CN113033600A - Rotor misalignment state identification method based on improved D-S evidence fusion - Google Patents
Rotor misalignment state identification method based on improved D-S evidence fusion Download PDFInfo
- Publication number
- CN113033600A CN113033600A CN202110140695.3A CN202110140695A CN113033600A CN 113033600 A CN113033600 A CN 113033600A CN 202110140695 A CN202110140695 A CN 202110140695A CN 113033600 A CN113033600 A CN 113033600A
- Authority
- CN
- China
- Prior art keywords
- evidence
- rotor
- source
- fusion
- samples
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/257—Belief theory, e.g. Dempster-Shafer
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention relates to a rotor misalignment state identification method based on improved D-S evidence fusion. The method comprises the following steps: acquiring acceleration vibration data of a rotor in an out-of-centering state; acquiring an initial evidence source of a D-S evidence theory according to the acceleration vibration data; obtaining mutual information measure between evidence bodies According to mutual information measure between evidencesObtaining a similarity measure function between the evidence bodies; obtaining an evidence body E according to a similarity measure functioniTotal similarity of (c); acquiring a credible function of the evidence body according to the total similarity; obtaining the credibility of each evidence body according to the credibility function; according to eachCorrecting the initial evidence source according to the credibility of the individual evidence body to obtain a corrected evidence source; and synthesizing the corrected evidence source by using a D-S synthesis rule to obtain a fault state identification result of multi-source decision fusion. Therefore, the method greatly optimizes the fault identification process and improves the identification accuracy of the rotor misalignment state.
Description
Technical Field
The invention relates to the technical field of double-rotor misalignment fault identification, in particular to a rotor misalignment state identification method based on improved D-S evidence fusion.
Background
The rotor system is a core part of equipment such as an aircraft engine, a steam turbine, a wind power generator and the like, and the rotor of the rotor system is in a non-centering state in various forms due to manufacturing and mounting errors, abrasion of a supporting bearing, flexibility and thermal deformation of the rotor and the like, so that the whole machine vibrates excessively if the rotor system is light, the whole rotor system is damaged if the rotor system is heavy, and the safe and stable operation of the equipment is threatened. The misalignment fault cannot be avoided, but if the misalignment state (type, position and degree) of the rotor system can be mastered in time, maintenance measures can be taken in time, and the accident rate caused by the misalignment fault is reduced.
In practical engineering application, a rotor system is generally in a sealed narrow inner space, the operating environment is severe, sensors can only be installed on an outer shell, weak data acquired by a single sensor cannot accurately acquire misalignment state information of the rotor system, and therefore a plurality of sensors are required to be adopted for monitoring, and data acquired by the plurality of sensors are subjected to fusion diagnosis and analysis. And the information fusion level is divided into data level fusion, feature level fusion and decision level fusion. The decision layer fusion is the highest level fusion, and has strong anti-interference performance, high flexibility and good fault tolerance, and the main fusion methods comprise a D-S evidence theory, a Bayes inference and fuzzy theory and the like. Currently, the most common is the D-S evidence theory. The D-S evidence theory was first proposed by Dempster in 1967, and the D-S evidence theory has made a significant application basis in the aspects of multi-attribute decision problem analysis, information fusion application, intelligence analysis and the like. Although the D-S evidence theory has wide practical application effect, when there is conflict between evidences, the D-S synthesis rule will generate reasoning results contradictory to the practical situation, and paradoxical phenomenon of the evidence theory is generated. To solve the problem of collisions between evidence bodies, many scholars have been expressing distance measure functions of evidence. However, the conventional distance measure function still cannot accurately express the conflict problem between evidences in some cases.
Disclosure of Invention
In view of the above, there is a need to provide a rotor misalignment state identification method based on improved D-S evidence fusion.
