 pivot main bearing outer race failure F 3;
FIG. 12 is an oil fusion diagnostic result F O;
FIG. 13 is a vibration and online oil dust data fusion diagnostic result;
fig. 14 is a normal engine No. 3 fulcrum main bearing failure frequency characteristic quantity F dB;
Fig. 15 is a normal engine band shift energy characteristic quantity FBE 5;
FIG. 16 is a membership of a normal engine No. 3 fulcrum bearing outer ring fault frequency characteristic F dB to a No. 3 fulcrum bearing outer ring fault F 3;
FIG. 17 is a membership of normal engine band transition energy feature FBE 5 to fulcrum No. 3 main bearing outer race fault F 3;
FIG. 18 is a normal engine vibration amount fusion diagnostic result Fv;
Detailed Description
The technical scheme of the invention is further described in detail below with reference to the accompanying drawings:
As shown in fig. 1, a multi-source multi-feature fusion diagnosis method for main bearing faults of an aeroengine comprises the following steps:
S1: constructing fault frequency characteristic quantity F based on frequency band envelope spectrum dB
S2: constructing band shift energy characteristic quantity FBE based on band energy shift 5
S3: ferromagnetic particle total accumulated mass characteristic quantity F of online lubricating oil metal scraps WMass
S4: ferromagnetic particle generation rate characteristic quantity F VMass
S5: setting a No. 3 fulcrum outer ring peeling fault F 3, and performing fuzzification processing on monitoring data
S6: defining a fuzzy relation between fault characteristic quantity and fault mode
S7: fuzzy reasoning of vibration data fusion
S8: fuzzy reasoning of online oil chip data fusion
S9: multi-source fuzzy reasoning fusion diagnosis based on vibration fuzzy fusion characteristics and oil fuzzy fusion characteristics
S10: aeroengine test run data analysis and verification of prefabricated No. 3 fulcrum outer ring spalling fault
The fault frequency characteristic quantity according to the step 1 is characterized in that the construction method comprises the following steps:
S1-1: selecting stable signals with the N2 rotating speed (namely the rotating speed of the high-voltage rotor) being more than 95% of rated maximum rotating speed, obtaining signals of 6 different frequency bands by adopting band-pass filters, wherein the frequency band ranges of the 6 band-pass filters are respectively as follows:
band 1:25000-50000Hz;
band 2:12500-25000Hz;
band 3:6250-12500Hz;
Band 4:3125-6250Hz;
Band 5:1560-3125Hz;
Band 6:10-1560Hz;
S1-2: envelope analysis is carried out on the 6 signals respectively to obtain 6 envelope signals;
s1-3: normalized autocorrelation noise reduction is carried out on the 6 envelope signals;
S1-4: finally, carrying out spectrum analysis on the envelope signal after normalization autocorrelation noise reduction to obtain 6 frequency band envelope spectrums;
s1-5: setting the empirical contact angle change range of a No. 3 fulcrum angular contact ball bearing of a certain engine above 95% working condition to be 20-45 degrees, obtaining a characteristic frequency change range f 01、f02 according to a characteristic frequency calculation formula, searching the maximum value of spectral lines at fault characteristic frequency from an envelope spectrum in the frequency range, and setting the characteristic frequency tolerance range as Envelope spectrum interval is/>The number of frequency points within the tolerance range is: /(I)The maximum value of the fault frequency in the first frequency band envelope spectrum is,
In the formula, W l means the respective spectral line amplitudes of the 6 envelope signals.
S1-6: the number of spectral lines of the envelope spectrum set in the frequency range of 10 Hz-f o1+fo2 is N e, the average value of the envelope spectrum is,
In the formula, f i means a frequency value in the above frequency range.
S1-7: a dimensionless feature quantity is constructed,
S1-8: calculating dimensionless characteristic values of detail signals of each layerThe characteristic values of the detail signals are then compared, and the maximum value thereof is taken as the final characteristic value, that is,
S1-9: converted into decibel (dB) value output, defines fault frequency characteristic quantity as,
FdB=20log(F0)
Fig. 3 is an engine on-line vibration outer ring failure frequency characteristic quantity F dB.
