CN112504808A - Aircraft thermal protection system damage diagnosis method based on machine learning algorithm - Google Patents
Aircraft thermal protection system damage diagnosis method based on machine learning algorithm Download PDFInfo
- Publication number
- CN112504808A CN112504808A CN202011201203.9A CN202011201203A CN112504808A CN 112504808 A CN112504808 A CN 112504808A CN 202011201203 A CN202011201203 A CN 202011201203A CN 112504808 A CN112504808 A CN 112504808A
- Authority
- CN
- China
- Prior art keywords
- protection system
- thermal protection
- damage
- aircraft
- neural network
- 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.)
- Pending
Links
- 230000006378 damage Effects 0.000 title claims abstract description 81
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000003745 diagnosis Methods 0.000 title claims abstract description 20
- 238000010801 machine learning Methods 0.000 title claims abstract description 10
- 238000013528 artificial neural network Methods 0.000 claims abstract description 36
- 238000012360 testing method Methods 0.000 claims abstract description 29
- 238000012549 training Methods 0.000 claims abstract description 27
- 239000013307 optical fiber Substances 0.000 claims abstract description 26
- 230000002159 abnormal effect Effects 0.000 claims abstract description 24
- 238000009413 insulation Methods 0.000 claims abstract description 11
- 230000008569 process Effects 0.000 claims abstract description 8
- 238000005259 measurement Methods 0.000 claims description 9
- 239000000835 fiber Substances 0.000 claims description 5
- 239000003292 glue Substances 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 238000002405 diagnostic procedure Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 3
- 230000002547 anomalous effect Effects 0.000 claims 1
- 238000013507 mapping Methods 0.000 abstract description 12
- 238000012545 processing Methods 0.000 abstract description 5
- 238000012423 maintenance Methods 0.000 abstract description 4
- 208000027418 Wounds and injury Diseases 0.000 abstract 1
- 208000014674 injury Diseases 0.000 abstract 1
- 238000001514 detection method Methods 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000005452 bending Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 230000003902 lesion Effects 0.000 description 3
- 238000003062 neural network model Methods 0.000 description 3
- 229910000838 Al alloy Inorganic materials 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- KXGFMDJXCMQABM-UHFFFAOYSA-N 2-methoxy-6-methylphenol Chemical compound [CH]OC1=CC=CC([CH])=C1O KXGFMDJXCMQABM-UHFFFAOYSA-N 0.000 description 1
- 239000004593 Epoxy Substances 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 239000000853 adhesive Substances 0.000 description 1
- 230000001070 adhesive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000032798 delamination Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 239000005011 phenolic resin Substances 0.000 description 1
- 229920001568 phenolic resin Polymers 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N3/00—Investigating strength properties of solid materials by application of mechanical stress
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N3/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N3/02—Details
- G01N3/06—Special adaptations of indicating or recording means
- G01N3/068—Special adaptations of indicating or recording means with optical indicating or recording means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/0058—Kind of property studied
- G01N2203/0069—Fatigue, creep, strain-stress relations or elastic constants
- G01N2203/0075—Strain-stress relations or elastic constants
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/02—Details not specific for a particular testing method
- G01N2203/0202—Control of the test
- G01N2203/0212—Theories, calculations
- G01N2203/0218—Calculations based on experimental data
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/02—Details not specific for a particular testing method
- G01N2203/06—Indicating or recording means; Sensing means
- G01N2203/0641—Indicating or recording means; Sensing means using optical, X-ray, ultraviolet, infrared or similar detectors
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/02—Details not specific for a particular testing method
- G01N2203/06—Indicating or recording means; Sensing means
- G01N2203/067—Parameter measured for estimating the property
- G01N2203/0682—Spatial dimension, e.g. length, area, angle
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Theoretical Computer Science (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses an aircraft thermal protection system damage diagnosis method based on a machine learning algorithm. Firstly, acquiring structural strain field distribution information by using an embedded optical fiber sensor network; then, positioning the position of the abnormal value by a method of combining polynomial fitting with a confidence interval, extracting the characteristics of the amplitude, the signal zero crossing rate, the signal fineness ratio and the like of the abnormal signal segment, and carrying out self-organizing mapping neural network training; after training is finished, test data are input into the self-organizing mapping neural network, and specific conditions such as the position and the type of the injury are obtained. In the flying process of the aircraft, the neural network can be trained according to real-time strain data reflected by factors such as service environment, special events, maintenance test and the like, the processing speed and the recognition precision of damage diagnosis are improved, and the problem of the integrity of the thermal insulation layer of the aircraft is solved.
