CN109711062A - A kind of equipment fault diagnosis method and device based on cloud service - Google Patents
A kind of equipment fault diagnosis method and device based on cloud service Download PDFInfo
- Publication number
- CN109711062A CN109711062A CN201811626867.2A CN201811626867A CN109711062A CN 109711062 A CN109711062 A CN 109711062A CN 201811626867 A CN201811626867 A CN 201811626867A CN 109711062 A CN109711062 A CN 109711062A
- Authority
- CN
- China
- Prior art keywords
- fault
- vectors
- diagnosis
- failure
- elm
- 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
Landscapes
- Test And Diagnosis Of Digital Computers (AREA)
Abstract
This application discloses a kind of equipment fault diagnosis method and device based on cloud service, wherein the method include that S1: be based on cloud memory technology, obtain in real time include automatic verification system equipment operating data fault diagnosis symptom vector;S2: fault vectors are obtained by fuzzy reasoning operation according to the fault diagnosis symptom vector got;S3: fault vectors are input to the ELM secondary failure diagnostic model after the completion of training and carry out secondary failure diagnosis, obtain fault type diagnostic result.The application is combined based on the data management performance of cloud service with fuzzy reasoning and ELM the two-stage Fault Tree Diagnosis Decision model formed, it solves in the automatic gauge detection system environment of the complexity such as multi-source, multidimensional, isomery, the conventional automated calibrating center technical problem low based on the equipment fault diagnosis accuracy rate of cloud service.
Description
Technical field
This application involves device intelligence management domain more particularly to a kind of equipment fault diagnosis method based on cloud service and
Device.
Background technique
Extreme learning machine (Extreme Learning Machine, ELM) is a kind of for single layer feedforward neural network
The machine learning algorithm of (Single Layer Feedforward neuron Network, SLFN) design, is mainly characterized by
Hidden layer node parameter can be random or given by man and not need to adjust, and learning process only needs to calculate output weight.ELM
Have the advantages that learning efficiency height and generalization ability are strong.
On the one hand coming into operation for a large amount of automatic calibration equipment improves the calibrating efficiency of measuring equipment, but also make simultaneously
Increase sharply at equipment O&M related data total amount, difficult management the problems such as, and the complicated feature such as show multi-source, multidimensional, isomery,
Existing Fault Tree Diagnosis Decision method and system is analyzed primarily directed to collected single data, video source information, difficult
To reach the Intelligent fault diagnosis to the high accuracy of equipment, especially to multi-source, multidimensional, isomery etc., structure is complicated, data
The accuracy rate for measuring big data information can further decrease, and result in the low skill of existing device intelligence fault diagnosis accuracy rate
Art problem.
Summary of the invention
This application provides a kind of equipment fault diagnosis method and device based on cloud service is used for existing device intelligence
Change the low technical problem of fault diagnosis accuracy rate.
The application first aspect provides a kind of equipment fault diagnosis method based on cloud service, comprising:
S1: being based on cloud memory technology, and acquisition in real time includes that the failure of the equipment operating data of automatic verification system is examined
Disconnected symptom vector;
S2: fault vectors are obtained by fuzzy reasoning operation according to the fault diagnosis symptom vector got;
S3: the fault vectors are input to the ELM secondary failure diagnostic model progress secondary failure after the completion of training and are examined
It is disconnected, obtain fault type diagnostic result.
Preferably, the step S2 is specifically included:
S21: the fault vectors in fuzzy reasoning operational model are passed through according to the fault diagnosis symptom vector got
The fault diagnosis symptom vector is converted into corresponding fault vectors, wherein the fault vectors conversion formula by conversion formula
Are as follows:
In formula, L is fault diagnosis symptom vector set, and F is fault vectors set, and R (E) is the equipment in expert knowledge library
The fuzzy reasoning matrix of environmental variance E, rij(E) the symptom vector l when environmental variance is E is indicatedjLead to failure type fi
Probability.
