CN115374960B - Method and system for managing aeroengine health - Google Patents

Method and system for managing aeroengine health Download PDF

Info

Publication number
CN115374960B
CN115374960B CN202210785464.2A CN202210785464A CN115374960B CN 115374960 B CN115374960 B CN 115374960B CN 202210785464 A CN202210785464 A CN 202210785464A CN 115374960 B CN115374960 B CN 115374960B
Authority
CN
China
Prior art keywords
data
engine
analysis
map
aeroengine
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202210785464.2A
Other languages
Chinese (zh)
Other versions
CN115374960A (en
Inventor
焦广胜
谢河金
康立
齐涛
李蒙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Meixin Testing Technology Co ltd
Original Assignee
Meixin Testing Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Meixin Testing Technology Co ltd filed Critical Meixin Testing Technology Co ltd
Priority to CN202210785464.2A priority Critical patent/CN115374960B/en
Publication of CN115374960A publication Critical patent/CN115374960A/en
Application granted granted Critical
Publication of CN115374960B publication Critical patent/CN115374960B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/20—Administration of product repair or maintenance
    • 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
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/04—Manufacturing

Abstract

The invention relates to a method and a system for managing the health of an aeroengine, which mainly utilize a mobile trolley provided with a dot matrix spectrometer carried by a multi-axis robot to carry out high-efficiency nondestructive detection on the aeroengine to be overhauled, so that on the basis of inputting related information of the aeroengine and experimental data, a convolutional neural network CNN (computer numerical network) is modeled by utilizing an automatic grabbing map analysis data result to obtain the number of possible unexpected sensitive material positions of the current engine to be tested, thereby judging the health condition of the current engine, and simultaneously realizing the organic organization and management of big data and providing a large amount of data for the safe flight of an airplane.