The step S2 specifically includes:
s2-1: band-pass filtering is carried out on the acquired vibration signals to obtain 6 frequency band signals, namely: xf i (n), i=1, 2, …; n=0, 1,2, … N, where xf i (N) represents the band-pass filtered signal, N is the signal length, N is the nth point; the frequency band ranges of the 6 band pass filters are respectively:
band 1:25000-50000Hz;
band 2:12500-25000Hz;
band 3:6250-12500Hz;
Band 4:3125-6250Hz;
Band 5:1560-3125Hz;
Band 6:10-1560Hz;
S2-2: in order to eliminate the interference of the random signal, the frequency band decomposition signal is subjected to noise reduction by adopting an autocorrelation noise reduction method to obtain a frequency band signal I=1, 2, …, 6; n=0, 1,2, …, N; wherein N is the signal length;
s2-3: then, the effective values of the 6 band signals are calculated to obtain: FBE 1、FBE2、FBE3、FBE4、FBE5、FBE6;
S2-4: and finally, taking the 5 th frequency band migration energy characteristic quantity FBE 5 as a main bearing fault evolution monitoring characteristic quantity. Fig. 4 shows a band shift energy characteristic quantity FBE 5.
The step S3 specifically includes:
S3-1: and accumulating the total mass of ferromagnetic metal particles with the diameter of more than 150 microns detected by an oil supply circuit lubricating oil metal chip sensor in each engine operation to obtain the total accumulated mass characteristic quantity F WMass of the lubricating oil metal chip ferromagnetic particles. FIG. 5 is a graph showing the total cumulative mass characteristic F WMass of ferromagnetic particles of an online lubricating oil metal chip;
The step S4 specifically includes:
S4-1: and the characteristic quantity F VMass of the generation rate of the ferromagnetic lubricating oil metal chip particles is obtained by the ratio of the total mass of ferromagnetic particles generated in the operation of the secondary engine to the current operation time of the engine. Fig. 6 is an on-line ferromagnetic lubricating oil metal chip particle generation rate characteristic quantity F VMass.
The step S5 specifically includes:
s5-1: in order to perform fusion diagnosis, the vibration and oil detection data needs to be subjected to blurring processing, namely, the data is uniformly processed into numbers between 0 and 1 according to a warning limit T 1 and an anomaly limit T 2. Therefore, a fuzzy membership function is constructed by using a fuzzy mathematical method:
in the above formula, x refers to the feature quantity to be fused uniformly.
S5-2: features involved in fusion
(1) Vibration characteristic quantity
1) Number 3 fulcrum outer ring fault frequency characteristic quantity F dB
Warning limit: 15. abnormal limit: 20, a step of;
2) Band shift energy feature quantity FBE of 5 th 5
Warning limit: 0.05, outlier limit: 0.1;
(2) Parameters of oil particles
1) On-line lubricating oil metal chip ferromagnetic particle total accumulated mass characteristic quantity F WMass, warning limit: 5mg, limit of abnormality: 50mg;
2) On-line ferromagnetic lubricating oil metal chip particle generation rate characteristic quantity F VMass, warning limit: 0.2mg/min, abnormal limit: 0.5mg/min;
the step S6 specifically includes:
S6-1: the relation R 11 between the characteristic F dB of the failure frequency of the fulcrum outer ring and the failure of the fulcrum No. 3 shows the relation between the characteristic F dB and the peeling failure of the fulcrum main bearing No. 3, and the characteristic frequency can be found to be affirmatively generated according to the failure vibration simulation and the signal analysis, so that no ambiguity exists between the sign and the failure mode, and the ambiguity relation is determined to be R 11 =1.0 according to experience;
S6-2: the relation R 12 of the 5 th band transition energy characteristic quantity FBE 5 to the No. 3 fulcrum main bearing outer ring fault, which represents the relation between the band transition energy characteristic quantity FBE 5 and the No. 3 fulcrum main bearing outer ring spalling fault, according to fault vibration simulation and signal analysis, it is found that as the No. 3 fulcrum outer ring spalling fault evolves, FBE 5 has a monotonically increasing trend, so that the relation between the symptom and the fault mode is determined, no certain ambiguity exists, and the ambiguity relation is determined as R 12 =1.0 according to experience;