A rotor misalignment state identification method based on improved D-S evidence fusion comprises the following steps:
acquiring acceleration vibration data of a rotor in an out-of-centering state;
acquiring an initial evidence source of a D-S evidence theory according to the acceleration vibration data;
According to mutual information measure between evidencesObtaining a similarity measure function between the evidence bodies;
obtaining an evidence body E according to a similarity measure functioniTotal similarity of (c);
acquiring a credible function of the evidence body according to the total similarity;
obtaining the credibility of each evidence body according to the credibility function;
correcting the initial evidence source according to the credibility of each evidence body to obtain a corrected evidence source;
and synthesizing the corrected evidence source by using a D-S synthesis rule to obtain a fault state identification result of multi-source decision fusion.
In some embodiments, the step of obtaining the acceleration vibration data of the rotor misalignment state comprises:
and acquiring a plurality of acceleration vibration data of the rotor in the non-centering state by using a double-rotor vibration test experiment mode.
In some embodiments, the step of obtaining an initial evidence source of D-S evidence theory from the acceleration vibration data comprises the steps of:
obtaining a data sample according to the acceleration vibration data; the data samples comprise training group samples and testing group samples;
using the training set samples to each DBNiThe model is trained to obtain a trained DBNiA model;
using trained DBNsiAnd the model carries out fault identification on the test group samples to obtain an initial evidence source of the D-S evidence theory.
In some of these embodiments, data samples are obtained from the acceleration vibration data; the data samples comprise training group samples and testing group samples, and the method comprises the following steps:
for the acceleration vibration data XiCarrying out filtering and denoising processing to obtain acceleration vibration data after the filtering and denoising processing;
equally dividing the acceleration vibration data subjected to filtering and denoising into l sections to obtain n groups of l sections of data samples Yi=[yi1,yi2,…,yil];
In some of these embodiments, each DBN is paired with a sample of the training setiThe model is trained to obtain a trained DBNiThe model comprises the following steps:
sampling the training setInputting corresponding number of i-th DBN as input sampleiModels and are respectively provided withDBN (double-sided tape)iThe number of layers of the network, the number of nodes of each layer, the iteration times of RBMs of each layer, the learning rate and the momentum items.
In some of these embodiments, trained DBNs are utilizediThe step that the model carries out fault identification on the test group samples to obtain an initial evidence source of the D-S evidence theory is as follows:
data samples of a test setInput ith DBNiTraining the model to obtain the ith initial fault state recognition result, and taking the ith initial state recognition result as the ith initial evidence source R of the D-S evidence theoryiWhere i ∈ (1,2, …, n).
In some of these embodiments, the similarity measure function between the evidence volumes is a similarity degree matrix SI, a similarity degree matrix
In some of these embodiments, the evidence volume E is obtained from a similarity measure functioniThe steps of total similarity of (1) are:
In some embodiments, the step of obtaining the credible function of the evidence body according to the total similarity comprises:
In some embodiments, the step of revising the initial evidence source according to the credibility of each evidence body to obtain a revised evidence source is as follows:
the credibility a of each evidence bodyi=C(mi) As a weight, weighting the initial evidence source according to the following formula to obtain a corrected evidence source;
According to the rotor misalignment state identification method based on improved D-S evidence fusion, the similarity degree between the evidence bodies is measured through mutual information measurement, and the initial evidence source is corrected through the similarity degree so as to solve the evidence conflict problem existing in the D-S evidence theory. Therefore, the rotor misalignment state identification method based on the improved D-S evidence fusion can effectively solve the problem that the D-S evidence theory has conflict among a plurality of evidence bodies, greatly optimizes the fault identification process and improves the identification accuracy of the rotor misalignment state.
Drawings
FIG. 1 is a schematic flow chart of a rotor misalignment state identification method based on D-S evidence fusion according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of step S20 of the rotor misalignment state identification method based on D-S evidence fusion shown in FIG. 1;
FIG. 3 is a schematic view of a rotor-support structure of a dual rotor test stand in accordance with an embodiment of the present invention;
FIG. 4 is a diagram illustrating a preliminary identification result of a rotor misalignment fault according to an embodiment of the present invention;
description of reference numerals: 110. an inner rotor; 120. an outer rotor; 130. a coupling is provided.