Description

Method and system for managing aeroengine health
Technical Field
The invention relates to a method and a system for managing the health of an aeroengine, in particular to a method and a system for capturing data and analyzing the health condition of the engine by applying Python.
Background
The problems of reliability, availability, economy, maintenance guarantee and the like of the current large-scale complex system are increasingly outstanding, and the traditional periodic maintenance mode is low in resource consumption and efficiency and high in cost. Therefore, on the premise of ensuring aviation safety, each large airline driver takes 'economic affordability' as a primary task without exception. The on-line maintenance has the remarkable advantages of small scale, high efficiency, good economic affordability, capability of avoiding serious disaster accidents and the like, and is very suitable for maintenance and guarantee of large-scale complex systems. One of the preconditions for performing an on-demand repair is the requirement that the fault be predictable.
The present age is an informatization age, so databases are also becoming more and more important for various industries. In the aeroengine health management system, a user can search and check corresponding detection data according to relevant information (such as serial numbers, registration numbers, airplane models and the like) of the engine, wherein the corresponding detection data comprises the information of chip morphology, size, components, approximate marks and the like, and can check past detection data of the engine, so that the health condition of the engine can be systematically analyzed and corresponding treatment measures can be made.
At present, only experimental data of an engine is generated into a corresponding detection report, and the problems of the corresponding engine are confirmed through the detection report. The health condition of the engine cannot be predicted in advance, meanwhile, because the aeroengine is huge, the parts to be monitored are also more, the number of damaged, scrapped, identified or suggested engines which cannot be served any more based on the damage detection is limited, and huge effective modeling data volume cannot be formed.
Disclosure of Invention
Aiming at the current situation, the invention provides a nondestructive analysis method for aeroengine data, which utilizes the maintenance period to realize the collection of general data, introduces accumulated engine detection data into a system, utilizes the data analysis to find out the rule therein, and can better predict and manage the engine health condition.
Based on the above considerations, the present invention provides a method for managing the health of an aeroengine, wherein the method uses MySQL database to collect the detection data of the aeroengine during maintenance, uses the program developed by Python to analyze and predict the stored detection data, and specifically comprises the following steps:
s1, inputting relevant information of an aeroengine, wherein the relevant information mainly comprises the following steps of: engine model (EngineType), engine number (ESN), aircraft model (AircraftType), aircraft registration number (AircraftReg), engine time since new use (ETSN), engine cycle since new use (ECSN), engine time since last repair (ETSR), engine cycle since last repair (ECSR), sampling location i, and selecting pairs againSample information: debris (Debris), oil filtration (OilFilter), oil (Oilsample). The data are stored in an engine information table T in a MySQL database i Is a kind of medium.
For the case of a plurality of sampling positions, the sampling Position information is stored in a Position field in a JSON form, and a sampling Position i and an engine information table T are established i First association mapping i-T between i The method comprises the steps of carrying out a first treatment on the surface of the The sampling locations i may be location numbers, location coordinates, and other identification codes or pattern marks that may be used to identify different sampling locations, etc.
S2, after the step S1 is completed, inputting experimental data, wherein the method mainly comprises the following steps: sample size and appearance photo, spectrum analysis photo and spectrum analysis data, and packaging the experimental data to form a data packet B i Numbering is carried outEstablishing a second association map +_with said engine information table in step S1>Thus, the sample position i can be correlated to the test data packing number +.>I.e. form a first composite associative mapping +.>
Thus, not only the engine information table but also the number can be obtained by the sampling position iExperimental data were obtained.
S3, inputting the map analysis data in the step S2, automatically grabbing the map analysis data by using a Python program, sequencing the names of various elements according to a preset sequence, calculating main element components and secondary element components, and endowing various elements with different color values to representDifferent contents to form a matching chart P i I.e. package numberingA third association map is formed with the matching graph>The materials in the element component comparison table are subjected to element sorting according to the same sequence and are endowed with specified content limit corresponding color values to form a limit map P j J represents a j-th Alloy material (Alloy) in the element composition comparison table;
preferably, the color value is
S4 at step S3, a matching pattern P is formed i The system then uses the Python program to match the map P i And limit map P j Performing a difference operation P i -P j The method comprises the steps of carrying out a first treatment on the surface of the If the differential result is a blank graph or the sum L of color values of all elements is in a specified range, then the matched result is indicated to determine the type of alloy material, otherwise, the alloy material cannot be matched, and the alloy material is indicated to be unknown, and the sampling position i is indicated to be normal or needs to be further detected;
wherein the specified range is specified as follows:
k is the sequential number of the preset sequence of elements, N is the total seed number of the elements, R k 、G k 、B k An allowable maximum deviation value for the three color components of element k;