S6-3: the relation R v of the vibration characteristic quantity to the faults of the outer ring of the No. 3 fulcrum main bearing represents the relation between the vibration fusion characteristic quantity F v obtained by fusion of the characteristic F dB and the FBE 5 and the peeling faults of the outer ring of the No. 3 fulcrum main bearing, the relation between the vibration and the fault mode is determined according to fault vibration simulation and signal analysis, no certain ambiguity exists, and the fuzzy relation is determined to be R v =1.0 according to experience;
S6-4: the relation R 13、R14 between the total accumulated mass characteristic quantity F WMass of the ferromagnetic particles of the online lubricating oil metal chips and the generation rate characteristic quantity F VMass of the online ferromagnetic lubricating oil metal chips and the faults of the number 3 fulcrum main bearings is represented by the relation R WMass of the total accumulated mass characteristic quantity F WMass of the ferromagnetic particles of the online lubricating oil metal chips and the generation rate characteristic quantity F VMass of the online ferromagnetic lubricating oil metal chips and the faults of the number 3 fulcrum main bearings, according to the analysis of a mechanical abrasion principle, the abrasion monitoring is a direct monitoring method of the main bearings, the quantity of the abrasion particles directly reflects the fault severity of the main bearings, the accumulated mass of the abrasion particles is found to be monotonically increased along with the evolution of the abrasion precision, so that once the abrasion particles are greatly increased, the bearing is positively predicted to be seriously abraded, and the fault is already developed to be advanced, so that the relation between the symptoms and the fault modes is determined, and the relation between the symptoms and the fault modes is determined according to experience, and the relation is determined to be R 13=1.0、R14 =1.0;
The step S7 specifically includes:
s7-1: obtaining a fusion diagnosis result of the two vibration quantities by using a fuzzy reasoning method, wherein, 、The membership degree of the fault stripping fault of the No. 3 fulcrum main bearing according to the fault frequency characteristic F dB and the frequency band migration energy characteristic quantity FBE 5 is respectively, and R 11、R12 is the fuzzy relation of the fault frequency characteristic and the frequency band energy characteristic to the fault F 3. FIG. 7 is a membership of the number 3 pivot main bearing outer race fault frequency feature F dB to the number 3 pivot main bearing outer race fault F 3; FIG. 8 is a membership of band transition energy feature quantity FBE 5 to fulcrum No. 3 main bearing outer race fault F 3; the fuzzy reasoning process is determined as follows:
The step S8 specifically includes:
S8-1: obtaining a fusion diagnosis result of two oil particle quantities by using a fuzzy reasoning method, wherein, 、/>The degree of membership of the total accumulated mass characteristic quantity F WMass of the ferromagnetic particles of the lubricating oil metal scraps and the characteristic quantity F VMass of the generation rate of the ferromagnetic lubricating oil metal scraps to the fault peeling fault of the No. 3 fulcrum main bearing is shown in FIG. 10, and the degree of membership of the accumulated abrasive particle mass to the fault F 3 of the outer ring of the No. 3 fulcrum main bearing is shown; FIG. 11 is a chart of cumulative abrasive particle velocity versus number 3 pivot main bearing outer race failure F 3 membership. R 13、R14 is the fuzzy relation of fault frequency characteristic and frequency band energy characteristic to fault F3. The fuzzy reasoning process is determined as follows:
The step S8 specifically includes:
S9-1: and according to the fusion result of the vibration quantity, combining oil dust data, and obtaining a final fusion diagnosis result by a fuzzy reasoning method, wherein F v、FO is a vibration characteristic quantity fusion result and an oil characteristic quantity fusion result respectively, and R v、Rw is a fuzzy relation of the vibration fusion characteristic quantity and the oil fusion characteristic quantity to a fault F 3 respectively. Fig. 9 is a vibration characteristic amount fusion diagnosis result Fv; FIG. 12 is a result F O of oil feature fusion diagnosis; FIG. 13 is a diagnostic result of vibration and online oil dust data fusion.