Preferably, the step S3 is specifically included:
S31: the fault vectors are input to the ELM secondary failure diagnostic model after the completion of training;
S32: carrying out secondary failure diagnosis by the fault type judgment formula in the ELM secondary failure diagnostic model,
And fault type diagnostic result is exported, wherein the fault type judgment formula are as follows:
In formula,For (t+1) secondary fault type judging result,For (t+1) secondary input failure to
Amount, β (t) are that the t times network exports weight vector, ωiFor the weight matrix for connecting input layer and hidden layer, biFor the inclined of hidden layer
Matrix is set, g is sigmoid function, and i is prototype network hidden layer number.
Preferably, the ELM secondary failure diagnostic model training complete before, it is described according to the fault diagnosis symptom to
Amount is obtained by fuzzy reasoning operation after fault vectors further include:
S4: the corresponding failure classes of the fault vectors are determined by fault threshold decision procedure according to the fault vectors
Type diagnostic result;
S5: using the fault type diagnostic result and the fault diagnosis symptom vector as model training input variable,
For training the ELM secondary failure diagnostic model.
Preferably, the step S4 is specifically included:
S41: according to the one-to-one failure fixed threshold of various fault types, by the fault vectors and the failure
Fixed threshold is compared, and filtering out first beyond the failure fixed threshold transfinites fault vectors, and obtains fault type
Preliminary judgement result;
S42: it according to the fault type preliminary judgement as a result, determining formula in conjunction with dynamic threshold, transfinites to described first
Fault vectors carry out secondary judgement, and filtering out second beyond the dynamic threshold transfinites fault vectors, obtain fault type most
Result is determined eventually.
The application second aspect provides a kind of equipment fault diagnosis device based on cloud service, comprising:
Symptom vector acquiring unit, for being based on cloud memory technology, obtaining in real time includes setting for automatic verification system
The fault diagnosis symptom vector of standby operation data;
Fault vectors converting unit, for passing through fuzzy reasoning operation according to the fault diagnosis symptom vector got
Obtain fault vectors;
Failure diagnosis unit, the ELM secondary failure diagnostic model for being input to the fault vectors after the completion of training
Secondary failure diagnosis is carried out, fault type diagnostic result is obtained.
Preferably, the fault vectors converting unit is specifically used for:
Pass through the fault vectors conversion in fuzzy reasoning operational model according to the fault diagnosis symptom vector got
The fault diagnosis symptom vector is converted into corresponding fault vectors, wherein the fault vectors conversion formula by formula are as follows:
In formula, L is fault diagnosis symptom vector set, and F is fault vectors set, and R (E) is the equipment in expert knowledge library
The fuzzy reasoning matrix of environmental variance E, rij(E) the symptom vector l when environmental variance is E is indicatedjLead to failure type fi
Probability.
Preferably, the failure diagnosis unit specifically includes:
Input quantity obtains subelement, diagnoses for the fault vectors to be input to the ELM secondary failure after the completion of training
Model;
Fault diagnosis subelement, for by the fault type judgment formula in the ELM secondary failure diagnostic model into
The diagnosis of row secondary failure, and fault type diagnostic result is exported, wherein the fault type judgment formula are as follows:
In formula,For (t+1) secondary fault type judging result,For (t+1) secondary input failure to
Amount, β (t) are that the t times network exports weight vector, ωiFor the weight matrix for connecting input layer and hidden layer, biFor the inclined of hidden layer
Matrix is set, g is sigmoid function, and i is prototype network hidden layer number.
Preferably, the ELM secondary failure diagnostic model training complete before, it is described according to the fault diagnosis symptom to
Amount is obtained by fuzzy reasoning operation after fault vectors further include:
Tentative diagnosis unit, for according to the fault vectors, by fault threshold decision procedure, determine the failure to
Measure corresponding fault type diagnostic result;
Model training unit, for using the fault type diagnostic result and the fault diagnosis symptom vector as model
Training input variable, for training the ELM secondary failure diagnostic model.
