CN117629637A - Aeroengine bearing fault diagnosis method and diagnosis system - Google Patents
Aeroengine bearing fault diagnosis method and diagnosis system Download PDFInfo
- Publication number
- CN117629637A CN117629637A CN202410098091.0A CN202410098091A CN117629637A CN 117629637 A CN117629637 A CN 117629637A CN 202410098091 A CN202410098091 A CN 202410098091A CN 117629637 A CN117629637 A CN 117629637A
- Authority
- CN
- China
- Prior art keywords
- confidence
- rule
- classifier
- layer
- representing
- 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
- 238000003745 diagnosis Methods 0.000 title claims abstract description 140
- 238000000034 method Methods 0.000 title claims abstract description 81
- 238000005457 optimization Methods 0.000 claims abstract description 26
- 230000008569 process Effects 0.000 claims abstract description 22
- 238000009826 distribution Methods 0.000 claims description 25
- 238000004364 calculation method Methods 0.000 claims description 18
- 230000004044 response Effects 0.000 claims description 14
- 238000005259 measurement Methods 0.000 claims description 12
- 238000010276 construction Methods 0.000 claims description 10
- 238000007781 pre-processing Methods 0.000 claims description 10
- 230000004913 activation Effects 0.000 claims description 9
- 238000011156 evaluation Methods 0.000 claims description 8
- 230000001133 acceleration Effects 0.000 claims description 7
- 238000006073 displacement reaction Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 6
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000004880 explosion Methods 0.000 abstract description 4
- 230000004927 fusion Effects 0.000 abstract description 4
- 239000010410 layer Substances 0.000 description 68
- 238000010586 diagram Methods 0.000 description 5
- 238000004590 computer program Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000036541 health Effects 0.000 description 3
- 238000013145 classification model Methods 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 239000002355 dual-layer Substances 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000000306 component Substances 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention provides a method and a system for diagnosing bearing faults of an aeroengine, which relate to the technical field of fault diagnosis, wherein a double-layer confidence rule classifier is constructed, a first-layer classifier of the double-layer confidence rule classifier completes rule fusion and preliminary fault diagnosis by using a evidence reasoning method, a second-layer classifier respectively constructs a plurality of affiliated confidence rule bases based on the preliminary diagnosis results of the first-layer classifier, each rule base respectively completes reasoning by applying the evidence reasoning method, and finally comprehensive fault diagnosis results are obtained by optimizing and iterating; according to the invention, the first layer classifier is used for solving the problem of combination rule explosion by using an interval structure, the second layer classifier is used for completing final fault diagnosis by using a plurality of auxiliary confidence rule bases, the multi-classification problem is converted into two classification problems of each auxiliary confidence rule base, and the interpretability constraint is added in the optimization process, so that the optimized parameters are ensured to accord with expert knowledge and actual conditions; and fault diagnosis of the aeroengine bearing is realized.
Description
Technical Field
The invention relates to the technical field of fault diagnosis, in particular to a method and a system for diagnosing bearing faults of an aero-engine.
Background
Aeroengines are used as the core components of aircraft, whose smooth operation is critical to ensuring flight safety. In the running process of the engine, the rolling bearing is a key component for bearing huge load, has a severe working environment and is extremely easy to generate faults. This may lead to serious flight accidents. Therefore, the fault diagnosis method of the aero-engine bearing is studied deeply, so that the early fault feature of the bearing can be found in time, and the accident can be effectively prevented.
Therefore, it is necessary to provide a method and a system for diagnosing bearing faults of an aeroengine to solve the above technical problems.
Disclosure of Invention
In order to solve the technical problems, the invention provides a method and a system for diagnosing the bearing faults of an aeroengine, wherein the method uses a double-layer confidence rule structure, a first layer classifier solves the problem of combination rule explosion by using a reference interval structure, a second layer classifier completes final fault diagnosis by using a plurality of auxiliary confidence rule bases, an interpretability constraint is added in the optimization process, and the optimized parameters are ensured to accord with a given range in an expert knowledge base; according to the method, the fault diagnosis of the aero-engine bearing is realized, so that the diagnosis result is more accurate and reliable, and the diagnosis process is more transparent.
The invention provides a method for diagnosing bearing faults of an aeroengine, which comprises the following steps:
obtaining actual measurement sample data of the aeroengine, wherein the actual measurement sample data are vibration signal data of the aeroengine under different fault conditions of the bearing, and the vibration signal data comprise displacement response and acceleration response;
constructing a fault diagnosis index system based on the measured sample data, wherein the fault diagnosis index system comprises at least one index;
determining reference values corresponding to the indexes according to the fault diagnosis index system, and constructing a reference value set based on the reference values corresponding to all the indexes;
constructing a double-layer confidence rule classifier based on the reference value set;
and performing fault diagnosis on the bearing of the aeroengine by using the double-layer confidence rule classifier to obtain a fault diagnosis result.
Preferably, the constructing a fault diagnosis index system based on the measured sample data includes:
performing data preprocessing on the vibration signal data, wherein the data preprocessing comprises cleaning and normalizing missing values, abnormal values and noise in the actually measured sample data;
extracting features of the vibration signal data subjected to data preprocessing to obtain a first feature set of an associated bearing, wherein the first feature set comprises frequency domain features, time domain features and statistical features;
performing feature screening on the first feature set to obtain a second feature set;
and constructing a fault diagnosis index system based on the second feature set.