Description
Technical Field
The invention belongs to the technical field of aerospace application, and particularly relates to an online identification method for damage of an aircraft thermal protection system based on a machine learning algorithm.
Background
The structural integrity of thermal protection systems, which are important components in aerospace vehicles to protect the safety of the overall structure, has become a critical issue in the development of reusable aerospace vehicle equipment. Compared with other structural components, the thermal protection system for the heat insulation layer bonded on the main body structure is more concealed from damage, so that the damage and failure of the thermal protection system are more sudden, and therefore, online identification, positioning and classification of the damage of the bonded heat insulation layer are particularly important.
In addition, in the service process of the aircraft, the heat insulation layer of the thermal protection system can generate cracks due to external impact or debond due to thermal mismatching, the two types of damage are clearly distinguished, and the method has important reference values for maintenance of the structure and decision making of a task envelope. Therefore, classification of these two types of lesions is an important issue and challenge for lesion diagnostic methods.
The traditional nondestructive detection method is used for identifying and positioning the damage of the structure in the shutdown state of the aircraft, external equipment needs to be assembled, data acquisition and manual analysis are needed, the time consumption is long, the detection cost is increased, and meanwhile, the damage type cannot be distinguished. The detection method for strain field reconstruction by using the embedded optical fiber sensor is widely applied to state evaluation of general structural members, and has the characteristics and advantages of high detection sensitivity and intuitive damage position information display, so that the existence of damage can be confirmed more efficiently and the general position of the damage can be positioned.
Disclosure of Invention
The invention aims to provide an aircraft thermal protection system damage diagnosis method based on a machine learning algorithm.
The technical scheme adopted by the invention for realizing the aim is as follows:
the invention provides an aircraft thermal protection system damage diagnosis method based on a machine learning algorithm, which comprises the following steps:
s1, performing physical model analysis according to the structural form and the typical loading form of the aircraft thermal protection system to obtain the strain distribution form of the healthy structure under the typical loading effect and the strain distribution change characteristics under the damage condition;
s2, designing the layout form of the optical fiber sensor in the thermal protection system according to the damage range obtained by physical model analysis and the minimum turning radius of the optical fiber;
s3, applying external load to the thermal protection system structure, and testing strain distribution data in the thermal protection system structure under different load states through the embedded optical fiber sensor;
s4, judging abnormal states according to the strain distribution data obtained by actual measurement, and detecting abnormal signals;
s5, randomly dividing abnormal signal data into a training data group and a testing data group, or taking test data before aircraft delivery as the training data group and taking the abnormal signal data as the testing data group; extracting abnormal signal features;
s6, inputting the training data set into the self-organizing map neural network to train the training data set, inputting the testing data set into the trained self-organizing map neural network to carry out validity check on the training data set, and if the testing data set is valid, the testing data set can be used for online damage diagnosis of the aircraft thermal protection system; if not, the process returns to step S5 to add the training data set and retrain the self-organizing map neural network.
Preferably, at least one fiber sensor is arranged in step S2, and the layout covers the damage range.
Preferably, in step S2, the optical fiber sensor is embedded in a glue layer between the thermal insulation layer and the main body structure in a revolving manner, and straight line segments of the optical fiber sensor are used as a measurement group.