Preferably, the tentative diagnosis unit specifically includes:
Fixed threshold determine subelement, according to the one-to-one failure fixed threshold of various fault types, will it is described therefore
Barrier vector be compared with the failure fixed threshold, filter out first beyond the failure fixed threshold transfinite failure to
Amount, and obtain fault type preliminary judgement result;
Dynamic threshold determines subelement, is used for according to the fault type preliminary judgement as a result, determining in conjunction with dynamic threshold
Formula carries out secondary judgement to described first fault vectors that transfinite, and filtering out second beyond the dynamic threshold transfinites failure
Vector obtains fault type and finally determines result.
As can be seen from the above technical solutions, the application has the following advantages:
This application provides a kind of equipment fault diagnosis methods based on cloud service, comprising: S1: storing skill based on cloud
Art, acquisition in real time includes the fault diagnosis symptom vector of the equipment operating data of automatic verification system, and stores and deposit beyond the clouds
It stores up in equipment;S2: fault vectors are obtained by fuzzy reasoning operation according to the fault diagnosis symptom vector got;S3:
The fault vectors are input to the ELM secondary failure diagnostic model after the completion of training and carry out secondary failure diagnosis, obtain failure
Type diagnostic result.
The two-stage Fault Tree Diagnosis Decision mould that data management performance of the application based on cloud service and fuzzy reasoning and ELM form
Type, which combines, to be solved in the automatic gauge detection system environment of the complexity such as multi-source, multidimensional, isomery, conventional automated calibrating
The center technical problem low based on the equipment fault diagnosis accuracy rate of cloud service.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of application without any creative labor, may be used also for those of ordinary skill in the art
To obtain other attached drawings according to these attached drawings.
Fig. 1 is a kind of process of one embodiment of the equipment fault diagnosis method based on cloud service provided by the present application
Schematic diagram;
Fig. 2 is a kind of process of second embodiment of the equipment fault diagnosis method based on cloud service provided by the present application
Schematic diagram;
Fig. 3 is a kind of process of the third embodiment of the equipment fault diagnosis method based on cloud service provided by the present application
Schematic diagram;
Fig. 4 is that a kind of construction of one embodiment of the equipment fault diagnosis device based on cloud service provided by the present application shows
It is intended to;
Fig. 5 is a kind of logical architecture figure of the equipment fault diagnosis method based on cloud service provided by the present application;
Fig. 6 is that a kind of secondary failure of equipment fault diagnosis method based on cloud service provided by the present application diagnoses decision
Extreme learning machine network structure.
Specific embodiment
The embodiment of the present application provides a kind of equipment fault diagnosis method and device based on cloud service, sets for existing
The standby low technical problem of Intelligent fault accuracy rate of diagnosis.
To enable present invention purpose, feature, advantage more obvious and understandable, below in conjunction with the application
Attached drawing in embodiment, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that disclosed below
Embodiment be only some embodiments of the present application, and not all embodiment.Based on the embodiment in the application, this field
Those of ordinary skill's all other embodiment obtained without making creative work belongs to the application protection
Range.
Referring to Fig. 1, the embodiment of the present application provides a kind of equipment fault diagnosis method based on cloud service, comprising:
101, it is based on cloud memory technology, acquisition in real time includes that the failure of the equipment operating data of automatic verification system is examined
Disconnected symptom vector;
It should be noted that the present embodiment is by unmanned each key link of measurement verification system and device location installation
Sensor acquires each type of structured, semi-structured or even unstructured data, and the sensing data to get is fault diagnosis
Symptom vector, while isomeric data is reached into cloud, and the data analysis and processing capacity powerful by cloud in real time, equipment
Health state evaluation and fault diagnosis, wherein fault diagnosis symptom vector is specially the practical fortune of equipment by automatic verification system
The characteristic vector data shown when row.
