CN113155469A - Engine fault diagnosis alarm system and device - Google Patents
Engine fault diagnosis alarm system and device Download PDFInfo
- Publication number
- CN113155469A CN113155469A CN202110297080.1A CN202110297080A CN113155469A CN 113155469 A CN113155469 A CN 113155469A CN 202110297080 A CN202110297080 A CN 202110297080A CN 113155469 A CN113155469 A CN 113155469A
- Authority
- CN
- China
- Prior art keywords
- fault diagnosis
- engine
- fault
- alarm
- vibration
- 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
- 238000003745 diagnosis Methods 0.000 title claims abstract description 83
- 230000002159 abnormal effect Effects 0.000 claims abstract description 17
- 238000000034 method Methods 0.000 claims abstract description 14
- 238000005516 engineering process Methods 0.000 claims abstract description 12
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 4
- 238000004458 analytical method Methods 0.000 claims description 27
- 238000013527 convolutional neural network Methods 0.000 claims description 18
- 238000013523 data management Methods 0.000 claims description 13
- 239000000203 mixture Substances 0.000 claims description 12
- 230000001133 acceleration Effects 0.000 claims description 10
- 238000007726 management method Methods 0.000 claims description 6
- 238000000354 decomposition reaction Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 230000035945 sensitivity Effects 0.000 claims description 3
- 230000003595 spectral effect Effects 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 3
- 230000004927 fusion Effects 0.000 claims description 2
- 230000006855 networking Effects 0.000 claims description 2
- 238000012549 training Methods 0.000 claims description 2
- 230000003993 interaction Effects 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 7
- 241000282414 Homo sapiens Species 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000010183 spectrum analysis Methods 0.000 description 2
- 239000004575 stone Substances 0.000 description 2
- 230000007123 defense Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000015541 sensory perception of touch Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M15/00—Testing of engines
- G01M15/04—Testing internal-combustion engines
- G01M15/12—Testing internal-combustion engines by monitoring vibrations
Landscapes
- Chemical & Material Sciences (AREA)
- Engineering & Computer Science (AREA)
- Combustion & Propulsion (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses an engine fault diagnosis alarm system and device, and relates to the field of engine fault diagnosis, in particular to the problems that the fault diagnosis is based on experience, the automation degree is low, and mass signals are difficult to manage. The engine fault diagnosis alarm system and the device are an integrated system integrating a sensor, a collection card, signal collection, signal display, data cloud storage, fault diagnosis and abnormal fault data alarm into a whole: the acquisition card is controlled by fault diagnosis client software to read vibration signals acquired by a vibration sensor attached to the surface of an engine in real time, the fault is automatically diagnosed by using an artificial intelligence technology, and real-time alarm on the fault is realized by combining software; the vibration signal and the diagnosis data are automatically stored and recorded, hands of workers are liberated, and the efficiency and the accuracy of engine fault diagnosis are greatly improved; the system has strong functionality and friendly human-computer interaction, and supports the expansion of multi-channel and multi-sensor; the method is applied to the fault diagnosis of the aircraft engine.
Description
Technical Field
The invention relates to an engine fault diagnosis alarm system and device, which mainly carry out diagnosis, alarm and mass data management on engine faults.
Background
The aircraft engine is a highly complex and precise machine, is a device for providing driving force or supporting force for an aircraft, is the 'heart' of a spacecraft, is known as a bright pearl on an industrial crown, is an important driving force for promoting the rapid development of aviation industry, is the most complex technical process, the highest technical difficulty and the most expensive part in the aerospace industry, and has the highest requirements on reliability and safety. The development of the aeroengine technology has an important role in national defense and national economy, and each time of human beings have great revolutionary progress in the aeronautical field, the development of the aeroengine technology is not closely related to the breakthrough and progress of the aeronautical power technology. If the potential fault existing in the conventional flight vibration cannot be detected in time and is not effectively checked and repaired, the damage of the aircraft engine is aggravated, the aircraft engine cannot be normally put into use, even an accident of machine damage and human death can occur in serious conditions, and huge loss is caused.