Preferably, the determining the reference value corresponding to the index according to the fault diagnosis index system, and constructing the reference value set based on the reference values corresponding to all the indexes includes:
constructing an expert knowledge base based on historical fault data of the bearing, wherein the expert knowledge base comprises typical vibration characteristics of the bearing in different fault modes and correlations between the different fault modes;
determining a reference value for each index based on the expert knowledge base and historical fault data for the bearing;
a set of reference values is constructed based on all the reference values.
Preferably, the constructing a dual-layer confidence rule classifier based on the reference value set includes:
constructing a first layer classifier of the double-layer confidence rule classifier according to the reference value set, and outputting a preliminary diagnosis result by the first layer classifier, wherein the first layer classifier comprises a confidence rule base of an interval structure, and the kth rule in the confidence rule base of the interval structure is expressed as:
；
wherein,fault indicator system for an aircraft engine bearing constructed in one piece>Indicating the number of indexes in the system->Reference interval representing each index, < >>Is a set of diagnostic results that are presented,indicating the confidence level of the jth result in the kth rule, +.>Representing rule weights, ++>Representing rule reliability;
constructing a second-layer classifier of the double-layer confidence rule classifier based on the preliminary diagnosis result and an expert knowledge base, wherein the first-layer classifier
The two-layer classifier comprises at least one affiliated confidence rule base, and the kth rule of the affiliated confidence rule base is expressed as:
；
wherein,representing a fault index system of the secondarily constructed aero-engine bearing,the p-th reference value representing the i-th index, num being the total number of reference values possessed by the i-th index,/->Representing attribute weights, ++>Representing rule weights, ++>Representing rule reliability.
Preferably, the constructing a first layer classifier of the dual-layer confidence rule classifier according to the reference value set, where the first layer classifier includes a confidence rule base of an interval structure constructed based on the reference value set, and includes:
constructing a confidence rule base of an interval structure based on a fault diagnosis index system and a reference value set, and taking the confidence rule base as a first layer classifier of the double-layer confidence rule classifier;
after the first layer classifier is constructed, fusing rules in a confidence rule base of the interval structure based on evidence reasoning rules to finish the primary diagnosis of faults so as to obtain a primary diagnosis result.
Preferably, the fusing of rules in the confidence rule base of the interval structure based on the evidence reasoning rule completes the preliminary diagnosis of the fault to obtain a preliminary diagnosis result, including:
the ith evidence, i.e. the ith rule is subjected to confidence distribution representation, specifically:
；
wherein,representation->Under evidence, the preliminary diagnosis result is evaluated as +.>Reliability of the evaluation plan of +.>Representing the i-th attribute in the recognition frame +.>Confidence in case;
based on the evidence weight and the evidence confidence, carrying out mixed weighting on the confidence distribution of the ith evidence, wherein the mixed weighted confidence distribution of the ith evidence is expressed as follows:
；
wherein,representing the power set of the recognition frame and requiring the condition +.>，/>Representing the i-th evidence in the recognition frame +.>Lower mixing probability mass, < >>Indicating that the ith evidence is inEvaluation grade->The following mix probability mass;
the ith evidence is at the rating scaleLower mixing probability mass->The calculation formula of (2) is as follows:
；
；
；
wherein,representing normalized coefficient,/->And->Evidence of->Reliability and weight of (2);
for any two independent pieces of evidenceAnd->Assuming that its confidence distribution form can be defined byExpressed, two evidences are about proposition->Is->Calculated from the following formula:
；
；
based on the calculation process supported by the combination of the two independent evidences, propositions of L independent evidences are calculatedIs->This can be obtained by iterating the following formula:
；
；
；
；
wherein, the first k pieces of evidence are fusedConfidence in the result is recorded as->，/>Reflects the proposition +.>Outputs confidence distribution +.>And the expected utility value->The method comprises the following steps:
；
；
wherein the method comprises the steps ofRepresenting the result->Under utility value->Representation identification frame->Under utility value->Representing L independent evidence pairs recognition frames>Is used for the joint support degree of the (a).
Preferably, a second-layer classifier of the double-layer confidence rule classifier is constructed based on the preliminary diagnosis result and an expert knowledge base, wherein the second-layer classifier comprises at least one affiliated confidence rule base, and the method comprises the following steps:
subdividing the reference value set based on a fault diagnosis index system and the preliminary diagnosis result;
constructing confidence distribution of each affiliated confidence rule base of the second-layer classifier;
based on the subdivided reference value set and the confidence distribution of each affiliated confidence rule base, reasoning the preliminary diagnosis result by utilizing a evidence reasoning analysis algorithm to obtain a reasoning result, wherein the reasoning result comprises confidence degree and utility value;
and carrying out iterative optimization on the confidence coefficient, the rule weight and the attribute weight input by the expert knowledge base based on the P-CMA-ES method to obtain a final diagnosis result.
Preferably, the reasoning is performed on the preliminary diagnosis result by using a evidence reasoning analysis algorithm based on the confidence distribution of the subdivided reference value set and each affiliated confidence rule base to obtain a reasoning result, including:
first, calculate the matching degreeThe calculation formula is as follows:
；
wherein,represents->No. 5 of the input>A plurality of reference values;
then, a rule activation weight is calculatedNamely, the activation degree of the input information on the rule, the calculation formula is as follows:
；
wherein the method comprises the steps ofIs the matching degree of the input information relative to the reference value, < >>Representing initial rule weights, ++>Representing the initial attribute weights. If->It indicates that the rule is not activated.