102, fault vectors are obtained by fuzzy reasoning operation according to the fault diagnosis symptom vector got;
It should be noted that the present embodiment be based on fuzzy reasoning mode, by will be made of real-time status data symptom to
It is first sent into after amount after fuzzy expert system carries out fuzzy reasoning and generates fault vectors.
103, fault vectors are input to the ELM secondary failure diagnostic model after the completion of training and carry out secondary failure diagnosis,
Obtain fault type diagnostic result.
It should be noted that will have the first order after the first order fuzzy reasoning by step 102 generates fault vectors
The fault vectors that fuzzy reasoning generates are sent to second level extreme learning machine fault diagnosis model (the ELM second level event for completing training
Hinder diagnostic model), further accident analysis is carried out, the fault type for obtaining the output of ELM secondary failure diagnostic model diagnoses knot
Fruit.
The two-stage Fault Tree Diagnosis Decision that data management performance of the present embodiment based on cloud service and fuzzy reasoning and ELM form
Model, which combines, to be solved in the automatic gauge detection system environment of the complexity such as multi-source, multidimensional, isomery, conventional automated inspection
The low technical problem of the equipment fault diagnosis accuracy rate based on cloud service that centers.
The above are a kind of the detailed of one embodiment of the equipment fault diagnosis method based on cloud service provided by the present application
Describe in detail bright, here is a kind of the detailed of second embodiment of equipment fault diagnosis method based on cloud service provided by the present application
Explanation.
Referring to Fig. 2, the embodiment of the present application provides a kind of equipment fault diagnosis method based on cloud service, comprising:
201, it is based on cloud memory technology, acquisition in real time includes that the failure of the equipment operating data of automatic verification system is examined
Disconnected symptom vector;
202, the fault vectors conversion in fuzzy reasoning operational model is passed through according to the fault diagnosis symptom vector got
Fault diagnosis symptom vector is converted into corresponding fault vectors by formula;
It should be noted that the cloud service intelligent trouble diagnosis decision system of unmanned measurement verification central apparatus O&M passes through
In the sensor of each key link of unmanned detection system and device location installation, each type of structured, semi-structured of equipment is acquired
Then these isomeric datas are reached cloud by mMTC and uRLLC technology by even unstructured data in real time.
Real time data can be analyzed and processed beyond the clouds, will test the equipment of automated system individual field devices first
Operation data is symptom vector as the environmental variance common definition of acquisition variable (or failure symptom variable) and equipment operation,
Wherein, acquisition variable can be divided into " upper and lower bound variable " and " 0/1 variable " again: upper and lower bound variable is that equipment should be kept when working normally
Parameter in a certain normal range (NR), as connecting terminal temperature, cylinder pressure, ammeter deviations, measuring point exception low-frequency noise,
Measuring point exception high-frequency noise etc.;0/1 variable refers to should not be in the presence of when equipment works normally, such as equipment to be checked and detection device
Communicate unstable, conveyer belt extremely vibration etc..And environmental variance refers to the working environment of detection automated system device context, such as temperature
Degree, humidity etc..In addition for having the variable of lower bound requirement (such as cylinder pressure), Fuzzy Representation method uses formula (1) institute
The subordinating degree function shown.Wherein, x is upper and lower bound variable instantaneous value, and ub and lb respectively indicate the bound of its normal value, and l (x) is
The Fuzzy Representation of x.