s5, after the steps are finished, the MySQL database continuously stores relevant experimental data of each model of engine, and the system grabs the experimental data and the quantity A j If the sensitive material is found and the quantity exceeds the limit value, the value is displayed in red, and the abnormal sample is regarded; if the sensitive material is found but the quantity is not beyond the limit, the material is displayed in yellow and is regarded as suspicious to be noticedA sample; if no sensitive material is found and the number does not exceed the limit, displaying green, regarding as a normal sample, the number being the total number of points in different sampling positions within the specified range;
s6, acquiring parameters EngineType, aircraftType, ETSN, ECSN, ETSR of different positions on a plurality of engines, and also endowing different color values of each parameter to represent different parameter values and match the parameters with a map P i Splicing to form predictive map p i The quantity of the corresponding materials at different positions is obtained, a prediction graph is divided into a training set and a verification set, the ratio of the training set to the verification set is 5:1-1:1, a convolutional neural network model CNN is established, the quantity of the materials is formed by outputting the quantity to a classification function softmax or sigmond through a full connection layer FC, and a quantity prediction result A is formed according to the rule j And display color C j Fourth association mapping A between j →C j Inputting parameters of different positions to be detected and a predictive graph to be detected formed by corresponding matching patterns into a trained CNN, and obtaining C through the fourth association mapping j Then a second composite mapping relation is formed
Therefore, the display color C can be obtained by using the CNN model and the fourth association mapping by using the formed matching diagram and the prediction diagram only by establishing the first composite mapping j 。
Preferably, the spectrum analysis adopts a lattice spectrometer for detection analysis, the lattice spectrometer comprises a probe consisting of a multichannel spectrum tester, a data processing module, a data analysis module and a WiFi system or a Bluetooth system for data transmission, the probe comprises a lattice formed by forming a pair of points by a light source and a detector, the light source in each point transmits light signals through an optical fiber, the emitting end of the optical fiber connected with the light source is provided with a collimator, the data processing module comprises a multichannel filter-photoelectric converter-pre-amplification group, a channel formed by each filter-photoelectric converter-pre-amplification group corresponds to the detector in the point to collect the spectroscopic light signals, and the data analysis module is used for carrying out intensity analysis on the collected pre-amplification signals to form spectrum analysis data, wherein the detector is a dispersion system.
The further detection comprises comprehensive analysis by adopting a scanning electron microscope SEM with an EDS spectrometer, a transmission electron microscope TEM with the EDS spectrometer, X-ray photoelectron spectroscopy XPS and X-ray diffraction XRD.
Preferably, the condition requiring further analysis is that the length of engine service exceeds 1/3 or 1/4 of the theoretical total expected service length or expected life.
Preferably, a dustproof protection cover is arranged outside the lattice spectrometer, the dustproof protection cover is provided with an automatic opening and closing door, and is in a closed state when the detection analysis is not performed, and otherwise, the probe is opened and exposed.
Preferably, the detection analysis is to locate the lattice spectrometer near the engine, and the motion controller controls the distance of the lattice spectrometer to be close to the surface of the engine, so as to detect and acquire the analysis data of the material on the engine.
Preferably, the motion controller comprises a cart provided with a multidirectional wheel with a pneumatic tire, a multi-axis robot arranged on a cart plate, wherein the limb end of the multi-axis robot is connected with the lattice spectrometer to realize multi-dimensional motion, and a power supply system for supplying power to the multi-axis robot and the lattice spectrometer is also arranged on the cart plate.
Preferably, the trolley is further equipped with a lifting system for enabling the required height of the lattice spectrometer arrangement when scanning the engine roof.
Preferably, the cart is provided with a control panel programmed for implementing the method of steps S1-S6 and programmed for the movements of the multi-axis robot and the lifting system.
Another aspect of the invention provides a system for aeroengine data storage and health management comprising a server, a MySQL database, a cart equipped with a lifting system and carrying a multi-axis robot, the ends of the limbs of the multi-axis robot being connected to a lattice spectrometer by means of joints on the lattice spectrometer, wherein,
the server is used for inputting relevant information and experimental data of the aeroengine and realizing the steps (2) -S6 in the method;
the multi-directional wheels with pneumatic tires are arranged at the bottom of a carriage plate or on a lifting system of the carriage, a control panel for programming the method in the steps S1-S6 and programming the motion of the multi-axis robot is arranged on the carriage, and a power supply system for supplying power to the multi-axis robot and the lattice spectrometer is arranged on the carriage;
and the trolley realizes data transmission between the server through a WiFi system or a Bluetooth system in the lattice spectrometer, and the server is communicated with the MySQL database.
Preferably, the control panel comprises two display screens respectively corresponding to a programming area and a main screen, the program file icons in the programming area can be manually dragged to the main screen to realize the program file operation in the main screen for program debugging to perform motion debugging of the multi-axis robot, the documents displayed in the main screen can also be manually dragged to the programming area for programming, the main screen is used for spectrum detection parameter adjustment, the display, the storage, the deletion and the restoration of spectrum data, the copy and paste of spectrum data and spectrum detection parameters, the input, the storage, the deletion and the restoration of spectrum data, the copy and paste, the processing analysis, the transmission, the debugging of the program, the adjustment and the control of motion parameters of the multi-axis robot and a lifting system and the power management,