The electromechanical equipment is developed in the direction of large-scale, high-speed, precision, systematization and automation. In order to maintain safe operation of the equipment, the health of the major components must be monitored in real time. The technical roadmap of the key field of "china manufacture 2025" indicates in particular: real-time fault diagnosis is required to be realized for aviation generators, advanced rail transit equipment, power generation equipment, agricultural equipment and the like. In particular, aircraft generators, reports clearly indicate that advanced health management systems with state monitoring, fault diagnosis and processing, fault prediction, and life management functions must be developed. Due to the fact that modern electromechanical equipment is large in scale, multiple monitoring points are arranged, the sampling frequency of the sensor is high, and the detection system is long in service period of the equipment and can obtain massive data. For example, 1149 devices are monitored by 29 enterprises in Beijing chemical university Gaojingji and the like, the total number of detection points is 18552, and the data volume also reaches 1.52 TB; the above examples show that condition monitoring of electromechanical equipment is advancing into the "big data" era.
The electromechanical big data era brings new challenges to fault diagnosis and has the following characteristics:
(1) the data volume is large. Accumulated data volume under actual conditions is bigger and bigger, but professional analyst's quantity is serious not enough, only relies on the manpower to detect and can not satisfy the requirement yet, needs urgently to carry out automated inspection's intelligent algorithm.
(2) The data types are diverse. The field data come from different acquisition devices, acquisition working conditions, acquisition environments and the like, the data types are complex and various, the characteristics are difficult to mine, and the detection difficulty is high.
(3) Data requirements are becoming more and more accurate. Under the condition of high speed, the connection of all parts in the equipment is tighter, and the tiny fault of one part possibly causes chain reaction, so that the whole equipment is broken down.
In China, mechanical equipment fault diagnosis based on vibration signals has been in history for more than thirty years, and initially, workers judge faults by knocking mechanical equipment with a hammer. However, if the analysis of the vibration signal only stays in the auditory sense and the tactile sense, the analysis effect is influenced by various factors such as the working experience of the diagnostician, the surrounding environment, the self condition of the equipment and the like, so that the analysis result has a large error, and a large amount of human resources are consumed. People's demand to vibration diagnosis constantly increases and impels the continuous maturity of vibration diagnosis technique, just vibration sensor has appeared, and the staff can utilize multichannel vibration sensor and data acquisition card to accomplish the collection work of data, converts analog signal into digital signal, and the data that so gather is accurate, the error is less, though vibration sensor's appearance makes the data of gathering more and more accurate, the data staff that so gathers is not convenient for directly to carry out the analysis to data to can't keep data. With the advancement of science and technology, the maintenance and management of equipment also needs to be more scientific. Therefore, the engine fault diagnosis alarm system and the engine fault diagnosis alarm device are produced at the same time.
The following problems exist in the current fault diagnosis based on vibration signals: (1) the fault diagnosis is empirical, and the automation degree is low: in the conventional fault diagnosis, based on software diagnosis, workers judge whether equipment is abnormal or not through experience by observing frequency spectrums of various sensors. The diagnosticians are required to observe the data and make records at any time. The use is inconvenient, and the great manpower and material resources are consumed. (2) Managing mass data: the diagnosis of the device will yield a huge amount of data. These data require archiving, viewing and statistical analysis. And therefore needs to be managed efficiently. However, at present, enterprise test beds may come from different manufacturers, and data which are mutually incompatible cannot be managed. Most of the data are saved by a magnetic disk and do not play a role in application. (3) And (4) fault alarm: traditional monitoring relies on human resources, judges the trouble manually, to the emergence of trouble, the system does not warn.
Disclosure of Invention
Aiming at the problems, the invention aims to provide an engine fault diagnosis alarm system and device, which read vibration signals collected by a vibration sensor attached to the surface of an engine in real time through a fault diagnosis client software control collection card, automatically diagnose faults by using an artificial intelligence technology and realize real-time alarm on the faults by combining software; the vibration signal and the diagnosis data are automatically stored, recorded and displayed, so that workers can conveniently watch the vibration signal and the diagnosis data, massive data generated by vibration of the engine is managed, and finally, the engine fault is diagnosed and alarmed.