The fuzzy reasoning process is determined as follows:
As a further optimization scheme of the airborne monitoring method for the main bearing faults of the aero-engine, the specific operation of the blurring processing in the step S5-1 is as follows:
S5-1-1: setting the warning limit T 1 =15 and the anomaly limit T 2 =20 of the F dB value, it can be seen that when the F dB value reaches 15, the membership degree is 0.5, namely the possibility of occurrence of a fault is considered to be 50%, and when the F dB value reaches 20, namely the possibility of occurrence of a fault is considered to be affirmed, and the membership degree is 1, namely the possibility of occurrence of a fault is considered to be 100%;
S5-1-2: assuming that the warning limit T 1 =0.05 and the anomaly limit T 2 =0.1 of the FBE 5, it can be seen that when the FBE 5 reaches 0.05, the membership is 0.5, i.e., the probability of occurrence of a fault is considered to be 50%, and when the FBE5 reaches 0.1, the probability of occurrence of a fault is considered to be affirmative, i.e., the membership is 1, i.e., the probability of occurrence of a fault is considered to be 100%.
S5-1-3: when the warning limit T 1 =5mg of F WMass and the abnormal limit T 2 =50mg are set, it can be seen that when F WMass reaches 5mg, the membership degree is 0.5, that is, the possibility of occurrence of a fault is considered to be 50%, and when F WMass reaches 50mg, it is considered that a fault is positively occurred, and when the membership degree is 1, that is, the possibility of occurrence of a fault is considered to be 100%.
S5-1-4: assuming that the warning limit t1=0.2 mg/min for F VMass and the anomaly limit t2=0.5 mg/min, it can be seen that when F VMass reaches 0.2mg/min, the membership is 0.5, i.e., the probability of occurrence of failure is considered to be 50%, and when F VMass reaches 0.5mg/min, it is considered that failure is definitely occurring, and the membership is1, i.e., the probability of failure is considered to be 100%.
Because the main bearing of the No. 3 fulcrum of the aeroengine has multiple faults, the characteristics and the rules of the main bearing of the No. 3 fulcrum under the typical spalling condition are studied in the test of the main bearing spalling fault complete machine of the engine. The fault preset state of the selected number 3 fulcrum main bearing outer ring is shown in figure 2.
The dither data sampling frequency was 200k, and each analysis period was 1s. And analyzing the high-frequency vibration acceleration data of more than 95% of rated maximum rotation speed of the intermediate case V2 measuring point.
And judging each data according to the final fusion diagnosis result, namely diagnosing the data as faults when the data are more than 0.5, and calculating the fault rate according to the fault. The fusion diagnosis results are shown in Table 1. Calculating to obtain the data fault diagnosis rate of 0.95151 which is higher than the rated maximum rotating speed of 95%; the fault diagnosis rate of 99% rated maximum rotation speed data is 0.96533. The detection rate is improved to a certain extent after the vibration characteristic quantity and the oil characteristic quantity are fused by comparison.
Table 1 comparison of fusion diagnostic test Rate results
To verify the false alarm rate condition of this method, and (3) analyzing vibration data of a V2 measuring point (namely a measuring point of the mounting edge of the bearing seat of the supporting point No. 3 of the intermediate case) in a state above 95% of the rated maximum rotating speed of the normal engine, wherein the fault characteristic quantity F dB of the supporting point No. 3 main bearing is shown in fig. 14. It can be seen that the failure frequency feature quantity F dB is substantially smaller than 15. The failure characteristic frequency of the outer ring of the No. 3 fulcrum main bearing of the engine is not obvious, so that the No. 3 fulcrum main bearing of the engine is predicted to have no outer ring peeling failure.
The analysis of the band shift energy characteristic quantity FBE 5 was performed on the V2-site vibration data of 95% or more of the normal engine, and the result is shown in fig. 15. It can be seen that the band shift energy characteristic amounts FBE 5 are substantially all smaller than 0.05. Indicating that the main bearing of the engine has no fault.
In order to perform fusion diagnosis on the peeling failure F3 of the outer ring of the fulcrum bearing No. 3, fuzzy fusion of two vibration amounts is performed first, fig. 16 is a fuzzy membership of the failure frequency characteristic amount F dB of the outer ring of the fulcrum bearing No. 3 to F3, fig. 17 is a fuzzy membership of the frequency band transfer energy characteristic amount FBE 5 to F3, and fig. 18 is a vibration fusion characteristic amount FV 3. It can be seen that, although the values of the feature quantity F dB and the FBE 5 are different in size, the fusion feature quantity FV 3 obtained by the fuzzy fusion has a value ranging from 0 to 1.
According to the formulated alarm threshold and fusion diagnosis rule, according to the fusion characteristic quantity FV 3, the diagnosis result is shown in Table 2, wherein the false alarm rate is 0.24969%. The diagnosis result is shown to reach a lower false alarm rate, and the correct validity of the diagnosis rule is fully shown.