the programming comprises programming for realizing the method, programming for realizing the steps S1-S6, programming for the movement of the multi-axis robot and the lifting system, transmitting a program file to the server through a programming area and/or a main screen by utilizing a WiFi system or a Bluetooth system, further modifying the program file in the server to simulate debugging, transmitting the program file back to the cart to debug and confirm in the main screen, continuing to debug in the programming area until reaching the standard if the program file still does not reach the standard, and/or transmitting a command again to enable the server to continue the re-debugging simulation, and transmitting the program file back to the cart to debug and confirm again until reaching the standard.
The invention also provides a computer readable non-transitory storage medium in which a program is stored that can be run by the server and that assists the control panel in implementing the above method of managing aeroengine health.
Advantageous effects
1. Related information, experimental data and health prediction of the aero-engine are organically organized and managed by utilizing the association mapping;
2. the method comprises the steps of fully utilizing engine experimental data to establish an aeroengine health prediction model, and predicting the health status of the aeroengine so as to guide overhaul;
3. the multipoint nondestructive detection of the materials of a plurality of interested parts of the engine is efficiently completed by adopting the dot matrix spectrometer, so that the formation and the predicted development situation of the sensitive materials are known, and a large amount of data are provided for the safe flight of the aircraft.
Drawings
Fig. 1 shows a schematic diagram for implementing information entry of an aeroengine according to embodiment 1 of the present invention, in which the number of sampling positions is not fixed and may be increased or decreased according to practical situations.
Fig. 2 shows a schematic diagram for realizing experimental data input of an aeroengine according to embodiment 2 of the present invention, the number of sample appearances can be increased or decreased according to actual situations, and a final data result system is obtained by automatically grabbing and analyzing the input data.
Fig. 3 shows a schematic diagram for analyzing the health condition of an aeroengine according to embodiment 3 of the present invention, wherein the analysis and statistics are performed on the health condition of the engine according to experimental data stored in a database, and the health condition reminding of the engine is performed with different colors.
Fig. 4 shows a detailed flowchart of a method for managing the health of an aircraft engine according to embodiment 4 of the present invention.
Fig. 5 shows a schematic structural diagram of a 220×220 pixel prediction map according to embodiment 4 of the present invention.
Fig. 6 shows a schematic diagram of a lattice spectrometer structure in embodiment 5 of the present invention, where a is a front view of a 4×4 probe, b and c are the closed and open states of the front end automatic door of the dust-proof shield, and d is the overall structure of the 4×4 probe, the data processing module, the data analysis module, and the WiFi system for data transmission.
Fig. 7 shows a schematic view of a situation of the cart in the lowest (a) and highest (b) positions of the lifting system in the analysis of a map of a certain position on the side of an aeroengine to be overhauled, wherein the cart is provided with a multidirectional wheel, a control panel and a multi-axis robot and a power supply system arranged on a cart board.
Fig. 8 shows a layout structure of an aeroengine data storage and health management prediction system and a data information interaction schematic diagram in embodiment 6 of the present invention.
Fig. 9 is a schematic diagram showing the structure of a control panel in embodiment 6 of the present invention.
Wherein the reference numerals: the device comprises a dot matrix spectrometer 1, a probe 2, a probe-containing channel 2-1, a polychromator-containing channel 2-2, a 2' rotating shaft, a 3 rotating shaft, a 4 rotating seat, a 5 base, 6 optical fibers, a 7 collimating lens and an 8 polychromator.
Detailed Description
Example 1
The embodiment 1 specifically provides an aeroengine information input mode, and specifically as shown in fig. 1, the following information is input: engine model, engine number, aircraft model, aircraft registration number, sampling date, engine time of use, engine cycle of use, engine time of use since last repair, engine cycle of use since last repair, sampling location, remarks, and corresponding sample types (chips/oil filters/oil products) are selected, and the data is stored in MySQL database after being submitted.
Example 2
In the embodiment 2, the method for inputting experimental data of the aero-engine is specifically provided, and as shown in fig. 2, the appearance photo of the sample is uploaded and can be freely added or subtracted; and uploading the corresponding map analysis data map, and freely adding or deleting. At this time, the background Python program identifies and grabs corresponding element data according to the uploaded map analysis result diagram, sorts the element contents, and automatically adds the sorted result into the main element and the secondary element. And finally, matching the primary element, the secondary element and the element comparison table by the system so as to analyze the corresponding material type.
Example 3
In this embodiment 3, the data obtained in embodiments 1 and 2 are used to perform an aggregate query from a database, perform a data analysis on the query result, match the corresponding result according to the rule of table 1, and show the abnormal health condition of the aeroengine in the form of a graph.
Example 4
Fig. 4 is a detailed flowchart of a method for managing health of an aeroengine according to embodiments 1-3 of the present invention, including:
s1, inputting relevant information of the aeroengine, wherein the relevant information mainly comprises the information listed in the embodiment 1. The data are stored in an engine information table T in a MySQL database i Is a kind of medium. For the case of a plurality of sampling positions, the sampling Position information is stored in a Position field in a JSON form, and a sampling Position i and an engine information table T are established i First association mapping i-T between i 。