The engine fault diagnosis alarm system and device include: the system comprises a signal acquisition module, a signal display module, a signal analysis module, a data management module, a signal playback module, a fault diagnosis module and an alarm module; the system work flow chart is shown in figure 1:
the data acquisition module controls the PCI board card to acquire vibration signals sensed by an acceleration sensor attached to the surface of the engine in real time; the collected data are transmitted to a signal analysis module;
the signal analysis module carries out spectrum analysis on the signals, and meanwhile, the data management module is combined to store the vibration signals in real time;
the fault diagnosis module automatically diagnoses faults by using a deep learning technology on the collected vibration signals; the alarm module alarms through the abnormal points received from the signal analysis module and the faults identified by the fault diagnosis module;
the playback module can play back the data collected once; the data management module is used for managing the vibration signal acquisition records of the equipment, and can trace back the vibration condition of each machine in the future to summarize the vibration rule of the engine from a long-term perspective.
S1, preparing an industrial personal computer, a vibration sensor, a connecting wire, a PCI acquisition card and engine fault diagnosis client software, installing the vibration acceleration sensor at the specified position of the engine, and connecting the vibration acceleration sensor with the PCI acquisition card installed on the host machine through the connecting wire;
s2, opening client software in the industrial personal computer, selecting acquisition parameters such as acquisition frequency, analysis frequency and spectral line number, setting channel parameters such as range and sensitivity of each channel, and controlling an acquisition card to acquire vibration signals in real time;
s3, converting the acquired time sequence signal into a frequency domain signal in real time by the signal analysis module, calculating the frequency, peak value and total value of the frequency domain signal, and transmitting the frequency domain signal to the signal display module for display;
s4, the data management module stores the acquired data in real time, stores and manages the channel parameters and the acquisition parameters, stores the data in a MySQL database, synchronizes the local database to a cloud server, and the cloud server collects vibration signals of all devices;
s5, the fault diagnosis module is used for analyzing and diagnosing faults of the collected real-time signals;
s6, the alarm template receives the abnormal points of the frequency, the peak value or the total value analyzed by the S3 signal analysis module and the faults identified by the S5 fault diagnosis module to alarm;
s7, the playback module plays back the engine diagnosis history, and the data is taken from the data management module and then sent to the signal display module for display.
S5, the fault diagnosis module diagnoses the fault of the aircraft engine and comprises the following four steps:
s51, combining the channels pairwise, averaging to obtain a mean channel, and expanding the number of the channels;
s52, extracting features of the data after the channel expansion by using a one-dimensional convolutional neural network (1DCNN) respectively;
s53, fusing the features of each channel extracted in the S52 into multi-dimensional features, and then decomposing the multi-dimensional features into one dimension by using tensor decomposition;
and S54, classifying the one-dimensional features of the S53 by using a Gaussian Mixture Model (GMM) to obtain a fault type.
Effects of the invention
The engine fault diagnosis alarm system and the device are a comprehensive system integrating a sensor, an acquisition card, signal acquisition, signal display, data cloud storage, fault diagnosis and abnormal fault data alarm; the acquisition card is controlled by fault diagnosis client software to read vibration signals acquired by a vibration sensor attached to the surface of an engine in real time, the fault is automatically diagnosed by using an artificial intelligence technology, and real-time alarm on the fault is realized by combining software; the vibration signal and the diagnosis data are automatically stored and recorded, hands of workers are liberated, and the efficiency and the accuracy of engine fault diagnosis are greatly improved; the system has strong functionality, covers all functions in engine fault diagnosis, provides a good user interface, has strong man-machine interaction and good expandability, and supports unlimited channel expansion under the condition of hardware permission, and the multi-channel and multi-sensor unified management is realized.
Drawings
A work flow diagram of the system of FIG. 1;
FIG. 2 is a software system interface;
FIG. 3 is a schematic diagram of an experimental procedure;
FIG. 41 is a parameter diagram of a DCNN network;
detailed description of the invention
The invention is further described in connection with the following application scenarios.