TABLE 2 fusion of normal engine vibration characteristics false alarm rate results
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
While the foregoing is directed to embodiments of the present invention, other and further details of the invention may be had by the present invention, it should be understood that the foregoing description is merely illustrative of the present invention and that no limitations are intended to the scope of the invention, except insofar as modifications, equivalents, improvements or modifications are within the spirit and principles of the invention.
Claims (8)
1. A multisource multi-feature fusion diagnosis method for main bearing faults of an aeroengine is characterized in that a main bearing dimensionless fault frequency feature quantity F dB is constructed by using a frequency band envelope spectrum method based on an onboard high-frequency vibration acceleration sensor, a frequency band migration energy feature quantity FBE 5 is constructed by using a frequency band energy migration method, and fuzzy fusion is carried out on the main bearing dimensionless fault frequency feature quantity F dB and the frequency band migration energy feature quantity FBE 5;
Constructing a total accumulated mass characteristic quantity F WMass of ferromagnetic particles of the lubricating oil metal chips and a generation rate characteristic quantity F VMass of the ferromagnetic particles of the lubricating oil metal chips based on a lubricating oil metal chip sensor installed on an aircraft, and carrying out fuzzy fusion on the total accumulated mass characteristic quantity F WMass of the ferromagnetic particles of the lubricating oil metal chips and the generation rate characteristic quantity F VMass of the ferromagnetic particles of the lubricating oil metal chips;
Finally, carrying out multi-source fuzzy fusion diagnosis on the main bearing dimensionless fault frequency characteristic quantity F dB and the frequency band migration energy characteristic quantity FBE 5, the total accumulated quality characteristic quantity F WMass of the ferromagnetic particles of the lubricating oil metal chips and the ferromagnetic particle generation rate characteristic quantity F VMass of the lubricating oil metal chips, and realizing accurate main bearing fault diagnosis and large-early-quantity failure early warning; wherein,
According to the acquired high-frequency vibration acceleration signals, 6 band-pass filters are respectively adopted to process the acquired high-frequency vibration acceleration signals to obtain 6 signals with different frequency bands, an autocorrelation method is applied to noise reduction of the frequency band decomposition signals, the noise-reduced frequency band decomposition signals are obtained, the effective value of the frequency band decomposition signals is calculated, and the migration energy characteristic quantity of each frequency band is obtained: FBE 1、FBE2、FBE3、FBE4、FBE5、FBE6; the migration energy feature quantity FBE 5 of the 5 th frequency band is selected as the main bearing fault monitoring feature.
2. The multi-source multi-feature fusion diagnosis method for main bearing faults of an aeroengine according to claim 1, wherein vibration acceleration stable signals with the high-voltage rotor rotating speed being more than 95% of rated maximum rotating speed are selected according to collected high-frequency vibration acceleration signals, 6 band-pass filters are respectively adopted to process the signals to obtain 6 different frequency bands, and envelope analysis, normalized autocorrelation noise reduction and spectrum analysis are respectively carried out to obtain 6 frequency band envelope spectrums; setting the variation range of the empirical contact angle of the main bearing in the state that the rotating speed of the high-pressure rotor of the engine is greater than 95% of the rated maximum rotating speed to be 20-45 degrees, obtaining the variation range of fault characteristic frequency to be F 01 -F 02, searching the maximum peak value of fault frequency in the variation range of the envelope spectrum characteristic frequency of each frequency band, calculating the dimensionless characteristic ratio of the maximum peak value to the average value of the energy of the frequency band, selecting the maximum value of the ratio in 6 frequency bands, and calculating the decibel value of the maximum value to obtain the dimensionless fault frequency characteristic quantity F dB of the main bearing.
3. The multi-source multi-feature fusion diagnosis method for main bearing faults of an aeroengine according to claim 1 or 2, wherein the frequency band range of the 6 band pass filters is specifically band 1:25000-50000Hz; band 2:12500-25000Hz; band 3:6250-12500Hz; band 4:3125-6250Hz; band 5:1560-3125Hz; band 6:10-1560Hz.