S2 after the step S1 is completed, inputting experimental data mainly comprising the data of the embodiment 2, and packaging the experimental data to form a data packet B i Numbering is carried outEstablishing a second association map with the engine information table in step S1Thus, the sample position i can be correlated to the test data packing number +.>I.e. form a first composite associative mapping +.>
S3, inputting the map analysis data in the step S2, and simultaneously, sequencing the names of various elements according to a preset sequence by using a system by utilizing the result of automatically grabbing the map analysis data, calculating the main element components and the secondary element components, and simultaneously, endowing various elements with different color values to represent different contents to form a matching map P i I.e. package numberingForm a third association mapping N with the matching graph Bi →P i The materials in the element component comparison table are subjected to element sorting according to the same sequence and are endowed with specified content limit corresponding color values to form a limit map P j J represents the j-th material (Alloy) in the element composition comparison table, the color value +.>The preset sequence may be a content sequence, and the specific sequence may be a sequence from right to left and then from top to bottom according to the periodic table sequence of the elements.
S4 at step S3, a matching pattern P is formed i The system then uses the Python program to match the map P i And limit map P j Performing a difference operation P i -P j The method comprises the steps of carrying out a first treatment on the surface of the If the difference result is a blank graph or the sum L of color values of all elements is in a specified range, the matching is indicated to determine the material type, otherwise, the matching is indicated as unknown, and the sampling position i is normal or needs to be further detected;
wherein the specified range is specified as follows:
k is the sequential number of the preset sequence of elements, N is the total seed number of the elements, R k 、G k 、B k An allowable maximum deviation value for the three color components of element k; default value is R k ＝G k ＝B k ＝10。
The further detection comprises comprehensive analysis by adopting a scanning electron microscope SEM with an EDS spectrometer, a transmission electron microscope TEM with an EDS spectrometer, X-ray photoelectron spectroscopy XPS and X-ray diffraction XRD
The condition to be further analyzed is that the service length of the engine exceeds 1/4 of the theoretical expected total service length or expected service life.
S5, after the steps are finished, continuously storing relevant experimental data of each model of engine by the database, and capturing data materials and quantity A by the system j Comparing the rules of table 1, if the sensitive material is found and the quantity exceeds the limit value, displaying in red, and treating as an abnormal sample; if the sensitive material is found but the quantity is not beyond the limit value, the sample is displayed in yellow and is regarded as suspicious to be noticed; if no sensitive material is found and the number does not exceed the limit, displaying green, regarding as a normal sample, the number being the total number of points in different sampling positions within the specified range;
TABLE 1
The M50 steel is aviation bearing steel with excellent comprehensive performance at present, and the M50Nil steel is a new generation of high-temperature bearing steel developed on the basis of the M50 steel. AMS6265, 3.25% nicrmo case hardened steel. The AISI9310 steel is a high strength steel with good quality and low alloy steel. The 17-4PH alloy is a precipitation hardening martensitic stainless steel composed of copper, niobium/columbium, and the 17-4PH stainless steel is a martensitic precipitation hardening stainless steel. 15-5PH is martensitic, precipitation-hardened, chromium nickel copper stainless steel.
S6, acquiring parameters EngineType, aircraftType, ETSN, ECSN, ETSR of different positions on a plurality of engines, and also endowing different color values of each parameter to represent different parameter values to form pixel columns of five colors and a matching chart P i SplicingForming a predictive map p i 。
A 5 x 5 prediction graph of one test point is shown in fig. 4. In practice 44×44 test points, a 220×220 pixel prediction map is formed, as shown in fig. 5.
Then, obtaining the corresponding quantity value of each material, dividing the side view of the 220×220 pixels of 44×44 test points on each different engine into a training set and a verification set, wherein the ratio of the training set to the verification set is 3:1, establishing a convolutional neural network model CNN, outputting the convolutional neural network model to a classification function softmax through a full connection layer FC to form a quantity prediction result A according to table 1, and forming the quantity prediction result A j And display color C j Fourth association mapping A between j →C j Inputting parameters of different positions to be detected and a predictive graph to be detected formed by corresponding matching patterns into a trained CNN, and obtaining C through the fourth association mapping j Then a second composite mapping relation is formed
Example 5
This example describes the analysis of the spectra described in example 4 using a lattice spectrometer detection analysis.
As shown in fig. 6 a, the lattice spectrometer 1 comprises a probe 2 consisting of a 4×4 channel spectrum tester, a data processing module, a data analysis module, and a WiFi system for data transmission.
Wherein, as shown in a and d in fig. 6, each probe 2 comprises a 4×4 lattice formed by forming a pair of points with light source-detector, each point has a channel 2-1 for accommodating a light source and a channel 2-2 for accommodating a polychromator 8, the light source transmits light signals through an optical fiber 6, the optical fiber 6 is connected through an optical fiber connector for fixing the optical fiber 6 and a collimator lens 7, the data processing module comprises 4×4 channel filter-photoelectric converter-preamplifier groups, the channels formed by each filter-photoelectric converter-preamplifier group correspond to the polychromator 8 in the points to collect spectroscopic signals, and the data analysis module is used for performing intensity analysis on the collected pre-amplifier signals to form spectrum analysis data, wherein the polychromator 8 is a detector as a dispersion system.