The engine fault diagnosis alarm system and the device are characterized in that (1) cloud storage of engine diagnosis data is realized, (2) a rotary mechanical fault diagnosis method based on a Convolutional Neural Network (CNN) and a Gaussian Mixture Model (GMM) is adopted, and (3) real-time alarm of abnormal values and faults of vibration signals is realized; the functions mainly comprise: the system comprises a signal acquisition module, a signal display module, a signal analysis module, a data management module, a signal playback module, a fault diagnosis module and an alarm module; the system work flow chart is shown in figure 1:
the data acquisition module controls the PCI board card to acquire vibration signals sensed by an acceleration sensor attached to the surface of the engine in real time; the collected data are transmitted to a signal analysis module;
the signal analysis module carries out spectrum analysis on the signals, and meanwhile, the data management module is combined to store the vibration signals in real time;
the fault diagnosis module automatically diagnoses faults by using a deep learning technology on the collected vibration signals; the alarm module alarms through the abnormal points received from the signal analysis module and the faults identified by the fault diagnosis module; the playback module can play back the data collected once; the data management module is used for managing the vibration signal acquisition records of the equipment, and can trace back the vibration condition of each machine in the future to summarize the vibration rule of the engine from a long-term perspective.
The fault diagnosis alarm system and the fault diagnosis alarm device comprise the following steps:
s1, preparing an industrial personal computer, a comet stone science and technology CA7002A type vibration acceleration sensor, a connecting wire, a porphyry PCI-1714U high-speed acquisition card and engine fault diagnosis client software, mounting the vibration acceleration sensor at a specified position of an engine, and connecting the vibration acceleration sensor with the porphyry acquisition card mounted on a host machine through the connecting wire;
s2, opening client software in the industrial personal computer, setting acquisition parameters such as acquisition frequency, analysis frequency and spectral line number, and setting channel parameters such as measuring range and sensitivity of each channel; then controlling a collection card to collect vibration time sequence signals in real time;
s3, converting the time sequence signal into a frequency domain signal in real time by the signal analysis module, calculating the frequency, peak value and total value of the frequency domain signal, and displaying the frequency, peak value and total value on the display module;
s4, the data management module stores the acquired vibration time sequence signals in real time and manages channel parameters and acquisition parameters; the data are stored in a MySQL database, the local database is synchronized to a cloud server, and the cloud server collects vibration signals of all the devices;
s5, the fault diagnosis module analyzes and diagnoses the vibration signal;
s6, the alarm module analyzes abnormal points according to the results of the steps S3 and S5 and gives an alarm;
s7, the playback module may present historical diagnostic information of the engine. The following examples illustrate the invention in detail:
1. the principle and function of the signal acquisition module are as follows:
the acquisition module sensor is a vibration measurement sensor of a comet stone technology, is CA7002A in model and is a vibration acceleration sensor based on voltage; the acquisition card adopts an independent high-speed analog input adapter card suitable for a porphyrized PCI bus, the model is 1714U, a 4-channel synchronous analog input card realizes multi-channel acquisition, the sampling frequency is supported by 23 ten thousand at least every second, the precision is stable (the 12-bit precision is ensured, no code hopping is realized in constant input, the conversion rate can reach 30MHz), the high-speed and medium-speed functions are realized, and the PCI card slot installation based on the industrial personal computer host can be expanded randomly according to the PCI card slot on the host; by using developed aircraft engine fault diagnosis client software, a channel parameter and an acquisition parameter are set, and then a 'start acquisition' button can be clicked to control the acquisition card to acquire vibration data simultaneously in multiple channels.
2. The principle and function of the signal analysis module is as follows:
time domain signals often cannot intuitively reflect the rule of vibration signals, so frequency domain analysis is used; converting the time domain signal into a frequency domain signal by Fast Fourier Transform (FFT), and analyzing a peak value, a total value, and a frequency in a corresponding spectrogram; some engine vibrations are concentrated on medium-low frequency, some engine vibrations are mainly concentrated on medium-high frequency, and when the engine breaks down, the frequency spectrum of the engine changes, the peak value and the frequency increase suddenly, and at this time, an abnormal point is diagnosed, and an alarm module gives an alarm.