Preferably, the step S4 locates the position of the outlier by a method of polynomial fitting in combination with the confidence interval.
Preferably, the polynomial fitting adopts straight line fitting, and the confidence interval of the fitting point is [0.25X,1.25X ].
Preferably, the abnormal signal characteristics include a signal peak-to-peak value, a signal zero crossing rate, a signal fineness ratio and a signal length.
Preferably, the method for diagnosing damage to the aircraft thermal protection system further comprises the step of continuously optimizing an algorithm by the self-organizing map neural network according to strain distribution data acquired by the optical fiber sensor in real time in the flight process of the aircraft.
Preferably, the damage diagnosis result is displayed through a display screen.
Compared with the prior art, the invention has the beneficial effects that:
compared with the traditional thermal protection layer ultrasonic or infrared nondestructive detection technology, the embedded optical fiber sensor network and the self-organizing map (SOM) neural network are applied to damage diagnosis of the aircraft thermal protection system, the online strain field data is directly adopted without stopping the service state of the aircraft, and the time and the detection cost are saved; compared with the traditional optical fiber online identification system, the method adopts the self-organizing mapping neural network technology, has high diagnosis speed and high learning efficiency, and can distinguish the structural damage types; the self-organizing mapping neural network has better operation performance and generalization capability, and can effectively solve the problems of hidden damage identification, positioning and classification by applying the self-organizing mapping neural network to the heat insulation layer damage feature extraction and pattern identification.
In the service process of the aircraft, the method can train the neural network according to real-time strain data reflected by factors such as service environment, special events, maintenance test and the like, can improve the processing speed and the recognition precision of damage diagnosis, and solves the problem of the integrity of the thermal insulation layer of the aircraft.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention. It is obvious that the drawings in the following description are only some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort.
FIG. 1 is a schematic diagram of the training of a typical SOM neural network;
FIG. 2 is a flow chart of a method for diagnosing damage to an aircraft thermal protection system based on a machine learning algorithm in accordance with the present invention;
FIG. 3 is a schematic diagram of a fiber sensor layout in an embodiment of the invention;
FIG. 4 is a graphical illustration of the results of using strain data to locate debond damage in an embodiment of the present invention;
FIG. 5 is a SOM neural network model trained using training data set data according to an embodiment of the present invention;
fig. 6 is a diagram illustrating the result of classifying the damage by using the trained SOM neural network model in the test data set according to the embodiment of the present invention.
Detailed Description
The following provides a detailed description of specific embodiments of the present invention. In the following description, for purposes of explanation and not limitation, specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details.
It should be noted that, in order to avoid obscuring the present invention with unnecessary details, only the device structures and/or processing steps that are closely related to the scheme according to the present invention are shown in the drawings, and other details that are not so relevant to the present invention are omitted.
According to the method for diagnosing the damage of the aircraft thermal protection system based on the machine learning algorithm, on one hand, the embedded optical fiber sensor network is used for collecting structural strain field distribution information, the embedded optical fiber sensor is high in detection sensitivity, damage position information is displayed visually, and the damage existence can be confirmed efficiently and the general position of the damage can be positioned. On the other hand, extracting the characteristics of the amplitude, the signal zero crossing rate, the signal fineness ratio and the like of the abnormal signal segment, and performing self-organizing map (SOM) neural network training, as shown in FIG. 1; the self-organizing map is an unsupervised artificial neural network, a competitive learning strategy is applied, the network is gradually optimized by means of mutual competition among neurons, and a topological structure of an input space is maintained by using a neighbor relation function; the self-organizing mapping neural network is integrated into a signal processing mechanism of a large number of human brain neurons and has unique structural characteristics, so that the self-organizing mapping neural network has good operational performance and generalization capability, and the problem of damage classification can be effectively solved by applying the self-organizing mapping neural network to the damage feature extraction and the mode identification of a thermal protection system.