As shown in b and c in fig. 6, a dust-proof cover is installed outside the lattice spectrometer 1, and is provided with an automatic opening and closing door driven by a motor, and is in a closed state (b in fig. 6) when the detection analysis is not performed, whereas all 4×4 probes 2 are exposed under rotation (c in fig. 6).
Wherein, the detection analysis is to arrange the lattice spectrometer 1 near the engine (as shown in fig. 7), press the lattice spectrometer to the surface of the engine through a cart, control the pressing distance and detect and acquire the spectrum analysis data of the materials on the engine through movement.
The 5×5 prediction map of one test point in fig. 4 belongs to the matching map P i The part of (2) includes a 4×5 pixel size, so that there is a content order of 20 elements, and a specific order may be from right to left and from top to bottom according to the periodic table order of the elements, and a color value is given to indicate the corresponding content size of the elements. If the number of element test seeds is less than 20, the corresponding pixel point is defined as white.
Thus to form a predictive map of 44 x 44 test points as in fig. 5, it is necessary to move the probes 2 to different positions on 11 x 11 engines, where the position on the engine illuminated by the light source in each probe 2 belongs to one of the sampling positions i.
As shown in fig. 7, when a map analysis is performed on an aeroengine to be overhauled, 11×11 cross-sectional directions in D and E directions are shown distributed on the side surfaces and both end positions of the engine.
Wherein, the shallow is equipped with the lifting system of installing at the shallow sweep that relies on a plurality of telescopic columns to realize the lifting, and fig. 7a is the state when lifting system the lowest position, and fig. 7b is the state when lifting to the highest height for be used for carrying out scanning detection to the engine top and make the setting of lattice spectrometer 1 can reach required height.
The cart comprises a multidirectional wheel with a pneumatic tire and arranged at the bottom of a cart plate or on a lifting system, and a multi-axis robot arranged on the cart, wherein the tail end of a limb of the multi-axis robot is connected with the lattice spectrometer 1 to realize multi-dimensional movement, so that the close distance and movement of the probe 2 are controlled.
The cart is provided with a control panel for programming the method of steps S1-S6 of embodiment 4 and programming the movements of the multi-axis robot and the lifting system, and a power supply system for providing power to the multi-axis robot and the lattice spectrometer.
Wherein the multiaxis robot includes the base 5 of internally mounted controller, with base swivelling joint's roating seat 4, but with the first linking arm of swivel seat 4 through pivot 3 connection's electric telescopic, and with the second linking arm that first linking arm passes through pivot 2' and connects, lattice spectrometer 1 passes through joint (b and c in fig. 6) and installs at first linking arm mould end and can rotate around first linking arm.
In this way, the cart is pushed to scan at 11×11 positions of the above distribution of the engine shown in fig. 7, thereby obtaining a plurality of prediction maps p i The CNN established according to the flow of FIG. 4 is mapped A by a fourth association j →C j The display color is obtained, so that the specific location in fig. 7 where the material of the aircraft engine to be tested takes place is obtained. Mapping by a first composite associationAnd/or a second composite mapping relation +.>And the effective association organization and management among the data are completed. And an maintainer can search the related information, experimental data and map analysis result data of each aeroengine only through a calibrated sampling position i on the engine.
Example 6
This embodiment will describe a system for aeroengine data storage and health management of the present invention, as shown in fig. 8, comprising a server, mySQL database, a cart as described in example 5, wherein,
the server is used for inputting relevant information and experimental data of the aeroengine and for realizing steps S2-S6 in the method described in the embodiment 4;
the trolley is in data transmission with the server through a WiFi system in the lattice spectrometer 1, and the server and the MySQL database are communicated with each other through data access requests and calls.
As shown in fig. 9, the control panel includes two display screens, corresponding to a programming area and a main screen, where the program file icons in the programming area can be manually dragged to the main screen to implement operation of the program file in the main screen for program debugging to perform motion debugging of the multi-axis robot, and the documents displayed in the main screen can also be manually dragged to the programming area for programming, where the main screen is used for spectrum detection parameter adjustment, display, saving, deleting and restoring of spectrum data, copy and paste, entry, storage, deleting and restoring of spectrum data and spectrum detection parameters, copy and paste, process analysis, transmission, debugging of the program, adjustment and control of motion parameters of the multi-axis robot and lifting system, and power management.
Wherein the programming comprises programming for implementing the aforementioned method, including programming for implementing steps S1-S6 in embodiment 4, programming for movements of the multi-axis robot and lifting system, the program document may be sent to the server via the programming area and/or the home screen using a WiFi system or a bluetooth system, and can be further modified in the server to simulate commissioning and be passed back to the cart for commissioning confirmation in the home screen (as shown in fig. 8).
And if the test result still does not reach the standard, continuing to debug in the programming area until the test result reaches the standard, and/or sending out an instruction again to enable the server to continue the repeated debugging simulation, and returning the simulation result to the cart for debugging and reconfirming again until the test result reaches the standard.