Similarly, for the variable (such as terminal temperature) for thering is the upper bound to require, shown in subordinating degree function such as formula (2):
For 0/1 variable, when its corresponding situation appearance then takes 1 (table is abnormal), does not occur, take 0 (table is normal).It is fuzzy
Reasoning is based on equipment data library and expert knowledge library: equipment data library is static database, and storage automatic Verification measurement centre is each
The data information and parameter information of equipment;Expert knowledge library is dynamic data base, on the one hand, is initialized by expertise and is being
It is constantly evolved update in system operational process by self study;On the other hand, expert knowledge library is the function of environmental variance, i.e. equipment
The different environmental condition in scene corresponds to different expertises.In addition, environmental variance is sent directly into expert knowledge library, transfer corresponding
Expertise under environmental condition carries out decision.These data are finally subjected to storage and management, it can also be a variety of with its analysis
Small factor superposition is influenced caused by equipment, to realize equipment health state evaluation and fault diagnosis, and then is equipment
Maintenance work provides decision-making foundation.
The first order Fuzzy Inference Model of the present embodiment is input with fault diagnosis symptom vector, is output with fault vectors
(output is then fault type when second level diagnostic model not enabled).Expert knowledge library is used for the mould for acquiring, storing symptom to failure
Mapping relations are pasted, use professional technician has had experience to initialize, and continues to optimize in system operation
It updates.Indistinct logic computer is then for realizing the reasoning process of symptom vector to fault vectors.Remember the fault diagnosis disease of Fuzzy Representation
Shape vector is L=[l1, l2..., 1p] ', the resulting fault vectors of reasoning are F=[f1, f2..., fq] ', the fuzzy reasoning of L and F
Shown in relationship such as formula (3):
F=R (E) L (3)
In formula, it is the function of environmental variance E, such as formula (4) institute that R (E), which is the fuzzy reasoning matrix in expert knowledge library,
Show:
In formula, rij(E) it indicates when environmental variance is E, symptom ljLead to failure fiProbability.
203, the ELM secondary failure diagnostic model being input to fault vectors after the completion of training;
It should be noted that will have the first order after the first order fuzzy reasoning by step 202 generates fault vectors
The fault vectors that fuzzy reasoning generates are sent to second level extreme learning machine fault diagnosis model (the ELM second level event for completing training
Hinder diagnostic model), to carry out further accident analysis.
204, secondary failure diagnosis is carried out by the fault type judgment formula in ELM secondary failure diagnostic model, and defeated
Be out of order type diagnostic result;
It should be noted that secondary failure diagnostic model can have been realized when secondary failure diagnostic knowledge base has been laid in sufficiently
Standby training can start extreme learning machine at this time and carry out second level fault diagnosis, and shield fault type diagnostic machine simultaneously.The limit
Habit machine proposes that compared with traditional BP neural network, it is asked without iterating for single hidden layer feedforward neural network
Solution, but weight and the biasing of hidden layer node are randomly selected, model training is rapidly completed by seeking generalized inverse matrix, relatively passes
System BP neural network, the limit learn and have better generalization ability.Extreme learning machine is single hidden layer neural network, comprising defeated
Enter layer, hidden layer and output layer.Policy Limit learning machine network structure, ω are diagnosed with reference to the secondary failure of Fig. 3iAnd βiRespectively
For connection input layer and hidden layer, the weight matrix for connecting hidden layer and output layer;biFor the bias matrix of hidden layer, xiWith yi
The respectively input vector and output vector of extreme learning machine.For L given data sample (xi,ti) (i=1,2 ...,
L), wherein xi=[xi1,xi2..., xiL] ', yi=[yi1, yi2,…,yiL]'.With n hidden layer node (n < L) and activation letter
The extreme learning machine output that number is g (x) may be expressed as:
If activation primitive g (x) infinitely can be micro-, extreme learning machine can be with zero error this true output of approximating spline ti,
I.e.
Formula (6) can be denoted as:
H β=T (7)
Wherein, H is hidden layer output matrix are as follows:
Output weight vector β can be obtained by seeking the generalized inverse H+ of matrix H, and generally use Moerr-Penrose
Generalized inverse provides output weight vector β, sees shown in formula (9), and wherein c is constant.