And then calculating an output fault diagnosis result according to the rule activation weight, wherein the calculation formula is as follows:
；
；
wherein,represents the nth output reference level +.>The combined output confidence of (2); above->And->Indicates the number of confidence rules and the number of output levels, respectively,/->Representing utility values.
And finally, calculating an expected utility value to obtain a final output result, wherein the calculation formula is as follows:
；
where y represents the final output of the confidence rule base and can be expressed as，/>Representing the number of output levels, +.>Representing a single evaluation result->Utility of->Indicating the final desired utility.
Preferably, the iterative optimization is performed on the confidence coefficient, the rule weight and the attribute weight input by the expert knowledge base based on the P-CMA-ES method to obtain a final diagnosis result, including:
constructing an optimization model, wherein the optimization model is as follows:
；
；
wherein,and->The predicted value and the true value of the confidence rule base are indicated respectively,/->Representing the number of data samples;
based on interpretability constraintsAnd the interpretability constraint->Optimizing the optimization model, wherein an interpretability constraint is +.>And the interpretability constraint->The method comprises the following steps of:
；
；
the invention also provides an aeroengine bearing fault diagnosis system, which comprises:
the data acquisition module is used for acquiring actual measurement sample data of the aeroengine, wherein the actual measurement sample data are vibration signal data of the aeroengine under different fault conditions of the bearing, and the vibration signal data comprise displacement response and acceleration response;
the fault diagnosis index system construction module is used for constructing a fault diagnosis index system based on the actually measured sample data, wherein the fault diagnosis index system comprises at least one index;
the reference value set construction module is used for determining reference values corresponding to the indexes according to the fault diagnosis index system and constructing a reference value set based on the reference values corresponding to all the indexes;
the double-layer confidence rule classifier construction module is used for constructing a double-layer confidence rule classifier based on the reference value set;
and the fault diagnosis module is used for carrying out fault diagnosis on the bearing of the aeroengine by utilizing the double-layer confidence rule classifier to obtain a fault diagnosis result.
Compared with the related art, the aeroengine bearing fault diagnosis method and the aeroengine bearing fault diagnosis system provided by the invention have the following beneficial effects:
according to the invention, a fault diagnosis index system is constructed by combining expert knowledge and extracting a second feature set affecting an aeroengine from acquired actually measured sample data, and a double-layer confidence rule classification model is constructed by combining the expert knowledge, wherein a first-layer classifier of the double-layer confidence rule classification model completes rule fusion and preliminary fault diagnosis by using a evidence reasoning method, a plurality of affiliated confidence rule libraries are respectively constructed by combining the preliminary diagnosis results of the first-layer classifier, and each rule library respectively carries out reasoning and optimization iteration processes to obtain comprehensive fault diagnosis results; the invention combines expert knowledge and measured data samples, and applies the basic method of a confidence rule base to realize the fusion processing of two kinds of information, so that the fault diagnosis result is more accurate and reliable; the method comprises the steps that a double-layer confidence rule structure is used, a first-layer classifier solves the problem of combination rule explosion by using a reference interval structure, a second-layer classifier completes final fault diagnosis by using a plurality of auxiliary confidence rule bases, and an interpretability constraint is added in the optimization process, so that the optimized parameters are ensured to accord with a given range in an expert knowledge base; according to the method, the fault diagnosis of the aero-engine bearing is realized, so that the diagnosis result is more accurate and reliable, and the diagnosis process is more transparent.
Drawings
FIG. 1 is a flow chart of an aero-engine bearing fault diagnosis method provided by the invention;
FIG. 2 is another flow chart of an aircraft engine bearing fault diagnosis method provided by the invention;
FIG. 3 is a flow chart of constructing a two-layer confidence rule classifier in the present invention;
FIG. 4 is a flow chart of constructing a first layer classifier in the present invention;
FIG. 5 is a flow chart of constructing a second layer classifier in the present invention;
fig. 6 is a block diagram of an aero-engine bearing fault diagnosis system provided by the invention.
Detailed Description
The invention is further described below with reference to the drawings and embodiments.
Example 1
The invention provides a method for diagnosing bearing faults of an aeroengine, which is shown by referring to fig. 1 and 2, and comprises the following steps:
s1: and obtaining actual measurement sample data of the aeroengine, wherein the actual measurement sample data are vibration signal data of the aeroengine under different fault conditions of the bearing, and the vibration signal data comprise displacement response and acceleration response.
Specifically, vibration signal data of the aero-engine is collected, tidied and obtained by arranging a plurality of vibration signal sensors at a plurality of points of the aero-engine under different fault conditions, and the obtained vibration signal data sample types include, but are not limited to, displacement response and acceleration response.
S2: and constructing a fault diagnosis index system based on the measured sample data, wherein the fault diagnosis index system comprises at least one index.
Specifically, feature extraction is performed on the actually measured sample data obtained in the step S1, main features affecting the health condition of the aeroengine bearing are selected by combining expert knowledge in the field, and an index system for fault diagnosis is constructed by the selected features.
In this embodiment, step S2 includes:
and carrying out data preprocessing on the vibration signal data, wherein the data preprocessing comprises the steps of cleaning missing values, abnormal values and noise in the actually measured sample data and normalizing.
Specifically, the data preprocessing operation cleans the missing values, abnormal values and noise, and includes normalization and other operations to ensure the quality and availability of the vibration signal data.
And carrying out feature extraction on the vibration signal data subjected to data preprocessing to obtain a first feature set of the associated bearing, wherein the first feature set comprises frequency domain features, time domain features and statistical features.