The invention provides a method for diagnosing damage of an aircraft thermal protection system based on a machine learning algorithm, which comprises the following steps as shown in figure 2:
1. and carrying out physical model analysis according to the structural form and the typical loaded form of the thermal protection system of the aircraft to obtain the strain distribution form of the healthy structure under the typical loading effect and the strain distribution change characteristics under the damage condition.
2. And designing a layout form of the optical fiber sensor in the thermal protection system according to the damage range obtained by the physical model analysis result and the minimum turning radius of the optical fiber.
3. And applying external load to the thermal protection system structure, and testing strain distribution data in the thermal protection system structure under different load states through the embedded optical fiber sensor.
4. And judging abnormal states according to the internal strain distribution data of the structure obtained by actual measurement to obtain abnormal signals.
5. And randomly dividing abnormal signal data into a training data group and a testing data group, and extracting the characteristics of the abnormal signals.
6. Inputting the training data set into a self-organizing mapping neural network to train the training data set, inputting the testing data set into the trained self-organizing mapping neural network to carry out validity check on the training data set, and if the testing data set is valid, the testing data set can be used for online damage diagnosis of an aircraft thermal protection system; and if the result is invalid, returning to the step 5, adding the training data set, and retraining the self-organizing mapping neural network.
Typical loading patterns in the above methods include bending, tensile, torsional, and the like. The damage condition is set in a physical model of the thermal protection system.
At least one optical fiber sensor is arranged according to the damage range, and the damage range is covered by the layout.
And (4) positioning the position of the abnormal value by using a polynomial fitting method combined with the confidence interval, wherein the local maximum value and the local minimum value of the signal in the abnormal value are selected as an abnormal signal segment.
The abnormal signal characteristics comprise the characteristics of signal peak-to-peak value, signal zero crossing rate, signal fineness ratio and the like.
In addition, in some embodiments, the result of the damage determination is output to an aircraft cockpit display screen for display.
The above method is described in detail with reference to the accompanying drawings and a specific embodiment. The damage identification method of the typical thermal protection system structure comprises the following steps:
1. in this embodiment, aircraft thermal protection system selects phenolic resin as the heat insulation layer, and the aluminum alloy plate is as the major structure, and epoxy couples together both as the test piece as the adhesive glue film, and the glue film of test piece is inside to have arranged the debonding damage, and the crack damage has been arranged to the heat insulation layer.
2. As shown in fig. 3, a single optical fiber sensor is embedded into a glue layer between a heat insulation layer and a main body structure in a rotary mode and is close to the aluminum alloy main body structure, two ends of the optical fiber sensor extend out of a thermal protection system to serve as an optical signal inlet and an optical signal outlet, six straight line segments in the rotary mode serve as measurement groups 1-6, the measurement groups cover a damaged area, two ends of each straight line segment serve as measurement nodes, and 12 measurement nodes are provided in total.
3. And analyzing the strain distribution form of the structural health state of the thermal protection system under the action of a typical load and the strain distribution rule of a typical damage under a loaded condition by adopting a physical model. Taking bending load as an example, the distribution of the internal strain field of the healthy structure of the thermal protection system is in a monotone decreasing mode; typical damage includes cracks and debonding, where the strain at the crack exhibits an "M" type profile, with two peaks corresponding to the two edges of the crack, the strain at the debonding damage exhibits an "N" type profile, and maxima and minima corresponding to the two edges of the debonding.
4. And applying bending load, and testing structural strain distribution data under different load states through the embedded optical fiber sensor.
5. Fitting the actually measured strain distribution data by utilizing a polynomial fitting method, and selecting a straight line for fitting because the analysis result of the physical model shows that the strain field in the healthy structure has the monotonous descending trend.
6. The [0.25X,1.25X ] of each fitted point is classified as a "confidence zone", and points outside the confidence zone are selected as outliers, which correspond to the location of the suspected lesion.