Detailed Description
To facilitate an understanding of the invention, the invention will now be described more fully with reference to the accompanying drawings. Preferred embodiments of the present invention are shown in the drawings. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
Unless defined otherwise, all 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. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
As described in the background, in order to improve the accuracy of rotor system misalignment identification, a plurality of misalignment data are generally subjected to fusion diagnosis analysis by using a D-S evidence theory and a DBN network. However, in the process of identifying the misalignment of the rotor system, when the D-S evidence theory faces the conflict between the evidences, the D-S synthesis rule generates an inference result contradicting with the actual situation, so that the accuracy of identifying and diagnosing the misalignment of the rotor is influenced.
Based on the reasons, the invention discloses a rotor misalignment state identification method based on improved D-S evidence fusion, and a method for improving D-S evidence combination by utilizing mutual information measure, which is used for solving the conflict problem between evidence bodies in the self-misalignment process.
As shown in fig. 1, the method for identifying the rotor misalignment state based on the improved D-S evidence fusion includes steps S10 to S90.
In step S10, acceleration vibration data of the double-rotor misalignment state is acquired.
Specifically, in step S10, acceleration vibration data of a misalignment fault of the dual rotors is obtained by using a dual rotor vibration test experiment.
More specifically, in the dual-rotor vibration test experiment process, sufficient acceleration vibration data of the dual-rotor misalignment fault are collected. In the embodiment of the invention, the double-rotor system refers to a double-rotor system which is nested inside and outside and rotates reversely.
Therefore, in the double-rotor vibration test experiment process, the misalignment state of the double-rotor system is adjusted by adding adjusting gaskets with different thicknesses at the bottoms of different supporting seats of the double-rotor vibration test bed, slight misalignment is defined when the thickness of the added adjusting gaskets is 0.2mm, and serious misalignment is defined when the thickness of the added adjusting gaskets is 0.5 mm.
When the acceleration vibration data in the rotor misalignment state is acquired by the dual-rotor vibration test stand, a plurality of acceleration vibration data in the rotor misalignment state are generally acquired by a plurality of acceleration sensors. In the rotor misalignment identification process, the ith acceleration sensor T is assumediThe acquired acceleration vibration data is XiWhere i ∈ (1,2, …, n).
And step S20, acquiring an initial evidence source of the D-S evidence theory according to the acceleration vibration data.
Specifically, as shown in fig. 2, step S20 includes steps S21 to S23.
In step S21, data samples are obtained from the acceleration vibration data. The data samples include training set samples and test set samples.
Specifically, for the acceleration vibration data XiCarrying out filtering and denoising processing to obtain acceleration vibration data after the filtering and denoising processing; equally dividing the acceleration vibration data subjected to filtering and denoising into l sections to obtain n groups of l sections of data samples Yi=[yi1,yi2,…,yil](ii) a Sample data YiRandom grouping into training set samplesAnd test group samples
Step S22, using training set samples for each DBNiThe model is trained to obtain a trained DBNiAnd (4) modeling.
Specifically, the training set samples areInputting corresponding number of i-th DBN as input sampleiSeparately set DBNiThe number of layers of the network,greedy training is performed layer by layer from the bottom layer to the high layer according to the number of nodes of each layer and parameters such as the iteration times, the learning rate and the momentum term of RBMs of each layer; according to the labels of the training set and the classification rules of the Soft-max classifier, parameters are finely adjusted from the highest layer to the lowest layer step by step in a reverse mode, and therefore each DBN is completediAnd (3) a training process of the model, wherein i belongs to (1,2, …, n).
Step S23, using the trained DBNiAnd the model carries out fault identification on the test group samples to obtain an initial evidence source of the D-S evidence theory.
Specifically, the group samples are testedInput ith DBNiIn the model, an ith preliminary fault state identification result is obtained, and the ith preliminary fault state identification result is used as an ith initial evidence source Ri of a D-S evidence theory, wherein i belongs to (1,2, …, n).
Step S40, according to mutual information measure between evidencesA similarity measure function between the evidence volumes is obtained.