Specifically, a first feature set related to the health condition of the aeroengine bearing is extracted from the vibration signal data, and the extracted first feature set includes, but is not limited to, frequency domain features, time domain features, statistical features, and the like.
And carrying out feature screening on the first feature set to obtain a second feature set.
Specifically, the first feature set is selected by combining expert knowledge in the field, and the second feature set with the most representation and key is selected, so that an index system for fault diagnosis is constructed.
And constructing a fault diagnosis index system based on the second feature set.
Specifically, the fault diagnosis index system comprises a series of numerical characteristics, represents indexes and reflects health conditions of different aspects of the bearing, and the indexes are used as inputs for subsequently constructing a double-layer confidence rule classifier.
S3: and determining reference values corresponding to the indexes according to the fault diagnosis index system, and constructing a reference value set based on the reference values corresponding to all the indexes.
Specifically, an expert knowledge base is constructed by relevant field experts according to fault data analysis of the aeroengine bearing and understanding of the working principle of the aeroengine bearing, and a reference value or a reference interval of an index is determined according to the established fault diagnosis index system, so that a corresponding reference value set is constructed.
In this embodiment, step S3 includes:
an expert knowledge base is constructed based on historical failure data of the bearing, wherein the expert knowledge base includes typical vibration characteristics of the bearing in different failure modes, and associations between the different failure modes.
Specifically, in cooperation with the expert in the related art, the expert is collected to collect the expert knowledge about the bearing failure of the aeroengine, wherein the expert may include mechanical engineers, aeroengineers and the like, needs to have deep understanding of the working principle and common failure modes of the aeroengine bearing, can determine the common failure modes, typical characteristics and relevant influencing factors of the aerobearing based on the analysis of historical failure data, and finally extracts key rules, correlations and experience knowledge from the expert cooperation and the failure data analysis to establish an expert knowledge base, wherein the knowledge may include typical vibration characteristics of different failure modes, correlations among failure modes and the like, and the representation form of the knowledge base includes but is not limited to failure trees and the like for failure diagnosis problems.
A reference value for each index is determined based on the expert knowledge base and historical fault data for the bearing.
Specifically, for each index in the fault diagnosis index system, a reference value or a reference interval is determined according to expert knowledge and historical data, and the reference values can be typical thresholds of fault states or typical ranges of normal working states.
A set of reference values is constructed based on all the reference values.
In particular, the reference values or reference intervals of all selected features are finally organized into a set, i.e. a reference value set.
S4: and constructing a double-layer confidence rule classifier based on the reference value set.
Specifically, when constructing a double-layer confidence rule classifier, firstly, constructing a confidence rule base of a first-layer classifier according to an initial reference value set to serve as a first layer of the classifier, then, respectively constructing a plurality of auxiliary confidence rule bases according to a preliminary diagnosis result of the first-layer classifier and combining an expert knowledge base to serve as a second-layer classifier of the classifier, and respectively carrying out reasoning and optimizing iteration processes on each auxiliary confidence rule base to obtain a fault diagnosis result.
In the present embodiment, referring to fig. 3 and 4, step S4 includes:
constructing a first layer classifier of the double-layer confidence rule classifier according to the reference value set, and outputting a preliminary diagnosis result by the first layer classifier, wherein the first layer classifier comprises a confidence rule base of an interval structure, and the kth rule in the confidence rule base of the interval structure is expressed as:
；
wherein,fault indicator system for an aircraft engine bearing constructed in one piece>Indicating the number of indexes in the system->Reference interval representing each index, < >>Is a set of diagnostic results that are presented,showing confidence of the jth result, < >>Representing rule weights, ++>Representing rule reliability.
Specifically, the steps include:
and constructing a confidence rule base of the interval structure based on the fault diagnosis index system and the reference value set, and taking the confidence rule base as a first layer classifier of the double-layer confidence rule classifier.
After the first layer classifier is constructed, fusing rules in a confidence rule base of the interval structure based on evidence reasoning rules to finish the primary diagnosis of faults so as to obtain a primary diagnosis result.
Specifically, the specific process of fusion is as follows:
the ith evidence, i.e. the ith rule is subjected to confidence distribution representation, specifically:
；
wherein the method comprises the steps of，Representation->Under evidence, the credibility of the evaluation plan, for which the preliminary diagnosis result is evaluated, +.>Representing the i-th attribute in the recognition frame +.>Confidence in case;
based on the evidence weight and the evidence confidence, carrying out mixed weighting on the confidence distribution of the ith evidence, wherein the mixed weighted confidence distribution of the ith evidence is expressed as follows:
；
wherein,representing the power set of the recognition frame and requiring the condition +.>，/>Representing the i-th evidence in the recognition frame +.>Lower mixing probability mass, < >>Indicating that the ith evidence is at the evaluation level +.>The following mix probability mass;
the ith evidence is at the rating scaleLower mixing probability mass->The calculation formula of (2) is as follows:
；
；
；
wherein,representing normalized coefficient,/->And->Evidence of->Reliability and weight of (2);
for any two independent pieces of evidenceAnd->Assuming that its confidence distribution form can be defined byExpressed, two evidences are about proposition->Is->Calculated from the following formula:
；
；
based on the calculation process supported by the combination of the two independent evidences, propositions of L independent evidences are calculatedIs->This can be obtained by iterating the following formula:
；
；
；
；
wherein, the first k pieces of evidence are fusedConfidence in the result is recorded as->，/>Reflects the proposition +.>Outputs confidence distribution +.>And the expected utility value->The method comprises the following steps:
；
；
wherein the method comprises the steps ofRepresenting the result->Under utility value->Representation identification frame->Under utility value->Representing L independent evidence pairs recognition frames>Is used for the joint support degree of the (a). />
Constructing a second-layer classifier of the double-layer confidence rule classifier based on the preliminary diagnosis result and an expert knowledge base, wherein the first-layer classifier
The two-layer classifier comprises at least one affiliated confidence rule base, and the kth rule of the affiliated confidence rule base is expressed as:
；
wherein,representing a fault index system of the secondarily constructed aero-engine bearing,the p-th reference value representing the i-th index, num being the total number of reference values possessed by the i-th index,/->Representing attribute weights, ++>Representing rule weights, ++>Representing rule reliability.