Claims (10)

1. A method for managing the health of an aeroengine, characterized in that it uses MySQL database to collect the detection data of the aeroengine during the overhaul, uses the program developed by Python to analyze and predict the stored detection data, and specifically comprises the following steps S1 to S6:
s1, inputting relevant information of an aeroengine, wherein the relevant information comprises the following steps: engine model (Engine Type), engine number (ESN), aircraft model (Aircraft Type), aircraft registration number (Aircraft Reg), engine self-service time (ETSN), engine self-service cycle (ECSN), engine self-last repair time (ETSR), engine self-last repair cycle (ECSR), sampling position i;
and selecting corresponding sample information: debris, oil filter and oil product, and storing the related information and the sample information in an engine information table T in a MySQL database i In (a) and (b);
for the case of a plurality of sampling positions, the sampling Position information is stored in a Position field in a JSON form, and a sampling Position i and an engine information table T are established i First association mapping i-T between i ；
S2, inputting experimental data, comprising: sample size and appearance photographs, profile analysis photographs, and profile analysis data; packaging the experimental data to form a data packet B i Numbering is carried outEstablishing a second association map +_with said engine information table in step S1>Thus, the sample position i can be correlated to the test data packing number +.>I.e. form a first composite associative mapping +.>
S3, while inputting the map analysis data in the step S1, the system automatically grabs the map analysis data by using a Python program, sorts the names of various elements according to a preset sequence, calculates the main element components and the secondary element components, and endows various elements with different color valuesRepresenting different contents to form a matching chart P i The method comprises the steps of carrying out a first treatment on the surface of the I.e. package numberingA third association map is formed with the matching graph>The materials in the element component comparison table are subjected to element sorting according to the same sequence and are endowed with specified content limit corresponding color values to form a limit map P j J represents the j-th alloy material in the element composition comparison table; the color value is->
S4 at S3 forming a match map P i The system then uses the Python program to match the map P i And limit map P j Performing a difference operation P i -P j The method comprises the steps of carrying out a first treatment on the surface of the If the differential result is a blank graph or the sum L of color values of all elements is in a specified range, then the matched result is indicated to determine the type of alloy material, otherwise, the alloy material cannot be matched, and the alloy material is indicated to be unknown, and the sampling position i is indicated to be normal or needs to be further detected;
wherein the specified range is specified as follows:
k is the sequential number of the preset sequence of elements, N is the total seed number of the elements, R k 、G k 、B k An allowable maximum deviation value for the three color components of element k;
s5, after the steps are finished, the MySQL database continuously stores relevant experimental data of each model of engine, and the system grabs the experimental data and the quantity A j A control rule, if the sensitive material is found and the quantity exceeds the limit value, displaying in red, and treating as an abnormal sample; if sensitive material is found but the amount does not exceed the limit, the color is displayed in yellow, and the sample is regarded as suspicious to be noticedThe method comprises the steps of carrying out a first treatment on the surface of the If no sensitive material is found and the amount does not exceed the limit, the color is green, and the sample is considered to be a normal sample, the amount A j Refers to the total points in the specified range in different sampling locations;
s6, acquiring parameters Engine Type and AircraftType, ETSN, ECSN, ETSR at different positions on a plurality of engines, and also endowing different color values of each parameter to represent different parameter values and match with a map P i Splicing to form predictive map p i The quantity of the corresponding materials at different positions is obtained, a prediction graph is divided into a training set and a verification set, the ratio of the training set to the verification set is 5:1-1:1, a convolutional neural network model CNN is established, the quantity of the materials is formed by outputting the quantity to a classification function softmax or sigmond through a full connection layer FC, and a quantity prediction result A is formed according to the rule j And display color C j Fourth association mapping A between j →C j Inputting parameters of different positions to be detected and a predictive graph to be detected formed by corresponding matching patterns into a trained CNN, and obtaining C through the fourth association mapping j Then a second composite mapping relation is formed
2. The method of claim 1, wherein the rules in step (5) are as shown in the following table:
3. the method according to claim 1 or 2, wherein the profile analysis uses a lattice spectrometer for detection analysis;
the lattice spectrometer comprises a probe consisting of a multichannel spectrum tester, a data processing module, a data analysis module and a WiFi system or a Bluetooth system for data transmission; the probe comprises a dot matrix formed by a pair of light source-detector components as dots, wherein the light source in each dot transmits light signals through optical fibers, and the emergent end of the optical fiber connected with the light source is provided with a collimator; the data processing module comprises a multi-channel filter-photoelectric converter-pre-amplification group, and a channel formed by each filter-photoelectric converter-pre-amplification group corresponds to the detector in the point so as to collect the spectroscopic optical signal; the data analysis module is used for carrying out intensity analysis on the collected pre-amplified signals to form map analysis data, wherein the detector is a dispersion system.
4. The method of claim 3, wherein the further testing comprises comprehensive analysis by scanning electron microscopy SEM with EDS spectrometer, transmission electron microscopy TEM with EDS spectrometer, X-ray spectroscopy XPS, X-ray diffraction XRD.
5. The method of claim 4, wherein the condition requiring further analysis is that the length of engine service exceeds 1/3 or 1/4 of the theoretical total expected length of service or expected life.
6. The method according to claim 4 or 5, wherein a dust-proof shield is installed outside the lattice spectrometer, the dust-proof shield is provided with an automatic opening and closing door, and is in a closed state when the detection analysis is not performed, and is opened to expose the probe;
the detection and analysis are that the lattice spectrometer is arranged near an engine, and the movement controller is used for controlling the distance close to the surface of the engine to detect and acquire the map analysis data of the materials on the engine;
the movable controller comprises a cart provided with a multidirectional wheel with a pneumatic tire, a multi-axis robot arranged on a cart plate, wherein the tail end of the multi-axis robot is connected with the lattice spectrometer to realize multi-dimensional movement, and a power supply system for supplying power to the multi-axis robot and the lattice spectrometer is also arranged on the cart plate;
the cart is also equipped with a lifting system for enabling the required height of the lattice spectrometer setup when scanning the engine roof.
7. The method according to claim 6, characterized in that the cart is provided with a control panel programmed for implementing the method of steps S1-S6 and programmed for the movements of the multi-axis robot and the lifting system.
8. A system for managing the health of an aeroengine, comprising a server, a MySQL database, a trolley equipped with a lifting system and carrying a multi-axis robot, the extremity of which is connected to a lattice spectrometer by means of a joint on the lattice spectrometer, wherein,
the server is used for inputting relevant information and experimental data of the aeroengine and for realizing steps S2-S6 in the method as claimed in any one of claims 1-7;
a multidirectional wheel with a pneumatic tire is arranged on the bottom of a trolley plate or a lifting system of the trolley, a control panel for programming the steps S1-S6 in the method of any one of claims 1-7 and programming the motion of the multi-axis robot is arranged on the trolley, and a power supply system for providing power for the multi-axis robot and the lattice spectrometer is arranged on the trolley;
and the trolley realizes data transmission between the server through a WiFi system or a Bluetooth system in the lattice spectrometer, and the server is communicated with the MySQL database.
9. The system according to claim 8, wherein the control panel comprises two display screens, corresponding to a programming area and a main screen respectively, wherein program file icons in the programming area can be manually dragged to the main screen to realize program file operation in the main screen for program debugging to perform motion debugging of the multi-axis robot, and documents displayed in the main screen can also be manually dragged to the programming area for programming, and the main screen is used for spectrum detection parameter adjustment, spectrum data display, storage, deletion and restoration, copy and paste, spectrum data and spectrum detection parameter input, storage, deletion and restoration, copy and paste, process analysis, transmission, program debugging, motion parameter adjustment and control of the multi-axis robot and lifting system and power management;
wherein the programming comprises programming for implementing the method according to any of claims 1-7, programming for movements of the multi-axis robot and the lifting system, the program document can be sent to the server by means of a WiFi system or a bluetooth system via the programming area and/or the main screen, and can be further modified in the server to simulate the commissioning and be transmitted back to the cart for commissioning confirmation in the main screen, and if still not up to standard, continuing the commissioning in the programming area until up to standard, and/or issuing instructions again to enable the server to continue the multi-tuning simulation, and be transmitted back again to the cart for commissioning reconfirmation until up to standard.
10. A computer readable non-transitory storage medium having stored therein a program executable by the server in the system of claim 8 or 9 and programmed by an auxiliary control panel to implement a method of managing aeroengine health of any of claims 1-7.
CN202210785464.2A 2022-07-05 2022-07-05 Method and system for managing aeroengine health Active CN115374960B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210785464.2A CN115374960B (en) 2022-07-05 2022-07-05 Method and system for managing aeroengine health