7. The local maximum and the local minimum of the signal among the outliers are selected as an outlier signal segment. The strain distribution rule of the abnormal signal segments can be used as the basis for damage diagnosis.
As shown in fig. 4, the horizontal axis represents strain data obtained along the length direction of a certain section of the optical fiber sensor, and the debonding damage is located by using the strain data, and the deviation between the debonding left and right boundaries and the actual debonding damage left and right boundaries of the diagnosis result is +1mm, +2mm, and the deviation of the width dimension is +1mm, respectively.
8. And detecting end points in the abnormal signal section according to the strain distribution change characteristics under the damage condition obtained by the physical model analysis, and extracting a signal characteristic vector. The signal feature vector comprises a signal peak-to-peak value, a fineness ratio, a zero-crossing rate and a signal length.
9. The data are divided into two groups, namely a training data group and a testing data group, the SOM neural network model is trained by using the training data group, as shown in figure 5, it can be seen that the debonding damage and the crack damage are divided into two areas in the self-organization weight neighborhood distance map.
10. The trained SOM neural network is used to diagnose damage to the test data set, and as shown in fig. 6, the results show that all debonding damage and crack damage are distributed into two regions in the self-organization weight neighborhood distance map, and the damage diagnosis accuracy is 100%.
The SOM neural network output data result shows that the SOM neural network can distinguish the positions and the types of different injuries in the neural network training and verifying process, and the distinguishing accuracy of the two injuries, namely delamination and crack, is 100%. The SOM neural network can accurately judge the damage state of the thermal protection system of the aircraft, including the damage position, the influence range and the type.
The training data set required by the training SOM neural network can be measured by pressurization, static force, wind tunnel and other tests before the aircraft leaves a factory and is input before delivery. In addition, a trained SOM neural network can be directly introduced into an airborne processing system of the aircraft, and then in the flying process of the aircraft, according to the environmental condition, the task envelope and the structure maintenance condition, the SOM neural network continuously optimizes the algorithm according to the strain distribution data acquired by the optical fiber sensor in real time, so that the accuracy is improved.
Compared with the traditional thermal protection layer ultrasonic or infrared nondestructive detection technology, the method does not need to stop the service state of the aircraft, but directly adopts on-line strain field data, thereby saving time and detection cost; compared with the traditional optical fiber online identification system, the method adopts the self-organizing mapping neural network technology, has high diagnosis speed and high learning efficiency, and can distinguish the structural damage types; the SOM neural network has better operational performance and generalization capability, and can effectively solve the problems of identification, positioning and classification of hidden damage by applying the SOM neural network to the damage feature extraction and pattern recognition of a thermal protection system.
Features that are described and/or illustrated above with respect to one embodiment may be used in the same way or in a similar way in one or more other embodiments and/or in combination with or instead of the features of the other embodiments.
It should be emphasized that the term "comprises/comprising" when used herein, is taken to specify the presence of stated features, integers, steps or components but does not preclude the presence or addition of one or more other features, integers, steps, components or groups thereof.
The many features and advantages of these embodiments are apparent from the detailed specification, and thus, it is intended by the appended claims to cover all such features and advantages of these embodiments which fall within the true spirit and scope thereof. Further, since numerous modifications and changes will readily occur to those skilled in the art, it is not desired to limit the embodiments of the invention to the exact construction and operation illustrated and described, and accordingly, all suitable modifications and equivalents may be resorted to, falling within the scope thereof.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
The invention has not been described in detail and is in part known to those of skill in the art.