The construction process of the second layer classifier specifically comprises the following steps:
and subdividing the reference value set based on a fault diagnosis index system and the preliminary diagnosis result.
Confidence distributions of the respective affiliated confidence rule bases of the second-layer classifier are constructed.
And based on the subdivided reference value set and the confidence distribution of each affiliated confidence rule base, reasoning the preliminary diagnosis result by utilizing a evidence reasoning analysis algorithm to obtain a reasoning result, wherein the reasoning result comprises confidence degree and utility value.
Obtaining a fault diagnosis result of the aeroengine bearing through evidence reasoning and parameter optimization iteration of an affiliated confidence rule base of the second-layer classifier, and displaying the fault diagnosis result in a form of corresponding confidence and expected utility; obtaining a fault diagnosis result of the aeroengine bearing through evidence reasoning and parameter optimization iteration of an auxiliary confidence rule base of the second-layer classifier, and reasoning by adopting an ER analysis algorithm for the auxiliary confidence rule base, wherein the specific reasoning process is as follows:
first, calculate the matching degreeThe calculation formula is as follows:
；
wherein,represents->No. 5 of the input>A plurality of reference values;
then, a rule activation weight is calculatedNamely, the activation degree of the input information on the rule, the calculation formula is as follows:
；/>
wherein the method comprises the steps ofIs the matching degree of the input information relative to the reference value, < >>Representing initial rule weights, ++>Representing the initial attribute weights. If->It indicates that the rule is not activated.
And then calculating an output fault diagnosis result according to the rule activation weight, wherein the calculation formula is as follows:
；
；
wherein,represents the nth output reference level +.>The combined output confidence of (2); above->And->Indicates the number of confidence rules and the number of output levels, respectively,/->Representing utility values.
And finally, calculating an expected utility value to obtain a final output result, wherein the calculation formula is as follows:
；
where y represents the final output of the confidence rule base and can be expressed as，/>Representing the number of output levels, +.>Representing a single evaluation result->Utility of->Indicating the final desired utility.
And carrying out iterative optimization on the confidence coefficient, the rule weight and the attribute weight input by the expert knowledge base based on the P-CMA-ES method to obtain a final diagnosis result.
The optimization process is as follows:
to construct an optimization model, firstly, the optimized function is defined, and the optimized function is expressed as:
；
；
wherein,and->Representing the predicted and the actual values of the confidence rule base, respectively,/->Representing the number of data samples.
The optimized parameters are rule weight, confidence and attribute reliability respectively, and because the optimization process can negatively affect the method interpretability, the interpretability constraint is added in the optimization process, so that the optimization result does not violate the initial expert judgment, and each optimized parameter keeps the original physical meaning.
Based on interpretability constraintsAnd the interpretability constraint->Optimizing the optimization model, wherein an interpretability constraint is +.>And the interpretability constraint->The method comprises the following steps of:
；
；
wherein the interpretability constraintEnsures that the index maintains the original physical meaning and the interpretation constraint condition>The confidence distribution after optimization is ensured to be in accordance with the actual diagnosis result, the confidence distribution curve of each rule is ensured to be monotonous or convex, and the relevant threshold and the limit are provided and determined by an expert knowledge base.
S5: and performing fault diagnosis on the bearing of the aeroengine by using the double-layer confidence rule classifier to obtain a fault diagnosis result.
The invention provides a method for diagnosing bearing faults of an aero-engine, which comprises the following working principles: the method comprises the steps that a double-layer confidence rule classifier is built, a first-layer classifier of the double-layer confidence rule classifier completes rule fusion and preliminary fault diagnosis by using a evidence reasoning method, a second-layer classifier respectively builds a plurality of auxiliary confidence rule libraries based on preliminary diagnosis results of the first-layer classifier, each rule library respectively applies a evidence reasoning method to complete reasoning, and finally comprehensive fault diagnosis results are obtained through optimization iteration; according to the invention, the first layer classifier is used for solving the problem of combination rule explosion by using an interval structure, the second layer classifier is used for completing final fault diagnosis by using a plurality of auxiliary confidence rule bases, the multi-classification problem is converted into two classification problems of each auxiliary confidence rule base, and the interpretability constraint is added in the optimization process, so that the optimized parameters are ensured to accord with expert knowledge and actual conditions; the fault diagnosis of the aero-engine bearing is realized, so that the diagnosis result is more accurate and reliable, and the diagnosis process is more transparent.
Examples
The invention also provides an aero-engine bearing fault diagnosis system, which is applied to the aero-engine bearing fault diagnosis method, and comprises the following steps:
the data acquisition module is used for acquiring actual measurement sample data of the aeroengine, wherein the actual measurement sample data are vibration signal data of the aeroengine under different fault conditions of the bearing, and the vibration signal data comprise displacement response and acceleration response.