Specifically, the similarity measure function between the evidence bodies is a similarity degree matrix SI. Similarity degree matrix
Step S50, obtaining an evidence body E according to the similarity measure functioniTotal similarity of (c).
Specifically, adding each row in the similarity degree matrix SI can obtain the evidence body EiTotal similarity of
And step S60, acquiring the credible function of the evidence body according to the total similarity.
And step S70, obtaining the credibility of each evidence body according to the credibility function.
In particular, the confidence level a of each evidence bodyi=C(mi)。
And step S80, correcting the initial evidence source according to the credibility of each evidence body to obtain a corrected evidence source.
Specifically, the confidence level a of each evidence bodyiAs a weight, weighting the initial evidence source according to the following formula to obtain a corrected evidence source;
And step S90, synthesizing the corrected evidence source by using a D-S synthesis rule to obtain a fault state identification result of multi-source decision fusion.
In the rotor misalignment state identification method based on D-S evidence fusion, the initial evidence source of the D-S evidence theory is obtained by executing the steps S10 to S20, the evidence source after D-S correction can be obtained by executing the steps S30 to S80, and then the corrected evidence source is fused and synthesized by using the D-S synthesis rule, so that the identification and diagnosis of the rotor misalignment state are realized, the problem of evidence conflict existing in the D-S evidence theory is solved, and the fault identification process is greatly optimized. Compared with the mode of carrying out rotor misalignment identification through a DBN + D-S, DBN + Pignistic improved D-S model in the prior art, the rotor misalignment state identification method based on D-S evidence fusion can improve the identification accuracy of the rotor misalignment state.
For the convenience of understanding and description of the implementation effect, the method for identifying the rotor misalignment state based on D-S evidence fusion is described in detail by the following embodiments.
The test was carried out on an aircraft engine dual-rotor vibration test platform manufactured by east diamond vibration testing apparatus of Suzhou, using Bruel, Denmark&The vibration test system collects vibration data of an out-of-alignment state. Fig. 3 shows a schematic structural view of the test stand rotor-support structure. For convenience of explanation, fig. 3 shows only the structure related to the embodiment of the present invention.
The test platform rotor is an inner and outer concentric nested double rotor, adopts a 5-pivot form, namely an inner rotor 110 adopts a 1-1-1 supporting form with the numbers of 1#, 2# and 3#, an outer rotor 120 adopts a 1-0-1 supporting form with the numbers of 4# and 5#, and is respectively driven by two servo motors, the highest rotating speed of the test platform rotor is 8000r/min, and the accurate control of the rotating speed and the steering can be realized.
On the above test platform, a plurality of acceleration sensors are respectively arranged in the vertical direction and the horizontal direction of the supporting seat of the left end inner rotor 110 and the supporting seat of the left end outer rotor 120, and are numbered as T1, T2, T3 and T4. In the experimental process, the simulation of the misalignment state of the rotor is realized by adding adjusting gaskets with different thicknesses at different supporting seats of the dual-rotor vibration test bed, and the experimental working condition description is shown in table 1. In the experiment, the rotating speed of the inner rotor 110 is 1500r/min, the rotating speed of the outer rotor 120 is 2400r/min, the steering is in a different direction, and the sampling frequency is 16384 Hz. The experiment has respectively gathered 7 vibration signals under centering state a plurality of.
Table 1 description of test conditions
The filtered and de-noised acceleration vibration data are equally divided into 400 groups, each group comprises 1024 data points, the 7 working condition signals comprise 2800 groups of data in total, and 4 sample sets with dimensions of 2800 x 1024 are formed. And respectively randomly selecting 320 groups from the samples corresponding to each working condition to form a training group sample, and forming the rest samples into a test group sample. And training corresponding classifiers DBN to be recorded as DBN1, DBN2, DBN3 and DBN4 by using the training group samples respectively. The DBN model adopts a classical 5-layer model, and the number of nodes is 1024-; the normalization method is a meanvar method, the maximum iteration number of the RBM is 100, the learning rate is 0.01, the momentum parameter is 0.6, and the initial identification results of the first 100 test group samples are shown in FIG. 4.