3. The principle and function of the data management module are as follows:
the software client controls the PCI acquisition card to acquire vibration signals in real time, stores the vibration signals in a local file in real time under an offline condition, and stores the vibration signal acquisition records, channel parameters, acquisition parameters and other project management data and frequency, peak values and total value data of each time period analyzed by the signal analysis module in a local MySQL database; the vibration signals are transmitted to the server in real time through the network under the networking condition to be stored, and meanwhile, local database records which are not uploaded to the server when the device is offline are uploaded to the server, so that the data are uniformly managed by the cloud platform, if the device has problems in the future, the diagnosis records of the device can be backtracked, and the diagnosis method of the faults can be learned and summarized from the historical records.
4. The rotary machine fault diagnosis method based on the convolutional neural network and the Gaussian mixture model comprises the following steps:
aiming at the problem of low accuracy of fault diagnosis of the rotary machine, deep-level feature information is extracted from data by using a Convolutional Neural Network (CNN) and fault modeling is performed by combining a Gaussian Mixture Model (GMM), and a fault diagnosis method fusing the Convolutional Neural network and the Gaussian Mixture Model is provided; in the training stage, firstly, a convolutional neural network is trained by utilizing a large number of vibration signals to extract the characteristics of the vibration signals, and then a GMM is trained for each type of fault; in the fault diagnosis stage, combining the multiple channels pairwise to calculate the mean value to obtain a mean value channel so as to expand a data set; then, feature extraction is carried out on the expanded data set one by using a trained one-dimensional convolutional neural network (1 DCNN); the features O after extraction1…Oi…Os，OiThrough feature fusion, a multidimensional feature O ═ O is obtained1...Oi...Os](ii) a Then decomposing the multidimensional characteristic O into one-dimensional characteristics through tensor decomposition, and inputting the one-dimensional characteristics to a GMM (Gaussian mixture model) model for classification to realize fault diagnosis; the method comprises the following specific steps:
s51, combining the channels pairwise, averaging to obtain a mean channel, and expanding the number of the channels;
s52, extracting features of the data after the channel expansion by using a one-dimensional convolutional neural network (1DCNN) respectively;
s53, fusing the features of each channel extracted in the S52 into multi-dimensional features, and then decomposing the multi-dimensional features into one dimension by using tensor decomposition;
s54, classifying the one-dimensional features of the S53 by using a Gaussian Mixture Model (GMM) to obtain a fault type;
a large number of experimental results show that the fault identification accuracy of the method is higher than that of a KNN algorithm, an SVM algorithm, an LSTM algorithm, a CNN algorithm and a method combining MFCC and GMM, and fault diagnosis of rotary mechanical equipment can be effectively realized.
5. The principle of the alarm module is as follows:
the alarm module is used for alarming abnormal points and faults of the vibration signals: (1) the frequency, the peak value and the total value analyzed by the signal analysis module are always within a certain range, experienced workers in a factory set a threshold value for each model of engine, and after the frequency, the peak value and the total value exceed the threshold values, the equipment is judged to be abnormal, at the moment, the software client starts to give an alarm, and an abnormal acquisition window turns red and flickers; (2) the real-time fault diagnosis result of the signal returned by the fault diagnosis module alarms according to the corresponding fault; while preserving the diagnosed anomaly and fault.
Claims (5)
1. The system and the device for diagnosing and alarming the engine fault are characterized in that a collection card is controlled by fault diagnosis client software to read vibration signals collected by a vibration sensor fixed on the surface of an engine in real time, the fault is automatically diagnosed by using an artificial intelligence technology, the real-time alarm of the fault is realized by combining software, and the vibration signals and diagnosis data are automatically stored and recorded; the fault diagnosis alarm system and the fault diagnosis alarm device comprise the following steps:
s1, preparing an industrial personal computer, a vibration sensor, a connecting wire, a PCI acquisition card and engine fault diagnosis client software, installing the vibration acceleration sensor at the specified position of the engine, and connecting the vibration acceleration sensor with the PCI acquisition card installed on the host machine through the connecting wire;
s2, opening client software in the industrial personal computer, setting acquisition parameters such as acquisition frequency, analysis frequency and spectral line number, and setting channel parameters such as measuring range and sensitivity of each channel; then controlling a collection card to collect vibration time sequence signals in real time;
s3, converting the time sequence signal into a frequency domain signal in real time by the signal analysis module, calculating the frequency, peak value and total value of the frequency domain signal, and displaying the frequency, peak value and total value on the display module;
s4, the data management module stores the acquired vibration time sequence signals in real time and manages channel parameters and acquisition parameters; the data are stored in a MySQL database, the local database is synchronized to a cloud server, and the cloud server collects vibration signals of all the devices;
s5, the fault diagnosis module analyzes and diagnoses the vibration signal;
s6, the alarm module analyzes abnormal points according to the results of the steps S3 and S5 and gives an alarm;
s7, the playback module may present historical diagnostic information of the engine.