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210785464.2A CN115374960B (en) 2022-07-05 2022-07-05 Method and system for managing aeroengine health

Publications (2)

Publication Number Publication Date
CN115374960A CN115374960A (en) 2022-11-22
CN115374960B true CN115374960B (en) 2023-07-21

Family

ID=84061285

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210785464.2A Active CN115374960B (en) 2022-07-05 2022-07-05 Method and system for managing aeroengine health

Country Status (1)

Country Link
CN (1) CN115374960B (en)

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103728965B (en) * 2012-10-15 2017-02-22 中航商用航空发动机有限责任公司 Monitoring device and method for aircraft engine and FADEC system
CN109035488A (en) * 2018-08-07 2018-12-18 哈尔滨工业大学（威海） Aero-engine time series method for detecting abnormality based on CNN feature extraction
CN110569696A (en) * 2018-08-31 2019-12-13 阿里巴巴集团控股有限公司 Neural network system, method and apparatus for vehicle component identification
CN113987139A (en) * 2021-09-24 2022-01-28 南京航空航天大学 Knowledge graph-based visual query management system for software defect cases of aircraft engine FADEC system
CN114419026A (en) * 2022-02-10 2022-04-29 北京航空航天大学 Visual attention-based aeroengine fuse winding identification system and method

Also Published As

Publication number Publication date
CN115374960A (en) 2022-11-22

Similar Documents

Publication Publication Date Title
CA2780380C (en) Method of and system for evaluating the health status of a system using groups of vibration data comprising images of the vibrations of the system
RU2357215C2 (en) Method of repair and maintenance used in hardware-software complex for diagnostics and system for quality control of repair and maintenance
CA2924611C (en) Systems and methods for non-destructive testing involving remotely located expert
CN101937214B (en) Industrial Ethernet based monitoring system of sea-crossing suspension bridge
CN108229782A (en) A kind of visual production management platform based on cloud
CN111444169A (en) Transformer substation electrical equipment state monitoring and diagnosis system and method
CN111985112B (en) Digital twin system of blast furnace based on Unity3D
Yasuda et al. Aircraft visual inspection: A systematic literature review
CN110268425A (en) System for analyzing machine data
CN101995880A (en) System for diagnosing and testing abnormal operating condition during petrochemical process
CN117314094A (en) Energy management and control method and system based on AI intelligent park
CN1152327C (en) Maintenance system of tester
CN110103856A (en) Automobile, train, subway and aircraft find out fault repair method automatically
DE102020005611A1 (en) vehicle surface analysis system
CN115374960B (en) Method and system for managing aeroengine health
EP2154591B1 (en) Device and method for testing systems with visual output
EP4211419A1 (en) Vehicle surface analysis system
CN107392506A (en) A kind of self-study line status evaluation system and evaluation method for merging development of Mobile Internet technology
CN109614298A (en) A kind of hardware system status monitoring and method for early warning based on monitor video
Mueller et al. Intelligent real time inspection of rivet quality supported by human-robot-collaboration
Kai et al. Semiconductor chip’s quality analysis based on its high dimensional test data
Mridha et al. Evaluate and Predict Concentration of Particulate Matter (PM10) Using Machine Learning Approach
CN104076715A (en) Automobile trunk lock function detecting device
CN211785145U (en) Oil element detection device based on image-assisted atomic emission spectroscopy
Kananadze System for Automated Infrared Image Processing Based on Neural Network Technologies

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