4. The multi-source multi-feature fusion diagnosis method for main bearing faults of an aeroengine according to claim 1, wherein the total mass of ferromagnetic metal particles above 150 microns detected by an oil supply way lubricating oil metal chip sensor in each engine operation is accumulated to obtain a ferromagnetic particle total accumulated mass feature quantity F WMass of lubricating oil metal chips; and calculating the ratio of the total mass of the ferromagnetic particles with the diameter of more than 150 microns to the current working time of the engine in the current engine working process to obtain the characteristic quantity F VMass of the ferromagnetic particle generation rate of the lubricating oil metal scraps.
5. The multi-source multi-feature fusion diagnosis method for main bearing faults of an aeroengine according to claim 4, wherein in order to perform fusion diagnosis, firstly, vibration and lubricating oil metal chip detection data are required to be subjected to fuzzification treatment, namely, the data are uniformly processed into numbers between 0 and 1 according to a warning limit T 1 and an anomaly limit T 2; for this purpose, a fuzzy membership function is constructed by using a fuzzy mathematical method, and four feature quantities are set: the main bearing dimensionless fault frequency characteristic quantity F dB, the migration energy characteristic quantity FBE 5 of the frequency band 5, the total accumulated mass characteristic quantity F WMass of the ferromagnetic particles of the lubricating oil metal chips and the generating rate characteristic quantity F VMass of the ferromagnetic particles of the lubricating oil metal chips are respectively 1 in relation with the main bearing fault R 11、R12、R13、R14, and the vibration fuzzy fusion characteristic and the lubricating oil metal chip fuzzy fusion characteristic are determined according to the fuzzy membership function and the correlation coefficient of the four characteristic quantities.
6. The method for multi-source multi-feature fusion diagnosis of main bearing faults of an aeroengine according to claim 5, wherein each warning limit T 1 and anomaly limit T 2 are empirically set as: main bearing dimensionless fault frequency characteristic quantity F dB: warning limit: 15. abnormal limit: 20, a step of; the transition energy characteristic quantity FBE 5 of the 5 th frequency band: warning limit 0.05, anomaly limit: 0.1, a ferromagnetic particle total accumulated mass characteristic quantity F WMass of lubricating oil metal scraps: warning limit: 5mg, limit of abnormality: 50mg; ferromagnetic particle generation rate characteristic quantity F VMass of lubricating oil metal chip: warning limit: 0.2 mg/min, abnormal limit: 0.5mg/min.
7. The multi-source multi-feature fusion diagnosis method for main bearing faults of an aeroengine according to claim 5 is characterized in that multi-source fuzzy reasoning fusion diagnosis is carried out on the vibration fuzzy fusion feature and the lubricating oil metal chip fuzzy fusion feature, a fault detection threshold is set, and faults are judged if the threshold is reached.
8. The multi-source multi-feature fusion diagnosis method for main bearing faults of an aeroengine according to claim 1 or 2, wherein the band-pass filter is a band-pass filter which is optimally designed by comprehensively considering fault impact and cycle period features.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410384395.3A CN117972638B (en) | 2024-04-01 | 2024-04-01 | Multi-source multi-feature fusion diagnosis method for main bearing faults of aero-engine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410384395.3A CN117972638B (en) | 2024-04-01 | 2024-04-01 | Multi-source multi-feature fusion diagnosis method for main bearing faults of aero-engine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117972638A CN117972638A (en) | 2024-05-03 |
| CN117972638B true CN117972638B (en) | 2024-06-07 |
Family
ID=90853608
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410384395.3A Active CN117972638B (en) | 2024-04-01 | 2024-04-01 | Multi-source multi-feature fusion diagnosis method for main bearing faults of aero-engine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117972638B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0632201A1 (en) * | 1993-06-28 | 1995-01-04 | Copeland Corporation | Scroll compressor oil pumping system |