2. The system and apparatus for engine fault diagnosis and alarm as claimed in claim 1, wherein the principle of the signal analysis module in step S3 is as follows:
time domain signals often cannot intuitively reflect the rule of vibration signals, so frequency domain analysis is used; converting the time domain signal into a frequency domain signal by Fast Fourier Transform (FFT), and analyzing a peak value, a total value, and a frequency in a corresponding spectrogram; some engine vibrations are concentrated on medium-low frequency, some engine vibrations are mainly concentrated on medium-high frequency, and when the engine breaks down, the frequency spectrum of the engine changes, the peak value and the frequency increase suddenly, and at this time, an abnormal point is diagnosed, and an alarm module gives an alarm.
3. The system and apparatus for engine fault diagnosis and alarm as claimed in claim 1, wherein the data management module in step S4 is based on the following principles:
the software client controls the PCI acquisition card to acquire vibration signals in real time, stores the vibration signals in a local file in real time under an offline condition, and stores project management data such as signal acquisition records, channel parameters and acquisition parameters and frequency, peak value and total value data of each time period analyzed by the signal analysis module in a local MySQL database; the vibration signals are transmitted to the server in real time through the network under the networking condition to be stored, and meanwhile, local database records which are not uploaded to the server when the device is offline are uploaded to the server, so that the data are uniformly managed by the cloud platform, if the device has problems in the future, the diagnosis records of the device can be backtracked, and the diagnosis method of the faults can be learned and summarized from the historical records.
4. The system and apparatus for engine fault diagnosis and alarm as claimed in claim 1, wherein the principle of the method for diagnosing faults of rotating machinery based on the convolutional neural network and the gaussian mixture model in step S5 is as follows:
aiming at the problem of low accuracy of fault diagnosis of the rotary machine, deep-level feature information is extracted from data by using a Convolutional Neural Network (CNN) and fault modeling is performed by combining a Gaussian Mixture Model (GMM), and a fault diagnosis method fusing the Convolutional Neural network and the Gaussian Mixture Model is provided; in the training stage, firstly, a convolutional neural network is trained by utilizing a large number of vibration signals to extract the characteristics of the vibration signals, and then a GMM is trained for each type of fault; in the fault diagnosis stage, combining the multiple channels pairwise to calculate the mean value to obtain a mean value channel so as to expand a data set; then, feature extraction is carried out on the expanded data set one by using a trained one-dimensional convolutional neural network (1 DCNN); the features O after extraction1…Oi…Os，OiThrough feature fusion, a multidimensional feature O ═ O is obtained1...Oi...Os](ii) a Then decomposing the multidimensional characteristic O into one-dimensional characteristics through tensor decomposition, and inputting the one-dimensional characteristics to a GMM (Gaussian mixture model) model for classification to realize fault diagnosis; the method comprises the following specific steps:
s51, combining the channels pairwise, averaging to obtain a mean channel, and expanding the number of the channels;
s52, extracting features of the data after the channel expansion by using a one-dimensional convolutional neural network (1DCNN) respectively;
s53, fusing the features of each channel extracted in the S52 into multi-dimensional features, and then decomposing the multi-dimensional features into one dimension by using tensor decomposition;
and S54, classifying the one-dimensional features of the S53 by using a Gaussian Mixture Model (GMM) to obtain a fault type.