The fault diagnosis index system construction module is used for constructing a fault diagnosis index system based on the actually measured sample data, wherein the fault diagnosis index system comprises at least one index.
And the reference value set construction module is used for determining reference values corresponding to the indexes according to the fault diagnosis index system and constructing a reference value set based on the reference values corresponding to all the indexes.
And the double-layer confidence rule classifier construction module is used for constructing a double-layer confidence rule classifier based on the reference value set.
And the fault diagnosis module is used for carrying out fault diagnosis on the bearing of the aeroengine by utilizing the double-layer confidence rule classifier to obtain a fault diagnosis result.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flowchart and/or block of the flowchart illustrations and/or block diagrams, and combinations of flowcharts and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Those of ordinary skill in the art will appreciate that all or part of the steps of the various methods of the above embodiments may be implemented by hardware associated with a program stored in a computer-readable storage medium, including Read-Only Memory (ROM), random-access Memory (Random Access Memory, RAM), programmable Read-Only Memory (Programmable Read-Only Memory, PROM), erasable programmable Read-Only Memory (Erasable Programmable Read Only Memory, EPROM), one-time programmable Read-Only Memory (OTPROM), electrically erasable programmable Read-Only Memory (EEPROM), compact disc Read-Only Memory (CD-ROM), or any other medium capable of being used for computer-readable storage or carrying data.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.
Claims (10)
1. The diagnosis method for the bearing faults of the aero-engine is characterized by comprising the following steps of:
obtaining actual measurement sample data of the aeroengine, wherein the actual measurement sample data are vibration signal data of the aeroengine under different fault conditions of the bearing, and the vibration signal data comprise displacement response and acceleration response;
constructing a fault diagnosis index system based on the measured sample data, wherein the fault diagnosis index system comprises at least one index;
determining reference values corresponding to the indexes according to the fault diagnosis index system, and constructing a reference value set based on the reference values corresponding to all the indexes;
constructing a double-layer confidence rule classifier based on the reference value set;
and performing fault diagnosis on the bearing of the aeroengine by using the double-layer confidence rule classifier to obtain a fault diagnosis result.
2. The method for diagnosing a bearing failure of an aircraft engine according to claim 1, wherein said constructing a failure diagnosis index system based on said measured sample data comprises:
performing data preprocessing on the vibration signal data, wherein the data preprocessing comprises cleaning and normalizing missing values, abnormal values and noise in the actually measured sample data;
extracting features of the vibration signal data subjected to data preprocessing to obtain a first feature set of an associated bearing, wherein the first feature set comprises frequency domain features, time domain features and statistical features;
performing feature screening on the first feature set to obtain a second feature set;
and constructing a fault diagnosis index system based on the second feature set.
3. The method according to claim 1, wherein determining the reference values corresponding to the indexes according to the fault diagnosis index system, and constructing the reference value set based on the reference values corresponding to all the indexes, comprises:
constructing an expert knowledge base based on historical fault data of the bearing, wherein the expert knowledge base comprises typical vibration characteristics of the bearing in different fault modes and correlations between the different fault modes;
determining a reference value for each index based on the expert knowledge base and historical fault data for the bearing;
a set of reference values is constructed based on all the reference values.
4. A method of diagnosing an aircraft engine bearing failure in accordance with claim 3, wherein said constructing a two-layer confidence rule classifier based on said set of reference values comprises:
constructing a first layer classifier of the double-layer confidence rule classifier according to the reference value set, and outputting a preliminary diagnosis result by the first layer classifier, wherein the first layer classifier comprises a confidence rule base of an interval structure, and the kth rule in the confidence rule base of the interval structure is expressed as:
；
wherein,fault indicator system for an aircraft engine bearing constructed in one piece>Indicating the number of indexes in the system->Reference interval representing each index, < >>Is a set of diagnostic results that are presented,indicating the confidence level of the jth result in the kth rule, +.>Representing rule weights, ++>Representing rule reliability;
constructing a second-layer classifier of the double-layer confidence rule classifier based on the preliminary diagnosis result and an expert knowledge base, wherein the second-layer classifier comprises at least one subordinate confidence rule base, and the kth rule of the subordinate confidence rule base is expressed as:
；
wherein,representing a fault index system of the secondarily constructed aero-engine bearing,the p-th reference value representing the i-th index, num being the total number of reference values possessed by the i-th index,/->Representing attribute weights, ++>Representing rule weights, ++>Representing rule reliability.
5. The method of claim 4, wherein constructing a first layer classifier of the two-layer confidence rule classifier according to the reference value set, wherein the first layer classifier includes a confidence rule base of an interval structure constructed based on the reference value set, and comprises:
constructing a confidence rule base of an interval structure based on a fault diagnosis index system and a reference value set, and taking the confidence rule base as a first layer classifier of the double-layer confidence rule classifier;
after the first layer classifier is constructed, fusing rules in a confidence rule base of the interval structure based on evidence reasoning rules to finish the primary diagnosis of faults so as to obtain a primary diagnosis result.