As can be seen from fig. 4, there are recognition-error samples of a single DBN model trained with a single sensor, and the recognition error samples of different DBN models are different, which results in unreliable recognition results. To avoid contingency, each classifier was run 100 times in duplicate and the preliminary fault condition identification results were averaged over the duplicate tests as shown in table 2. It can be seen from table 2 that a single DBN has a certain difference when performing fault diagnosis, and cannot meet the requirements of practical diagnostic applications.
TABLE 2 Primary Fault State identification error Rate
Therefore, the error rate of the primary fault state identification is used as the uncertainty of the evidence body, and the initial evidence sources are marked as R1, R2, R3 and R4 corresponding to the corresponding DBN numbers. Fusion diagnosis was performed for each initial evidence source using the improved D-S evidence theory presented herein. To analyze the fusion performance of the method herein, the fusion records of test sample No. 2, test sample No. 38, and test sample No. 85 of each DBN were selected as an example for illustration, as shown in table 3.
TABLE 3 fusion records
In the process of applying the D-S evidence theory, various mutually conflicting information often exists, so that the decision reasonability is objectified, for example, three test sample evidence sources in the table 3 have great difference, but the invention can accurately identify, because the invention measures the similarity degree between the evidence bodies by using mutual information, corrects the initial evidence source by weighting the similarity degree, solves the evidence conflict problem, can effectively solve the conflict problem among a plurality of evidence bodies, and can accurately perform fusion diagnosis.
In order to verify the effectiveness of the method, two fusion methods of DBN + D-S, DBN + Pignis t ic improved D-S are selected for comparison. The average error rate is shown in table 4.
TABLE 4 fusion diagnostic results
Comparing table 2 and table 4, it can be seen that the recognition rate of the out-of-alignment state after multi-source decision fusion diagnosis by the D-S evidence theory is significantly higher than that of any single DBN, because the D-S evidence combination method can handle information uncertainty caused by randomness and ambiguity lock of objective information, compared with the traditional reasoning method, the evidence theory adopts the concept of evidence to represent the uncertainty of information, and utilizes the combination rule to process uncertain information, which can obtain a maximum possibility from incomplete and inaccurate information, so that the recognition rate of the fusion diagnosis result is higher than that of a single classifier. From table 4, it can be seen that the recognition error rate of the misalignment state recognition of the invention is only 0.36%, which is better than other fusion models. The recognition error rate of the DBN + Pignistic improved D-S model is 0.91% higher than that of the DBN + Pignistic improved D-S model, but is 1.61% lower than that of the DBN + D-S model.
Therefore, compared with the application of the method for identifying the rotor misalignment state based on the improved D-S evidence fusion, the method for improving the D-S evidence combination method based on the improved D-S evidence fusion can better represent the conflict between the evidence bodies and solve the evidence conflict problem existing in the D-S evidence theory. Moreover, the rotor misalignment state identification method based on improved D-S evidence fusion is suitable for identification and diagnosis of the rotor misalignment state, greatly optimizes the fault identification process of the rotor misalignment state compared with other methods, and has lower identification error rate.
The technical features of the embodiments described above may be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the embodiments described above are not described, but should be considered as being within the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A rotor misalignment state identification method based on improved D-S evidence fusion is characterized by comprising the following steps:
acquiring acceleration vibration data of a rotor in an out-of-centering state;
acquiring an initial evidence source of a D-S evidence theory according to the acceleration vibration data;
According to mutual information measure between evidencesObtaining a similarity measure function between the evidence bodies;
obtaining an evidence body E according to a similarity measure functioniTotal similarity of (c);
acquiring a credible function of the evidence body according to the total similarity;
obtaining the credibility of each evidence body according to the credibility function;
correcting the initial evidence source according to the credibility of each evidence body to obtain a corrected evidence source;
and synthesizing the corrected evidence source by using a D-S synthesis rule to obtain a fault state identification result of multi-source decision fusion.