5. The system and the device for engine fault diagnosis and alarm as claimed in claims 1, 2 and 4, wherein the principle of the alarm module in the step S6 is as follows:
the alarm module is used for alarming abnormal points and faults of the vibration signals: (1) the frequency, the peak value and the total value analyzed by the signal analysis module are always within a certain range, experienced workers in a factory set a threshold value for each model of engine, and after the frequency, the peak value and the total value exceed the threshold values, the equipment is judged to be abnormal, at the moment, the software client starts to give an alarm, and an abnormal acquisition window turns red and flickers; (2) the real-time fault diagnosis result of the signal returned by the fault diagnosis module alarms according to the corresponding fault; while preserving the diagnosed anomaly and fault.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110297080.1A CN113155469A (en) | 2021-03-19 | 2021-03-19 | Engine fault diagnosis alarm system and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110297080.1A CN113155469A (en) | 2021-03-19 | 2021-03-19 | Engine fault diagnosis alarm system and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113155469A true CN113155469A (en) | 2021-07-23 |
Family
ID=76887715
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110297080.1A Pending CN113155469A (en) | 2021-03-19 | 2021-03-19 | Engine fault diagnosis alarm system and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113155469A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113607205A (en) * | 2021-08-02 | 2021-11-05 | 中国民航大学 | Method and device for detecting faults of aero-engine sensor |
| CN113776794A (en) * | 2021-08-13 | 2021-12-10 | 昆明理工大学 | Fault diagnosis method, device and system for embedded edge computing |
| CN113837245A (en) * | 2021-09-06 | 2021-12-24 | 西安理工大学 | Flexible printing electronic equipment core unit fault diagnosis method |
| CN114252716A (en) * | 2021-12-06 | 2022-03-29 | 深圳供电局有限公司 | Power grid fault diagnosis method and device and storage medium |
| CN114893428A (en) * | 2022-04-02 | 2022-08-12 | 上海大学 | Fine pre-diagnosis and accurate operation and maintenance device and system for ventilator |
| CN114910733A (en) * | 2022-07-15 | 2022-08-16 | 深圳益实科技有限公司 | Display fault intelligent diagnosis analysis system based on artificial intelligence |
| CN117189720A (en) * | 2023-09-14 | 2023-12-08 | 成都飞航智云科技有限公司 | Fault diagnosis method for hydraulic system of aircraft |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109211548A (en) * | 2018-08-31 | 2019-01-15 | 沃德传动（天津）股份有限公司 | A kind of mechanical failure diagnostic method |
| CN109506921A (en) * | 2018-12-24 | 2019-03-22 | 西安科技大学 | A kind of rotary machinery fault diagnosis and method for early warning |
| CN109765045A (en) * | 2019-03-07 | 2019-05-17 | 广东省智能制造研究所 | It is a kind of for rotating the system and method for class mechanical remote monitoring and fault diagnosis |
| CN110647830A (en) * | 2019-09-12 | 2020-01-03 | 华中科技大学 | Bearing fault diagnosis method based on convolutional neural network and Gaussian mixture model |
| CN110779716A (en) * | 2019-11-01 | 2020-02-11 | 苏州德姆斯信息技术有限公司 | Embedded mechanical fault intelligent diagnosis equipment and diagnosis method |
-
2021
- 2021-03-19 CN CN202110297080.1A patent/CN113155469A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109211548A (en) * | 2018-08-31 | 2019-01-15 | 沃德传动（天津）股份有限公司 | A kind of mechanical failure diagnostic method |
| CN109506921A (en) * | 2018-12-24 | 2019-03-22 | 西安科技大学 | A kind of rotary machinery fault diagnosis and method for early warning |
| CN109765045A (en) * | 2019-03-07 | 2019-05-17 | 广东省智能制造研究所 | It is a kind of for rotating the system and method for class mechanical remote monitoring and fault diagnosis |
| CN110647830A (en) * | 2019-09-12 | 2020-01-03 | 华中科技大学 | Bearing fault diagnosis method based on convolutional neural network and Gaussian mixture model |
| CN110779716A (en) * | 2019-11-01 | 2020-02-11 | 苏州德姆斯信息技术有限公司 | Embedded mechanical fault intelligent diagnosis equipment and diagnosis method |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113607205A (en) * | 2021-08-02 | 2021-11-05 | 中国民航大学 | Method and device for detecting faults of aero-engine sensor |
| CN113607205B (en) * | 2021-08-02 | 2023-09-19 | 中国民航大学 | Method and device for detecting sensor faults of aero-engine |
| CN113776794A (en) * | 2021-08-13 | 2021-12-10 | 昆明理工大学 | Fault diagnosis method, device and system for embedded edge computing |
| CN113776794B (en) * | 2021-08-13 | 2023-04-28 | 昆明理工大学 | Fault diagnosis method, device and system for embedded edge calculation |
| CN113837245A (en) * | 2021-09-06 | 2021-12-24 | 西安理工大学 | Flexible printing electronic equipment core unit fault diagnosis method |
| CN113837245B (en) * | 2021-09-06 | 2024-04-09 | 西安理工大学 | Flexible printed electronic equipment core unit fault diagnosis method |
| CN114252716A (en) * | 2021-12-06 | 2022-03-29 | 深圳供电局有限公司 | Power grid fault diagnosis method and device and storage medium |
| CN114893428A (en) * | 2022-04-02 | 2022-08-12 | 上海大学 | Fine pre-diagnosis and accurate operation and maintenance device and system for ventilator |
| CN114893428B (en) * | 2022-04-02 | 2023-02-21 | 上海大学 | Fine pre-diagnosis and accurate operation and maintenance device and system for ventilator |
| CN114910733A (en) * | 2022-07-15 | 2022-08-16 | 深圳益实科技有限公司 | Display fault intelligent diagnosis analysis system based on artificial intelligence |
| CN114910733B (en) * | 2022-07-15 | 2022-09-30 | 深圳益实科技有限公司 | Display fault intelligent diagnosis analysis system based on artificial intelligence |
| CN117189720A (en) * | 2023-09-14 | 2023-12-08 | 成都飞航智云科技有限公司 | Fault diagnosis method for hydraulic system of aircraft |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113155469A (en) | Engine fault diagnosis alarm system and device | |
| CN112085930B (en) | Intelligent monitoring and early warning system and method for generator set | |
| JP5530020B1 (en) | Abnormality diagnosis system and abnormality diagnosis method | |
| CN107168201B (en) | A kind of real-time watch device operation management system of threst stand | |
| JP5530045B1 (en) | Health management system and health management method | |
| JP5431235B2 (en) | Equipment condition monitoring method and apparatus | |
| US11138817B2 (en) | Vehicle system prognosis device and method | |
| US9483049B2 (en) | Anomaly detection and diagnosis/prognosis method, anomaly detection and diagnosis/prognosis system, and anomaly detection and diagnosis/prognosis program | |
| US6999884B2 (en) | Bearing anomaly detection and location | |
| EP1297313B1 (en) | Monitoring the health of a power plant | |
| CN109186744B (en) | Turbine blade crack evaluation system and turbine blade crack early warning method | |
| CN113776794B (en) | Fault diagnosis method, device and system for embedded edge calculation | |
| CN113400652A (en) | 3D printer monitoring and diagnosing knowledge base device and system based on vibration signals | |
| US20120271587A1 (en) | Equipment status monitoring method, monitoring system, and monitoring program | |
| CN111639467A (en) | Aero-engine service life prediction method based on long-term and short-term memory network | |
| JP2015172945A (en) | Facility state monitoring method and apparatus for the same | |
| CN103839302A (en) | Intelligent inspection system and method in thermal power plant | |
| CN111946559B (en) | Method for detecting structures of wind turbine foundation and tower | |
| CN103541948B (en) | Test stand for hydraulic element Distributed Status Monitoring network system | |
| CN110320041B (en) | Multi-bearing embedded state monitoring system of large-scale vibrating table | |
| CN113341879A (en) | Data acquisition and monitoring method and system for numerical control machine tool | |
| CN117215258A (en) | Numerical control machine tool real-time state monitoring system and method based on Flink | |
| CN2890860Y (en) | Real-time detection and diagnosis system torsional vibration | |
| CN109299201A (en) | Power plant's production subsystem method for monitoring abnormality and device based on two-phase analyzing method | |
| JP6641056B1 (en) | Device abnormality diagnosis method and device abnormality diagnosis system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210723 |
|
| RJ01 | Rejection of invention patent application after publication |