Claims (8)
1. A method for diagnosing damage of an aircraft thermal protection system based on a machine learning algorithm is characterized by comprising the following steps:
s1, performing physical model analysis according to the structural form and the typical loading form of the aircraft thermal protection system to obtain the strain distribution form of the healthy structure under the typical loading effect and the strain distribution change characteristics under the damage condition;
s2, designing the layout form of the optical fiber sensor in the thermal protection system according to the damage range obtained by physical model analysis and the minimum turning radius of the optical fiber;
s3, applying external load to the thermal protection system structure, and testing strain distribution data in the thermal protection system structure under different load states through the embedded optical fiber sensor;
s4, judging abnormal states according to the strain distribution data obtained by actual measurement, and detecting abnormal signals;
s5, randomly dividing abnormal signal data into a training data group and a testing data group, or taking test data before aircraft delivery as the training data group and taking the abnormal signal data as the testing data group; extracting abnormal signal features;
s6, inputting the training data set into the self-organizing map neural network to train the training data set, inputting the testing data set into the trained self-organizing map neural network to carry out validity check on the training data set, and if the testing data set is valid, the testing data set can be used for online damage diagnosis of the aircraft thermal protection system; if not, the process returns to step S5 to add the training data set and retrain the self-organizing map neural network.
2. The aircraft thermal protection system damage diagnosis method according to claim 1, wherein at least one fiber sensor is provided in step S2, and the layout covers the damage range.
3. The aircraft thermal protection system damage diagnosis method according to claim 2, wherein in step S2, the fiber optic sensor is embedded in a glue layer between the thermal insulation layer and the main body structure in a rotary manner, and straight line segments of the fiber optic sensor are used as a measurement group.
4. The method for diagnosing damage to an aircraft thermal protection system according to claim 1, wherein the step S4 is performed by a method of combining a polynomial fit with a confidence interval to locate the position of the outlier.
5. The aircraft thermal protection system damage diagnostic method of claim 4, wherein the polynomial fit employs a straight line fit with a confidence interval of the fit points of [0.25X,1.25X ].
6. The aircraft thermal protection system damage diagnostic method of claim 1, wherein the anomalous signal characteristics include signal peak-to-peak, signal zero crossing rate, signal fineness ratio, signal length.
7. The aircraft thermal protection system damage diagnosis method according to claim 1, further comprising a step of continuously optimizing an algorithm by the self-organizing map neural network according to strain distribution data collected by the optical fiber sensor in real time during the flight of the aircraft.
8. The aircraft thermal protection system damage diagnostic method of claim 1, wherein the damage diagnostic result is displayed via a display screen.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011201203.9A CN112504808A (en) | 2020-11-02 | 2020-11-02 | Aircraft thermal protection system damage diagnosis method based on machine learning algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011201203.9A CN112504808A (en) | 2020-11-02 | 2020-11-02 | Aircraft thermal protection system damage diagnosis method based on machine learning algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112504808A true CN112504808A (en) | 2021-03-16 |
Family
ID=74954823
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011201203.9A Pending CN112504808A (en) | 2020-11-02 | 2020-11-02 | Aircraft thermal protection system damage diagnosis method based on machine learning algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112504808A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023014298A3 (en) * | 2021-08-06 | 2023-04-13 | 脸萌有限公司 | Neural network construction method and apparatus |
| CN117347378A (en) * | 2023-08-18 | 2024-01-05 | 大连理工大学 | Composite material damage identification method based on optical fiber measurement and neural network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101544064A (en) * | 2009-05-08 | 2009-09-30 | 南京航空航天大学 | Intelligent structure self-healing method and health monitoring system based on light repairing technology |