2. The method for identifying the rotor misalignment state based on the improved D-S evidence fusion as claimed in claim 1, wherein the step of obtaining the acceleration vibration data of the rotor misalignment state comprises the following steps:
and acquiring a plurality of acceleration vibration data of the rotor in the non-centering state by using a double-rotor vibration test experiment mode.
3. The method for identifying a rotor misalignment state based on improved D-S evidence fusion as claimed in claim 1, wherein the step of obtaining an initial evidence source of D-S evidence theory from the acceleration vibration data comprises the steps of:
obtaining a data sample according to the acceleration vibration data; the data samples comprise training group samples and testing group samples;
using the training set samples to each DBNiThe model is trained to obtain a trained DBNiA model;
using trained DBNsiAnd the model carries out fault identification on the test group samples to obtain an initial evidence source of the D-S evidence theory.
4. The improved D-S evidence fusion based rotor misalignment state identification method according to claim 3, wherein data samples are obtained from the acceleration vibration data; the data samples comprise training group samples and testing group samples, and the method comprises the following steps:
for the acceleration vibration data XiFiltering and de-noising to obtain the filtered and de-noised acceleration vibration data;
Equally dividing the acceleration vibration data subjected to filtering and denoising into l sections to obtain n groups of l sections of data samples Yi=[yi1,yi2,...,yil];
5. The method for rotor misalignment state recognition based on improved D-S evidence fusion of claim 4, wherein the training set samples are used for each DBNiThe model is trained to obtain a trained DBNiThe model comprises the following steps:
6. The method for rotor misalignment state recognition based on improved D-S evidence fusion as claimed in claim 5, characterized in that trained DBN is utilizediThe step that the model carries out fault identification on the test group samples to obtain an initial evidence source of the D-S evidence theory is as follows:
8. The method for rotor misalignment state recognition based on improved D-S evidence fusion as claimed in claim 7, wherein the evidence body E is obtained from a similarity measure functioniThe steps of total similarity of (1) are:
10. The method for identifying rotor misalignment state based on improved D-S evidence fusion according to claim 9, wherein the step of revising the initial evidence source according to the credibility of each evidence body to obtain a revised evidence source comprises:
the credibility a of each evidence bodyi=C(mi) As weights, and weights the initial evidence sources according to the following formula to obtain modifiedA source of evidence;
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110140695.3A CN113033600B (en) | 2021-02-02 | 2021-02-02 | Rotor misalignment state identification method based on improved D-S evidence fusion |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110140695.3A CN113033600B (en) | 2021-02-02 | 2021-02-02 | Rotor misalignment state identification method based on improved D-S evidence fusion |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113033600A true CN113033600A (en) | 2021-06-25 |
CN113033600B CN113033600B (en) | 2022-05-27 |
Family
ID=76459967
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110140695.3A Active CN113033600B (en) | 2021-02-02 | 2021-02-02 | Rotor misalignment state identification method based on improved D-S evidence fusion |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113033600B (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101071505A (en) * | 2007-06-18 | 2007-11-14 | 华中科技大学 | Multi likeness measure image registration method |
CN102589890A (en) * | 2012-03-01 | 2012-07-18 | 上海电力学院 | Integrated fault diagnostic method of steam turbine based on CPN (counter-propagation network) and D-S (dempster-shafer) evidences |
CN105675274A (en) * | 2016-01-07 | 2016-06-15 | 西安交通大学 | Time-domain parameter and D-S evidence theory-based rotor running state monitoring method |
CN110276303A (en) * | 2019-06-25 | 2019-09-24 | 湖南科技大学 | Rotor misalignment quantitative identification method based on VMD and DBN |
US20200387785A1 (en) * | 2019-06-05 | 2020-12-10 | Wuhan University | Power equipment fault detecting and positioning method of artificial intelligence inference fusion |