6. The method for diagnosing an aeroengine bearing fault according to claim 5, wherein the merging rules in the confidence rule base of the interval structure based on the evidence reasoning rules to complete the preliminary diagnosis of the fault so as to obtain a preliminary diagnosis result comprises:
the ith evidence, i.e. the ith rule is subjected to confidence distribution representation, specifically:
；
wherein,representation->Under evidence, the preliminary diagnosis result is evaluated as +.>Reliability of the evaluation plan of +.>Representing the i-th attribute in the recognition frame +.>Confidence in case;
based on the evidence weight and the evidence confidence, carrying out mixed weighting on the confidence distribution of the ith evidence, wherein the mixed weighted confidence distribution of the ith evidence is expressed as follows:
；
wherein,representing the power set of the recognition frame and requiring the condition +.>，/>Representing the i-th evidence in the recognition frame +.>Lower mixing probability mass, < >>Indicating that the ith evidence is at the evaluation level +.>The following mix probability mass;
the ith evidence is at the rating scaleLower mixing probability mass->The calculation formula of (2) is as follows:
；
；
；
wherein,representing normalized coefficient,/->And->Evidence of->Reliability and weight of (2);
for any two independent pieces of evidenceAnd->It is assumed that its confidence distribution form can be defined by +.>Expressed, two evidences are about proposition->Is->Calculated from the following formula:
；
；
based on the calculation process supported by the combination of the two independent evidences, propositions of L independent evidences are calculatedIs->This can be obtained by iterating the following formula:
；
；
；
；
wherein, the first k pieces of evidence are fusedConfidence in the result is recorded as->，/>Reflects the proposition +.>Outputs confidence distribution +.>And the expected utility value->The method comprises the following steps:
；
；
wherein the method comprises the steps ofRepresenting the result->Under utility value->Representation identification frame->Under utility value->Representing L independent evidence pairs recognition frames>Is used for the joint support degree of the (a).
7. The method of claim 4, wherein constructing a second-tier classifier of the two-tier confidence rule classifier based on the preliminary diagnosis and expert knowledge base, wherein the second-tier classifier includes at least one affiliated confidence rule base, comprising:
subdividing the reference value set based on a fault diagnosis index system and the preliminary diagnosis result;
constructing confidence distribution of each affiliated confidence rule base of the second-layer classifier;
based on the subdivided reference value set and the confidence distribution of each affiliated confidence rule base, reasoning the preliminary diagnosis result by utilizing a evidence reasoning analysis algorithm to obtain a reasoning result, wherein the reasoning result comprises confidence degree and utility value;
and carrying out iterative optimization on the confidence coefficient, the rule weight and the attribute weight input by the expert knowledge base based on the P-CMA-ES method to obtain a final diagnosis result.
8. The method for diagnosing an aircraft engine bearing fault according to claim 7, wherein the reasoning about the preliminary diagnosis result using a evidence reasoning parsing algorithm based on the subdivided reference value set and the confidence distribution of each affiliated confidence rule base to obtain a reasoning result includes:
first, calculate the matching degreeThe calculation formula is as follows:
；
wherein,represents->No. 5 of the input>A plurality of reference values;
then, a rule activation weight is calculatedNamely, the activation degree of the input information on the rule, the calculation formula is as follows:
；
wherein the method comprises the steps ofIs the matching degree of the input information relative to the reference value, < >>Representing initial rule weights, ++>Representing an initial attribute weight; if->And then the rule is not activated, and the output fault diagnosis result is calculated according to the activation weight of the rule, wherein the calculation formula is as follows:
；
；
wherein,represents the nth output reference level +.>The combined output confidence of (2); above->And->Indicates the number of confidence rules and the number of output levels, respectively,/->And finally, calculating an expected utility value to obtain a final output result, wherein the calculation formula is as follows:
；
where y represents the final output of the confidence rule base and can be expressed as，/>Representing the number of output levels, +.>Representing a single evaluation result->Utility of->Indicating the final desired utility.
9. The method for diagnosing an aeroengine bearing fault according to claim 7, wherein the iterative optimization of the confidence level, the rule weights and the attribute weights inputted by the expert knowledge base based on the P-CMA-ES method to obtain a final diagnosis result comprises:
constructing an optimization model, wherein the optimization model is as follows:
；
；
wherein,and->Representing the predicted and the actual values of the confidence rule base, respectively,/->Representing the number of data samples;
based on the interpretability constraint stripPieceAnd the interpretability constraint->Optimizing the optimization model, wherein an interpretability constraint is +.>And the interpretability constraint->The method comprises the following steps of:
；
。
10. an aircraft engine bearing fault diagnosis system, comprising:
the data acquisition module is used for acquiring actual measurement sample data of the aeroengine, wherein the actual measurement sample data are vibration signal data of the aeroengine under different fault conditions of the bearing, and the vibration signal data comprise displacement response and acceleration response;
the fault diagnosis index system construction module is used for constructing a fault diagnosis index system based on the actually measured sample data, wherein the fault diagnosis index system comprises at least one index;