| CN104200004A (en) * | 2014-07-28 | 2014-12-10 | 东北大学 | Optimized bridge damage identification method based on neural network |
| CN104200005A (en) * | 2014-07-28 | 2014-12-10 | 东北大学 | Bridge damage identification method based on neural network |
| CN111024821A (en) * | 2019-12-30 | 2020-04-17 | 大连理工大学 | Composite material storage box health monitoring system and method |
-
2020
- 2020-11-02 CN CN202011201203.9A patent/CN112504808A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101544064A (en) * | 2009-05-08 | 2009-09-30 | 南京航空航天大学 | Intelligent structure self-healing method and health monitoring system based on light repairing technology |
| CN104200004A (en) * | 2014-07-28 | 2014-12-10 | 东北大学 | Optimized bridge damage identification method based on neural network |
| CN104200005A (en) * | 2014-07-28 | 2014-12-10 | 东北大学 | Bridge damage identification method based on neural network |
| CN111024821A (en) * | 2019-12-30 | 2020-04-17 | 大连理工大学 | Composite material storage box health monitoring system and method |
Non-Patent Citations (4)
| Title |
|---|
| 易鸿: "基于神经网络对光纤智能结构的损伤评估", 《四川理工学院学报(自然科学版)》 * |
| 袁颖 等: "《结构损伤识别理论及其应用》", 31 December 2008, 中国大地出版社 * |
| 郁可 等: "光纤智能材料与结构损伤估计的神经网络处理", 《山东建材学院学报》 * |
| 郭林峰: "状态实时监控与损伤快速修复的光纤智能结构研究", 《中国优秀博硕士学位论文全文数据库(博士) 信息科技辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023014298A3 (en) * | 2021-08-06 | 2023-04-13 | 脸萌有限公司 | Neural network construction method and apparatus |
| CN117347378A (en) * | 2023-08-18 | 2024-01-05 | 大连理工大学 | Composite material damage identification method based on optical fiber measurement and neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106006344B (en) | Staircase On-line Fault early warning system and method for diagnosing faults | |
| CN112504808A (en) | Aircraft thermal protection system damage diagnosis method based on machine learning algorithm | |
| CN104897403B (en) | Self-adaption fault diagnosis method based on permutation entropy (PE) and manifold-based dynamic time warping (MDTW) | |
| Lorenzo et al. | Damage detection in wind turbine blades by using operational modal analysis | |
| CN104833534A (en) | Train running fault diagnosis device based on multi-source information fusion, and method | |
| CN108932711B (en) | Image analysis neural network system | |
| CN102175449B (en) | Blade fault diagnostic method based on strain energy response of wind-driven generator | |
| Sun et al. | In-situ condition monitoring of wind turbine blades: A critical and systematic review of techniques, challenges, and futures | |
| CN103310438B (en) | Based on the contact net rod insulator fault detection method of Harris angle point and image difference | |
| CN108092623B (en) | A kind of photovoltaic array multisensor fault detecting and positioning method | |
| CN109187060B (en) | The detection of train speed sensor abnormal signal and axis locking method for diagnosing faults | |
| CN108647642A (en) | Multisensor Crack Damage error comprehensive diagnosis method based on fuzzy Fusion | |
| CN108760305A (en) | A kind of Bearing Fault Detection Method, device and equipment | |
| CN108981796A (en) | A kind of five in one hydraulic method for diagnosing faults | |
| CN110704911A (en) | Bridge damage cross positioning method based on cluster structure similarity | |
| CN114755017B (en) | Variable-speed bearing fault diagnosis method of cross-domain data driving unsupervised field shared network | |
| CN109681391A (en) | A kind of blade root bolt fracture fault detection method and medium | |
| Mujica et al. | Impact damage detection in aircraft composites using knowledge-based reasoning | |
| CN108267312A (en) | A kind of subway train bearing intelligent diagnostic method based on fast search algorithm | |
| CN107543581A (en) | Multi-functional substation framework health monitoring and damnification recognition method | |
| Wu et al. | Blade crack detection using blade tip timing | |
| Zhang et al. | Data anomaly detection for bridge SHM based on CNN combined with statistic features | |
| Junior et al. | Unbalance evaluation of a scaled wind turbine under different rotational regimes via detrended fluctuation analysis of vibration signals combined with pattern recognition techniques | |
| Wang et al. | EEMD-based videogrammetry and vibration analysis method for rotating wind power blades | |
| CN110210326A (en) | A kind of identification of online train and speed estimation method based on fiber-optic vibration signal |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20210316 |