CN112101161A (en) * | 2020-09-04 | 2020-12-18 | 西安交通大学 | Evidence theory fault state identification method based on correlation coefficient distance and iteration improvement |
-
2021
- 2021-02-02 CN CN202110140695.3A patent/CN113033600B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101071505A (en) * | 2007-06-18 | 2007-11-14 | 华中科技大学 | Multi likeness measure image registration method |
CN102589890A (en) * | 2012-03-01 | 2012-07-18 | 上海电力学院 | Integrated fault diagnostic method of steam turbine based on CPN (counter-propagation network) and D-S (dempster-shafer) evidences |
CN105675274A (en) * | 2016-01-07 | 2016-06-15 | 西安交通大学 | Time-domain parameter and D-S evidence theory-based rotor running state monitoring method |
US20200387785A1 (en) * | 2019-06-05 | 2020-12-10 | Wuhan University | Power equipment fault detecting and positioning method of artificial intelligence inference fusion |
CN110276303A (en) * | 2019-06-25 | 2019-09-24 | 湖南科技大学 | Rotor misalignment quantitative identification method based on VMD and DBN |
CN112101161A (en) * | 2020-09-04 | 2020-12-18 | 西安交通大学 | Evidence theory fault state identification method based on correlation coefficient distance and iteration improvement |
Non-Patent Citations (2)
Title |
---|
A REVIEW OF MISALIGNMENT: "A review of misalignment", 《ACTA AERONAUTICA ET ASTRONAUTICA》 * |
张帆宇 等: "变分模态分解与深度信念网络的双转子不对中程度识别", 《机械科学与技术》 * |
Also Published As
Publication number | Publication date |
---|---|
CN113033600B (en) | 2022-05-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Nasiri et al. | Fracture mechanics and mechanical fault detection by artificial intelligence methods: A review | |
CN107657250B (en) | Bearing fault detection and positioning method and detection and positioning model implementation system and method | |
CN112161784B (en) | Mechanical fault diagnosis method based on multi-sensor information fusion migration network | |
US20200387785A1 (en) | Power equipment fault detecting and positioning method of artificial intelligence inference fusion | |
CN112183581A (en) | Semi-supervised mechanical fault diagnosis method based on self-adaptive migration neural network | |
CN110334764B (en) | Rotary machine intelligent fault diagnosis method based on integrated depth self-encoder | |
Li et al. | WavCapsNet: An interpretable intelligent compound fault diagnosis method by backward tracking | |
Di et al. | Ensemble deep transfer learning driven by multisensor signals for the fault diagnosis of bevel-gear cross-operation conditions | |
CN108256556A (en) | Wind-driven generator group wheel box method for diagnosing faults based on depth belief network | |
CN112326276B (en) | High-speed rail steering system fault detection LSTM method based on generation countermeasure network | |
CN111022313B (en) | Ocean platform air compressor fault diagnosis method based on LSTM | |
CN111637045B (en) | Fault diagnosis method for air compressor of ocean platform | |
CN113375941A (en) | Open set fault diagnosis method for high-speed motor train unit bearing | |
CN111523081A (en) | Aircraft engine fault diagnosis method based on enhanced gated cyclic neural network | |
Zhou et al. | Automated model generation for machinery fault diagnosis based on reinforcement learning and neural architecture search | |
CN111860839A (en) | Shore bridge fault monitoring method based on multi-signal fusion and Adam optimization algorithm | |
Shang et al. | Fault diagnosis method of rolling bearing based on deep belief network | |
Wang et al. | A remaining useful life prediction model based on hybrid long-short sequences for engines | |
CN114429152A (en) | Rolling bearing fault diagnosis method based on dynamic index antagonism self-adaption | |
CN116358871A (en) | Rolling bearing weak signal composite fault diagnosis method based on graph rolling network | |
Li et al. | Early gear pitting fault diagnosis based on bi-directional LSTM | |
Chou et al. | SHM data anomaly classification using machine learning strategies: A comparative study | |
CN117076935B (en) | Digital twin-assisted mechanical fault data lightweight generation method and system | |
CN113033600B (en) | Rotor misalignment state identification method based on improved D-S evidence fusion | |
CN112163630A (en) | Compound fault diagnosis method and device based on unbalanced learning |
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 |