| CN115993245A (en) * | 2022-10-24 | 2023-04-21 | 中国人民解放军93208部队 | Special tester for bearings between rotors of military turbofan engine |
| CN116659860A (en) * | 2022-10-24 | 2023-08-29 | 中国人民解放军93208部队 | Novel method for monitoring main bearing fault evolution of aeroengine in service environment |
| CN117113049A (en) * | 2023-07-13 | 2023-11-24 | 中国航发沈阳发动机研究所 | Airborne early warning method for peeling fault of ball bearing outer ring in aero-engine |
| JP2023167548A (en) * | 2022-05-12 | 2023-11-24 | 三菱電機株式会社 | Electric blower and vacuum cleaner |
-
2024
- 2024-04-01 CN CN202410384395.3A patent/CN117972638B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0632201A1 (en) * | 1993-06-28 | 1995-01-04 | Copeland Corporation | Scroll compressor oil pumping system |
| JP2023167548A (en) * | 2022-05-12 | 2023-11-24 | 三菱電機株式会社 | Electric blower and vacuum cleaner |
| CN115993245A (en) * | 2022-10-24 | 2023-04-21 | 中国人民解放军93208部队 | Special tester for bearings between rotors of military turbofan engine |
| CN116659860A (en) * | 2022-10-24 | 2023-08-29 | 中国人民解放军93208部队 | Novel method for monitoring main bearing fault evolution of aeroengine in service environment |
| CN117113049A (en) * | 2023-07-13 | 2023-11-24 | 中国航发沈阳发动机研究所 | Airborne early warning method for peeling fault of ball bearing outer ring in aero-engine |
Non-Patent Citations (3)
| Title |
|---|
| 一种混合有限元模型的航空发动机主轴承剥落故障仿真分析方法;吴英祥等;《航空发动机》;20240314;1-8 * |
| 滑油中金属屑分析在航空发动机状态监控中的应用;姚红宇;《失效分析与预防》;20060831;60-63 * |
| 航空发动机预测健康管理关键技术分析;姜繁生等;《 第十五届中国科协年会第13分会场：航空发动机设计、制造与应用技术研讨会论文集》;20130525;1149-1154 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117972638A (en) | 2024-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109506921B (en) | Fault diagnosis and early warning method for rotary machine | |
| CN113204837B (en) | Rolling bearing fault diagnosis method based on multi-source information | |
| CN104990709A (en) | Method for detecting locomotive bearing fault | |
| CN109944809B (en) | Method for diagnosing faults of slurry circulating pump | |
| Kurfess et al. | Advanced diagnostic and prognostic techniques for rolling element bearings | |
| CN114636555B (en) | Fuzzy fusion diagnosis method and system for abrasion faults of aero-engine | |
| CN116659860B (en) | Novel method for monitoring main bearing fault evolution of aeroengine in service environment | |
| CN116304848B (en) | Rolling bearing fault diagnosis system and method | |
| CN117972638B (en) | Multi-source multi-feature fusion diagnosis method for main bearing faults of aero-engine | |
| CN111521399A (en) | Kurtosis index and envelope spectrum based early fault diagnosis method for space bearing | |
| Xia et al. | An improved underdamped asymmetric bistable stochastic resonance method and its application for spindle bearing fault diagnosis | |
| CN109828548B (en) | Performance degradation characteristic evaluation method based on time series change mutation point detection | |
| Liu et al. | A review of current condition monitoring and fault diagnosis methods for low-speed and heavy-load slewing bearings | |
| CN116644280A (en) | High-frequency vibration diagnosis method for failure of main bearing of engine | |
| CN111503242B (en) | Fault determination method, device and system, and computer readable storage medium | |
| CN117113049A (en) | Airborne early warning method for peeling fault of ball bearing outer ring in aero-engine | |
| CN112161805A (en) | Bearing tiny fault diagnosis method based on time series scale analysis and CNN | |
| Narendiranath Babu et al. | Fault analysis on continuous variable transmission using DB-06 wavelet decomposition and fault classification using ANN | |
| Zhen | Rotating Machinery Fault Diagnosis Based on Adaptive Vibration Signal Processing under Safety Environment Conditions | |
| Narendiranath Babu et al. | Vibration analysis of planetary gearbox using empirical mode decomposition and automatic fault prediction using artificial neural network | |
| WO2021128985A1 (en) | Device fault recognition systen and method | |
| CN112393907B (en) | Wind turbine generator bearing typical fault automatic diagnosis method based on sweep frequency analysis technology | |
| Guo et al. | Sparse Representation Based on MCKD and Periodic Dictionary for Bearing Fault Diagnosis | |
| Gopikuttan et al. | Maintenance Initiation Prediction Incorporating Vibrations and System Availability | |
| CN116956202B (en) | Engine fault detection method, system, equipment and storage medium |
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 |