the reference value set construction module is used for determining reference values corresponding to the indexes according to the fault diagnosis index system and constructing a reference value set based on the reference values corresponding to all the indexes;
the double-layer confidence rule classifier construction module is used for constructing a double-layer confidence rule classifier based on the reference value set;
and the fault diagnosis module is used for carrying out fault diagnosis on the bearing of the aeroengine by utilizing the double-layer confidence rule classifier to obtain a fault diagnosis result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410098091.0A CN117629637B (en) | 2024-01-24 | 2024-01-24 | Aeroengine bearing fault diagnosis method and diagnosis system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410098091.0A CN117629637B (en) | 2024-01-24 | 2024-01-24 | Aeroengine bearing fault diagnosis method and diagnosis system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117629637A true CN117629637A (en) | 2024-03-01 |
| CN117629637B CN117629637B (en) | 2024-04-30 |
Family
ID=90035896
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410098091.0A Active CN117629637B (en) | 2024-01-24 | 2024-01-24 | Aeroengine bearing fault diagnosis method and diagnosis system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117629637B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113569478A (en) * | 2021-07-27 | 2021-10-29 | 郑州铁路职业技术学院 | Rotary machine fault diagnosis method based on element influence degree and confidence rule base |
| US20220083851A1 (en) * | 2020-09-11 | 2022-03-17 | Acoem France | Vibrating Machine Automated Diagnosis with Supervised Learning |
| CN115655717A (en) * | 2022-10-21 | 2023-01-31 | 厦门大学 | Bearing fault diagnosis method based on depth domain generalization network |
| CN116625686A (en) * | 2023-05-04 | 2023-08-22 | 中国航发沈阳发动机研究所 | On-line diagnosis method for bearing faults of aero-engine |
| CN117421633A (en) * | 2023-10-25 | 2024-01-19 | 哈尔滨师范大学 | Intelligent and credible rotary machine hierarchy fault diagnosis method |
| CN117421556A (en) * | 2023-10-25 | 2024-01-19 | 哈尔滨师范大学 | Trusted aviation equipment fault diagnosis method for solving ambiguity and uncertainty |
-
2024
- 2024-01-24 CN CN202410098091.0A patent/CN117629637B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20220083851A1 (en) * | 2020-09-11 | 2022-03-17 | Acoem France | Vibrating Machine Automated Diagnosis with Supervised Learning |
| CN113569478A (en) * | 2021-07-27 | 2021-10-29 | 郑州铁路职业技术学院 | Rotary machine fault diagnosis method based on element influence degree and confidence rule base |
| CN115655717A (en) * | 2022-10-21 | 2023-01-31 | 厦门大学 | Bearing fault diagnosis method based on depth domain generalization network |
| CN116625686A (en) * | 2023-05-04 | 2023-08-22 | 中国航发沈阳发动机研究所 | On-line diagnosis method for bearing faults of aero-engine |
| CN117421633A (en) * | 2023-10-25 | 2024-01-19 | 哈尔滨师范大学 | Intelligent and credible rotary machine hierarchy fault diagnosis method |
| CN117421556A (en) * | 2023-10-25 | 2024-01-19 | 哈尔滨师范大学 | Trusted aviation equipment fault diagnosis method for solving ambiguity and uncertainty |
Non-Patent Citations (2)
| Title |
|---|
| 胡昌华;冯志超;周志杰;胡冠宇;贺维;曹友;: "考虑环境干扰的液体运载火箭结构安全性评估方法", 中国科学:信息科学, no. 10, 16 October 2020 (2020-10-16), pages 1559 - 1573 * |
| 钱虹等: "基于故障样本的滚动轴承故障诊断置信规则库研究", 电力科学与技术学报, vol. 35, no. 1, 28 January 2020 (2020-01-28), pages 144 - 150 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117629637B (en) | 2024-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109583570B (en) | Method for determining abnormal data source of bridge health monitoring system based on deep learning | |
| Ragab et al. | Remaining useful life prediction using prognostic methodology based on logical analysis of data and Kaplan–Meier estimation | |
| CN110807257A (en) | Method for predicting residual life of aircraft engine | |
| CN114757309B (en) | Multi-physical-field monitoring data collaborative fusion engineering disaster early warning method and system | |
| CN109766583A (en) | Based on no label, unbalanced, initial value uncertain data aero-engine service life prediction technique | |
| CN112347898B (en) | Rolling bearing health index construction method based on DCAE neural network | |
| CN111337244B (en) | Method and device for monitoring and diagnosing faults of input shaft of fan gearbox | |
| CN112257914B (en) | Aviation safety causal prediction method based on random forest | |
| CN114048468A (en) | Intrusion detection method, intrusion detection model training method, device and medium | |
| CN112488142A (en) | Radar fault prediction method and device and storage medium | |
| CN115221973A (en) | Aviation bearing fault diagnosis method based on enhanced weighted heterogeneous ensemble learning | |
| CN110956331A (en) | Method, system and device for predicting operation state of digital factory | |
| CN117076869B (en) | Time-frequency domain fusion fault diagnosis method and system for rotary machine | |
| CN113988220A (en) | Method for evaluating health state of coal mining machine | |
| CN111474476B (en) | Motor fault prediction method | |
| CN117629637B (en) | Aeroengine bearing fault diagnosis method and diagnosis system | |
| CN114580472B (en) | Large-scale equipment fault prediction method with repeated cause and effect and attention in industrial internet | |
| CN112069621B (en) | Method for predicting residual service life of rolling bearing based on linear reliability index | |
| CN115660101A (en) | Data service providing method and device based on service node information | |
| CN113347014B (en) | Industrial control system situation combination prediction method based on time sequence | |
| CN114118249A (en) | Structure damage diagnosis method based on optimized stacked self-encoder and multi-signal fusion | |
| Tang et al. | The improvement of remaining useful life prediction for aero-engines by classification and deep learning | |
| Mostafaei et al. | Engine fault diagnosis decision-making with incomplete information using Dempster-Shafer theory | |
| CN116305531B (en) | Spacecraft health evolution model modeling method, device, equipment and medium | |
| CN118010103B (en) | Intelligent monitoring method and system for equal-thickness cement soil stirring wall in severe cold environment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |