CN114021621A - Fault diagnosis method, system, storage medium and edge computing device - Google Patents
Fault diagnosis method, system, storage medium and edge computing device Download PDFInfo
- Publication number
- CN114021621A CN114021621A CN202111192866.3A CN202111192866A CN114021621A CN 114021621 A CN114021621 A CN 114021621A CN 202111192866 A CN202111192866 A CN 202111192866A CN 114021621 A CN114021621 A CN 114021621A
- Authority
- CN
- China
- Prior art keywords
- fault
- fault diagnosis
- state
- model
- data
- 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 112
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000012549 training Methods 0.000 claims description 41
- 238000012795 verification Methods 0.000 claims description 33
- 238000012806 monitoring device Methods 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 9
- 230000008878 coupling Effects 0.000 claims description 6
- 238000010168 coupling process Methods 0.000 claims description 6
- 238000005859 coupling reaction Methods 0.000 claims description 6
- 239000000725 suspension Substances 0.000 claims description 6
- 230000008569 process Effects 0.000 description 11
- 238000004364 calculation method Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000005309 stochastic process Methods 0.000 description 3
- 230000007704 transition Effects 0.000 description 3
- 238000004140 cleaning Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000012535 impurity Substances 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/29—Graphical models, e.g. Bayesian networks
- G06F18/295—Markov models or related models, e.g. semi-Markov models; Markov random fields; Networks embedding Markov models
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
A fault diagnosis method, system, storage medium and edge computing device, wherein the method comprises: when a system to be diagnosed runs, acquiring real-time observation data reflecting the states of a plurality of devices of the system; sending the real-time observation data to a trained fault diagnosis model constructed based on a hidden Markov model; and performing fault diagnosis through the fault diagnosis model, and outputting a diagnosis result.
Description
Technical Field
The present disclosure relates to fault automatic diagnosis technologies, and more particularly, to a fault diagnosis method, system, storage medium, and edge computing device.
Background
The platform safety door serving as a component in the platform safety door system is arranged along the edge of a platform to isolate a train from a platform waiting room, so that not only can the danger caused by the falling or jumping off of passengers from a track be prevented, but also a safe, comfortable and beautiful riding environment can be provided for the passengers. The platform safety door system is used as a public safety protection system, is widely applied to overhead, ground and underground platforms in rail transit such as subways, light rails, intercity railways and the like, and has the functions of energy conservation, environmental protection and safety.
The safety, the reliability and the operation and maintenance rapidness of the platform safety door system directly influence the transportation safety and the efficiency of rail transit, so that the fault diagnosis is carried out on the platform safety door system, and the normal operation of the platform safety door system is guaranteed to be very important.
Disclosure of Invention
The application provides a fault diagnosis method, a fault diagnosis system, a storage medium and an edge calculation device, which can accurately diagnose faults.
The fault diagnosis method provided by the embodiment of the application comprises the following steps:
when a system to be diagnosed runs, acquiring real-time observation data reflecting the states of a plurality of devices of the system;
sending the real-time observation data to a trained fault diagnosis model constructed based on a hidden Markov model;
and performing fault diagnosis through the fault diagnosis model, and outputting a diagnosis result.
Optionally, the method for training the fault diagnosis model includes:
acquiring multiple groups of real-time observation data of each device in the multiple devices under different states to obtain a training data set and a verification data set;
training the fault diagnosis model through the training data set;
and verifying the trained fault diagnosis model through the verification data set.
Optionally, the training data included in the training data set is a two-dimensional matrix, different horizontal columns of the matrix represent the acquired device state data monitored by the data monitoring device on different unit time nodes, and different vertical rows of the matrix represent different data monitoring devices;
the verification data contained in the verification data set is training data carrying a label, and the label is used for indicating the state of each device.
Optionally, training the fault diagnosis model through the training data set includes:
training through the training data set to generate a device fault state classifier comprising a plurality of device fault models, wherein each device fault model is described by a hidden markov model λ ═ ([ pi ], a, b), the device fault models being used to reflect device states, the device states comprising: a normal state and a fault state;
wherein pi is the initial state of the equipment; a is the probability that the equipment is changed from the state of the current moment to a different state of the next moment; and b is the probability of the appearance of each observation state corresponding to the state of the equipment at the current moment.
Optionally, verifying the trained fault diagnosis model by the verification data set includes:
sending verification data in the verification data set to the equipment fault state classifier;
obtaining a fault prediction result according to an output result of each equipment fault model in the equipment fault state classifier;
comparing the obtained fault prediction results with the labels carried by the verification data respectively;
obtaining the fault diagnosis accuracy of the fault diagnosis model according to the comparison result;
and when the fault diagnosis accuracy reaches a preset threshold value, the verification is passed.
Optionally, performing fault diagnosis by using the fault diagnosis model, and outputting a diagnosis result, including:
respectively sending the real-time observation data to L equipment fault models of the equipment fault state classifier to obtain { lambda1，λ2，...，λL}；
According to Pl＝logP(O/λl) L e {1, 2, 3.., L } calculates the log-likelihood value that each device fault model produces observed data, O ═ O · O ·1，O2，…，OnN is the total number of the real-time observed data;
and selecting the output result of the equipment fault model generating the maximum log likelihood value as a diagnosis result.
Optionally, the system is a platform door system in rail transit;
the plurality of devices comprising the system may include any of:
the platform door comprises a platform door component, a sliding door unlocking device, a screw rod, a coupling and a suspension device;
the platform door states include: the roughness value of the component is normal or the roughness of the component is high;
the sliding door states include: normal or foreign body jamming;
the sliding door unlocking device state includes: normal or functional failure;
the screw state includes: normal or worn;
the coupling state includes: normal or loose;
the suspension state includes: normal or loose.
Embodiments of the present application also provide a computer readable storage medium storing one or more programs, which are executable by one or more processors to implement a method as described in any of the preceding.
An embodiment of the present application further provides an edge computing apparatus, including: a memory storing a program which, when read and executed by the processor, implements a method as in any preceding claim.
An embodiment of the present application further provides a fault diagnosis system, where the system includes:
the edge computing device as described above;
and the central server is used for storing and displaying the diagnosis result output by the edge computing device.
An embodiment of the present application further provides a fault diagnosis system, where the system includes:
the edge computing device is set to obtain real-time observation data reflecting the states of a plurality of devices of the system to be diagnosed when the system to be diagnosed runs; sending the real-time observation data to a central server;
and the central server is used for sending the real-time observation data to a trained fault diagnosis model constructed based on a hidden Markov model, carrying out fault diagnosis through the fault diagnosis model and outputting a diagnosis result.
Because the hidden Markov model has reliable calculation performance, better classification and identification capabilities and small sample data amount required by learning (such as the sample data amount far smaller than that required by neural network learning), the fault diagnosis model constructed based on the hidden Markov model in the scheme disclosed by the application can accurately identify the fault type and is simple to implement. Furthermore, hidden markov models are probabilistic based stochastic processes whose learning and classification interpretability is far superior to neural networks.
Additional features and advantages of the application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the application. Other advantages of the present application may be realized and attained by the instrumentalities and combinations particularly pointed out in the specification and the drawings.
Drawings
The accompanying drawings are included to provide an understanding of the present disclosure and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the examples serve to explain the principles of the disclosure and not to limit the disclosure.
Fig. 1 is a flowchart of a fault diagnosis method provided in an embodiment of the present application;
FIG. 2 is a flow chart of a method for training a fault diagnosis model according to an embodiment of the present disclosure;
FIG. 3 is a flowchart of a method for validating a trained fault diagnosis model through a validation data set according to an embodiment of the present application;
fig. 4 is a flowchart of a method for performing fault diagnosis by a fault diagnosis model according to an embodiment of the present application;
FIG. 5 is a block diagram of an edge computing device according to an embodiment of the present disclosure;
fig. 6 is a component diagram of a fault diagnosis system according to an embodiment of the present application.
Detailed Description
The present application describes embodiments, but the description is illustrative rather than limiting and it will be apparent to those of ordinary skill in the art that many more embodiments and implementations are possible within the scope of the embodiments described herein. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are possible. Any feature or element of any embodiment may be used in combination with or instead of any other feature or element in any other embodiment, unless expressly limited otherwise.
The present application includes and contemplates combinations of features and elements known to those of ordinary skill in the art. The embodiments, features and elements disclosed in this application may also be combined with any conventional features or elements to form a unique inventive concept as defined by the claims. Any feature or element of any embodiment may also be combined with features or elements from other inventive aspects to form yet another unique inventive aspect, as defined by the claims. Thus, it should be understood that any of the features shown and/or discussed in this application may be implemented alone or in any suitable combination. Accordingly, the embodiments are not limited except as by the appended claims and their equivalents. Furthermore, various modifications and changes may be made within the scope of the appended claims.
Further, in describing representative embodiments, the specification may have presented the method and/or process as a particular sequence of steps. However, to the extent that the method or process does not rely on the particular order of steps set forth herein, the method or process should not be limited to the particular sequence of steps described. Other orders of steps are possible as will be understood by those of ordinary skill in the art. Therefore, the particular order of the steps set forth in the specification should not be construed as limitations on the claims. Further, the claims directed to the method and/or process should not be limited to the performance of their steps in the order written, and one skilled in the art can readily appreciate that the sequences may be varied and still remain within the spirit and scope of the embodiments of the present application.
An embodiment of the present application provides a fault diagnosis method, as shown in fig. 1, the method includes the following steps:
step S101, when a system to be diagnosed runs, acquiring real-time observation data reflecting the states of a plurality of devices of the system;
step S102, the real-time observation data are sent to a trained fault diagnosis model constructed based on a hidden Markov model;
step S103, fault diagnosis is carried out through the fault diagnosis model, and a diagnosis result is output.
Because the hidden Markov model has reliable calculation performance, better classification and identification capabilities and small sample data amount required by learning (such as the sample data amount far smaller than that required by neural network learning), the fault diagnosis model constructed based on the hidden Markov model can accurately identify the fault type and is simple to implement. Furthermore, hidden markov models are probabilistic based stochastic processes whose learning and classification interpretability is far superior to neural networks.
In an exemplary embodiment, the method for training the fault diagnosis model, as shown in fig. 2, includes:
step S201, acquiring multiple groups of real-time observation data of each device in different states to obtain a training data set and a verification data set; the different states of the equipment refer to that the equipment is in a normal state or a fault state; the fault condition may in turn comprise one or more; the training data set and the verification data set can jointly form a sample data set; for example, the data volume ratio of the training data set and the verification data set in the sample data set may be 7: 3;
step S202, training the fault diagnosis model through the training data set;
step S203 verifies the trained fault diagnosis model by the verification data set.
In an exemplary embodiment, the training data contained in the training data set is a two-dimensional matrix, different horizontal columns of the matrix represent the acquired device state data monitored by the data monitoring device on different unit time nodes, and different vertical rows of the matrix represent different data monitoring devices;
different data monitoring devices can be used for acquiring state data of different devices (such as the data monitoring device 1 monitors the state of the device 1, and the data monitoring device 2 monitors the state of the device 2), or different data monitoring devices can also be used for acquiring state data of the same device (such as the data monitoring devices 1 and 2 monitor the state of the device 1 together); the data monitoring device can be a sensor, and the data to be monitored can be in various forms, such as pressure data, light data and the like, and the types of the corresponding sensors can also be various;
different training data forming the training data set can correspond to data acquired under different equipment states;
the verification data contained in the verification data set is training data carrying a label, and the label is used for indicating the state of each device.
In an exemplary embodiment, step S202 trains the fault diagnosis model through the training data set, including:
training through the training data set to generate a device fault state classifier comprising a plurality of device fault models, wherein each device fault model is described by a hidden markov model λ ═ ([ pi ], a, b), the device fault models being used to reflect device states, the device states comprising: a normal state and a fault state; wherein pi is the initial state of the equipment; a is the probability that the equipment is changed from the state of the current moment to a different state of the next moment; and b is the probability of the appearance of each observation state corresponding to the state of the equipment at the current moment.
The parameters in the hidden markov model λ ═ (pi, a, b) can be determined using a maximum likelihood estimation method, as shown in the following example:
let the training data set time t be in state i, and the frequency of transition from time t +1 to state j be AijThen the state transition probability is aijThe estimation of (d) is:
n represents the number of states that each device may have;
let B be the frequency of the training data set at time t in state j and observed as kjkThen the probability b that the state is j observed as state kj(k) The estimation of (d) is:
m represents the number of possible observations corresponding to each state;
probability of initial state iiThe estimation of (d) is: the frequency of the initial state i in the training dataset.
From the above parameter determination process, it can be found that the hidden markov model is a double stochastic process: one is a hidden state transition random process, and the other is a random process for generating an observation value in each hidden state. The hidden state is used for representing the internal state, and the observation value represents the characteristic shown in the hidden state. This structure of hidden markov models makes it suitable for modeling dynamic process time series, especially non-stationary time series with poor reproducibility of repetition. Furthermore, hidden markov models can process variable length feature sequences, while most of the existing model classification methods can only process fixed length feature sequences. Therefore, the fault diagnosis model is established based on the hidden Markov model, the fault diagnosis method and the fault diagnosis system can be suitable for various fault diagnosis scenes, and the applicability is wide.
In an exemplary embodiment, step S203 verifies the trained fault diagnosis model through the verification data set, as shown in fig. 3, including:
step S2031, sending the verification data in the verification data set to the equipment fault state classifier;
step S2032, a fault prediction result is obtained according to the output result of each equipment fault model in the equipment fault state classifier;
step S2033 is to compare the obtained failure prediction results with the labels carried by the verification data respectively;
step S2034 of obtaining the fault diagnosis accuracy of the fault diagnosis model according to the comparison result;
and step S2035, when the fault diagnosis accuracy reaches a preset threshold value, the verification is passed.
In an exemplary embodiment, obtaining the fault diagnosis accuracy of the fault diagnosis model according to the comparison result includes: determining the fault diagnosis accuracy according to comparison results of a preset number of verification data in the verification data set, for example, taking an arithmetic mean value of the comparison results of the preset number as the fault diagnosis accuracy; the preset number can be 1 or more, and the larger the value of the preset number is, the more samples are used for obtaining the fault diagnosis accuracy, and the higher the reliability of the obtained fault accuracy is; however, too large a preset number also increases the amount of calculation, which results in time-consuming model verification.
In an exemplary embodiment, step S103 performs fault diagnosis by the fault diagnosis model, and outputs a diagnosis result, as shown in fig. 4, including:
step S1031 of the method is to respectively send the real-time observation data to L equipment fault models of the equipment fault state classifier to obtain { lambda1，λ2，...，λL}；
Step S1032 is according to Pl＝logP(O/λl) L e {1, 2, 3.., L } calculates the log-likelihood value that each device fault model produces observed data, O ═ O · O ·1，O2，…，OnN is the total number of the real-time observed data;
step S1033 selects an output result of the equipment failure model that produces the maximum log likelihood value as a diagnosis result.
In an exemplary embodiment, the method further comprises: after acquiring real-time observation data, processing the real-time observation data, wherein the processing comprises the following steps: at least one of a cleaning and a normalization process;
and performing subsequent operations based on the processed real-time observation data, such as sending the processed real-time observation data to a trained fault diagnosis model constructed based on a hidden Markov model, or obtaining a training data set and a verification data set based on the processed real-time observation data.
According to the embodiment, impurity data in the real-time observation data can be eliminated through cleaning, and the fault diagnosis efficiency is improved by reducing the processing flow of the impurity data; through normalization processing, adverse effects of data with different dimensions and dimension units can be eliminated, and the accuracy of fault diagnosis is improved.
The fault diagnosis method can be applied to a platform safety door system of rail transit.
The above-described fault diagnosis method will be described below by taking a platform safety door system as an example.
Firstly, a well-trained fault diagnosis model constructed based on a hidden Markov model is obtained for a platform safety door system:
the fault diagnosis model comprises L equipment fault models, each equipment fault model is described by a hidden Markov model lambda (pi, a, b), and the equipment fault models are used for reflecting equipment states;
the L devices include: the platform door comprises a platform door component, a sliding door unlocking device, a screw rod, a coupling and a suspension device;
the state of each device includes a normal state and a fault state; wherein the content of the first and second substances,
the platform door states include: the roughness value of the component is normal or the roughness of the component is high;
the sliding door states include: normal or foreign body jamming;
the sliding door unlocking device state includes: normal or functional failure;
the screw state includes: normal or worn;
the coupling state includes: normal or loose;
the suspension state includes: normal or loose.
Acquiring real-time observation data O ═ { O } reflecting the states of a plurality of devices of a platform safety door system to be diagnosed when the platform safety door system to be diagnosed runs1，O2，…，On}；
The real-time observation data O is set as { O ═ O1，O2，…，OnSending the L equipment fault models to the equipment fault state classifier respectively to obtain { lambda }1，λ2，...，λL}；
According to Pl＝logP(O/λl) L belongs to {1, 2, 3.,. L }, calculating the log-likelihood value of each equipment fault model generating observation data;
and selecting the output result of the equipment fault model generating the maximum log likelihood value as a diagnosis result.
Embodiments of the present application also provide a computer-readable storage medium storing one or more programs, which are executable by one or more processors to implement the method according to any of the foregoing embodiments.
An embodiment of the present application further provides an edge computing apparatus, as shown in fig. 5, including: a memory 501 and a processor 502, the memory 501 storing a program which, when read and executed by the processor 502, implements a method as in any of the previous embodiments.
An embodiment of the present application further provides a fault diagnosis system, as shown in fig. 6, the system includes:
the edge calculation means 601 as described previously;
a central server 602 configured to store and display the diagnosis result output by the edge computing device 601.
According to the embodiment of the application, the edge computing device undertakes data computing work, computing resources of the edge computing device are fully utilized, and computing pressure of the central server is reduced. For a scene with high demand on instantaneity, the scheme can feed back the diagnosis result more quickly.
An embodiment of the present application further provides a fault diagnosis system, where the system includes:
the edge computing device is set to obtain real-time observation data reflecting the states of a plurality of devices of the system to be diagnosed when the system to be diagnosed runs; sending the real-time observation data to a central server;
and the central server is used for sending the real-time observation data to a trained fault diagnosis model constructed based on a hidden Markov model, carrying out fault diagnosis through the fault diagnosis model and outputting a diagnosis result.
The edge computing device in the embodiment finishes data acquisition, and the central server finishes data analysis and fault diagnosis; the computing pressure of the edge computing device is not large and other data streams can be processed concurrently.
It will be understood by those of ordinary skill in the art that all or some of the steps of the methods, systems, functional modules/units in the devices disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. In a hardware implementation, the division between functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed by several physical components in cooperation. Some or all of the components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as is well known to those of ordinary skill in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media as known to those skilled in the art.
Claims (11)
1. A fault diagnosis method comprising:
when a system to be diagnosed runs, acquiring real-time observation data reflecting the states of a plurality of devices of the system;
sending the real-time observation data to a trained fault diagnosis model constructed based on a hidden Markov model;
and performing fault diagnosis through the fault diagnosis model, and outputting a diagnosis result.
2. The method of claim 1,
a method of training the fault diagnosis model, comprising:
acquiring multiple groups of real-time observation data of each device in the multiple devices under different states to obtain a training data set and a verification data set;
training the fault diagnosis model through the training data set;
and verifying the trained fault diagnosis model through the verification data set.
3. The method of claim 2,
the training data contained in the training data set is a two-dimensional matrix, different horizontal columns of the matrix represent the acquired equipment state data monitored by the data monitoring device on different unit time nodes, and different vertical rows of the matrix represent different data monitoring devices;
the verification data contained in the verification data set is training data carrying a label, and the label is used for indicating the state of each device.
4. The method of claim 3,
training the fault diagnosis model with the training data set, including:
training through the training data set to generate a device fault state classifier comprising a plurality of device fault models, wherein each device fault model is described by a hidden markov model λ ═ ([ pi ], a, b), the device fault models being used to reflect device states, the device states comprising: a normal state and a fault state;
wherein pi is the initial state of the equipment; a is the probability that the equipment is changed from the state of the current moment to a different state of the next moment; and b is the probability of the appearance of each observation state corresponding to the state of the equipment at the current moment.
5. The method of claim 4,
verifying the trained fault diagnosis model by the verification data set, comprising:
sending verification data in the verification data set to the equipment fault state classifier;
obtaining a fault prediction result according to an output result of each equipment fault model in the equipment fault state classifier;
comparing the obtained fault prediction results with the labels carried by the verification data respectively;
obtaining the fault diagnosis accuracy of the fault diagnosis model according to the comparison result;
and when the fault diagnosis accuracy reaches a preset threshold value, the verification is passed.
6. The method of claim 5,
carrying out fault diagnosis through the fault diagnosis model and outputting a diagnosis result, wherein the fault diagnosis comprises the following steps:
respectively sending the real-time observation data to L equipment fault models of the equipment fault state classifier to obtain { lambda1，λ2，...，λL}；
According to Pl＝log P(O/λl) L e {1, 2, 3.., L } calculates the log-likelihood value that each device fault model produces observed data, O ═ O · O ·1，O2，...，OnN is the total number of the real-time observed data;
and selecting the output result of the equipment fault model generating the maximum log likelihood value as a diagnosis result.
7. The method according to any one of claims 4 to 6,
the system is a platform door system in rail transit;
the plurality of devices comprising the system may include any of:
the platform door comprises a platform door component, a sliding door unlocking device, a screw rod, a coupling and a suspension device;
the platform door states include: the roughness value of the component is normal or the roughness of the component is high;
the sliding door states include: normal or foreign body jamming;
the sliding door unlocking device state includes: normal or functional failure;
the screw state includes: normal or worn;
the coupling state includes: normal or loose;
the suspension state includes: normal or loose.
8. A computer readable storage medium storing one or more programs, the one or more programs being executable by one or more processors to implement the method of any of claims 1 to 7.
9. An edge computing device, comprising: a memory and a processor, the memory storing a program that, when read and executed by the processor, implements the method of any of claims 1 to 7.
10. A fault diagnosis system, characterized in that the system comprises:
the edge computing device of claim 9;
and the central server is used for storing and displaying the diagnosis result output by the edge computing device.
11. A fault diagnosis system, characterized in that the system comprises:
the edge computing device is set to obtain real-time observation data reflecting the states of a plurality of devices of the system to be diagnosed when the system to be diagnosed runs; sending the real-time observation data to a central server;
and the central server is used for sending the real-time observation data to a trained fault diagnosis model constructed based on a hidden Markov model, carrying out fault diagnosis through the fault diagnosis model and outputting a diagnosis result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111192866.3A CN114021621A (en) | 2021-10-13 | 2021-10-13 | Fault diagnosis method, system, storage medium and edge computing device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111192866.3A CN114021621A (en) | 2021-10-13 | 2021-10-13 | Fault diagnosis method, system, storage medium and edge computing device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114021621A true CN114021621A (en) | 2022-02-08 |
Family
ID=80056036
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111192866.3A Pending CN114021621A (en) | 2021-10-13 | 2021-10-13 | Fault diagnosis method, system, storage medium and edge computing device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114021621A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114996258A (en) * | 2022-06-23 | 2022-09-02 | 中铁第四勘察设计院集团有限公司 | Contact network fault diagnosis method based on data warehouse |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102073318A (en) * | 2009-11-17 | 2011-05-25 | 通用汽车环球科技运作公司 | Fault diagnosis and prognosis using diagnostic trouble code markov chains |
| CN106483405A (en) * | 2016-09-30 | 2017-03-08 | 江苏大学 | The method for diagnosing faults of the NPC photovoltaic DC-to-AC converter based on hidden Markov model |
| CN106803101A (en) * | 2016-12-30 | 2017-06-06 | 北京交通大学 | Odometer method for diagnosing faults based on HMM |
| CN110262463A (en) * | 2019-07-09 | 2019-09-20 | 河海大学常州校区 | A kind of rail traffic platform door fault diagnosis system based on deep learning |
-
2021
- 2021-10-13 CN CN202111192866.3A patent/CN114021621A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102073318A (en) * | 2009-11-17 | 2011-05-25 | 通用汽车环球科技运作公司 | Fault diagnosis and prognosis using diagnostic trouble code markov chains |
| CN106483405A (en) * | 2016-09-30 | 2017-03-08 | 江苏大学 | The method for diagnosing faults of the NPC photovoltaic DC-to-AC converter based on hidden Markov model |
| CN106803101A (en) * | 2016-12-30 | 2017-06-06 | 北京交通大学 | Odometer method for diagnosing faults based on HMM |
| CN110262463A (en) * | 2019-07-09 | 2019-09-20 | 河海大学常州校区 | A kind of rail traffic platform door fault diagnosis system based on deep learning |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114996258A (en) * | 2022-06-23 | 2022-09-02 | 中铁第四勘察设计院集团有限公司 | Contact network fault diagnosis method based on data warehouse |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111124840B (en) | Method and device for predicting alarm in business operation and maintenance and electronic equipment | |
| Li et al. | Trend modeling for traffic time series analysis: An integrated study | |
| CN109978379B (en) | Time series data abnormity detection method and device, computer equipment and storage medium | |
| CN111178523B (en) | Behavior detection method and device, electronic equipment and storage medium | |
| US20190228312A1 (en) | Unsupervised model building for clustering and anomaly detection | |
| US20200272823A1 (en) | Weakly-Supervised Action Localization by Sparse Temporal Pooling Network | |
| CN109447263B (en) | Space abnormal event detection method based on generation of countermeasure network | |
| CN111914873A (en) | Two-stage cloud server unsupervised anomaly prediction method | |
| CN111460392B (en) | Magnetic suspension train and suspension system fault detection method and system thereof | |
| CN111026915A (en) | Video classification method, video classification device, storage medium and electronic equipment | |
| KR102274069B1 (en) | Apparatus and method for generating prediction model | |
| CN114297036B (en) | Data processing method, device, electronic equipment and readable storage medium | |
| CN111901296A8 (en) | Network attack behavior detection method and detection system | |
| CN109117774B (en) | Multi-view video anomaly detection method based on sparse coding | |
| CN109357747A (en) | A kind of identification of online train and speed estimation method based on fiber-optic vibration signal | |
| Pereira et al. | Using deep autoencoders for in-vehicle audio anomaly detection | |
| CN112967252A (en) | Rail vehicle machine sense hanger assembly bolt loss detection method | |
| CN114021621A (en) | Fault diagnosis method, system, storage medium and edge computing device | |
| CN114528190A (en) | Single index abnormality detection method and device, electronic equipment and readable storage medium | |
| CN114860542A (en) | Trend prediction model optimization method, trend prediction model optimization device, electronic device, and medium | |
| CN113589230A (en) | Target sound source positioning method and system based on joint optimization network | |
| CN113758709A (en) | Rolling bearing fault diagnosis method and system combining edge calculation and deep learning | |
| Latyshev | Sensor Data Preprocessing, Feature Engineering and Equipment Remaining Lifetime Forecasting for Predictive Maintenance. | |
| CN116594057A (en) | Earthquake early warning method and device based on deep learning and edge calculation | |
| CN111160594A (en) | Method and device for estimating arrival time and storage medium |
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 |