Referring to Fig. 5, the secondary failure diagnostic knowledge base in Fig. 5 is responsible for collecting, lays in fault vectors Fi=[fi1,
fi2..., fiq]TAnd its corresponding fault type k through fault recognitioni, then by the fault vectors of chopped-off head diagnosis accumulation in step 2
And its fault type (Fi,ki) (i=1,2 ... Nc(t)) carry out composing training sample, Nc (t) is indicated by current time t limit inferior
The effective number of training stored in learning machine fault diagnosis knowledge base.In the hidden layer output matrix H of current t moment model
(t) are as follows:
Weight vector β is exported by the network hidden layer of the available current t moment of extreme learning machine network architecture
(t):
After model training complete enabling, the fault vectors of indistinct logic computer outputAs mode input, then pole
Limit learning machine can directly export fault type kNc(t+1), realize that secondary failure diagnoses decision.
The present embodiment by the data acquisition based on cloud service and mMTC and uRLLC technology, analysis, processing and management method with
The multiple knowledge base two-stage Fault Tree Diagnosis Decision model of fuzzy expert system and extreme learning machine combines, in complicated automation meter
In amount detection systems environment, it is quasi- that this method not only overcomes equipment fault diagnosis of the conventional automated calibrating center based on cloud service
The low technical problem of true rate.
The above are a kind of the detailed of second embodiment of the equipment fault diagnosis method based on cloud service provided by the present application
Describe in detail bright, here is a kind of the detailed of the third embodiment of equipment fault diagnosis method based on cloud service provided by the present application
Explanation.
Referring to Fig. 3, the equipment fault diagnosis method provided by the embodiments of the present application based on cloud service is in above-mentioned implementation
The ELM secondary failure diagnostic model that example refers to not yet has been realized because of equipment update or hardware platform upgrading or extreme learning machine
The method for diagnosing faults implemented when standby training, comprising:
301, it is based on cloud memory technology, acquisition in real time includes that the failure of the equipment operating data of automatic verification system is examined
Disconnected symptom vector;
302, the fault vectors conversion in fuzzy reasoning operational model is passed through according to the fault diagnosis symptom vector got
Fault diagnosis symptom vector is converted into corresponding fault vectors by formula;
303, the corresponding fault type diagnosis of fault vectors is determined by fault threshold decision procedure according to fault vectors
As a result;
It should be noted that the chopped-off head fault diagnosis that the present embodiment carries out is in equipment update or hardware platform upgrading
Or extreme learning machine carries out diagnosis decision when not yet realizing complete trained, mainly determines failure by fault type threshold determination
Type.In particular, the fixed threshold of each fault type is set according to expertise first, as shown in formula (12):
In formula, λiIndicate fault type fiThreshold value, rijFor the corresponding element in fuzzy reasoning matrix, p is Fuzzy Representation
Symptom vector number, q is the obtained fault vectors number of fuzzy reasoning.
304, using fault type diagnostic result and fault diagnosis symptom vector as model training input variable, for training
ELM secondary failure diagnostic model.
In addition, the fault vectors F=[f provided for indistinct logic computer1,f2,…,fq] ', compared one by one with threshold value λ
Compared with finding out the fault type { f beyond threshold valuem,…,fn}.Dynamic threshold is calculated subsequently, for the equipment beyond fixed threshold/element
Value τ, as shown in formula (13):
In formula, τsFor equipment/element s dynamic threshold, tsFor equipment/element s accumulation non-failure operation time, TsFor
Equipment/element s mean time between failures.If equipment/element beyond fixed threshold further exceeds its dynamic threshold,
It then can determine that its fault type.
The above are a kind of the detailed of the third embodiment of the equipment fault diagnosis method based on cloud service provided by the present application
Describe in detail it is bright, here be a kind of one embodiment of equipment fault diagnosis device based on cloud service provided by the present application specifically
It is bright.
Referring to Fig. 4, the embodiment of the present application provides a kind of equipment fault diagnosis device based on cloud service, comprising:
Symptom vector acquiring unit 401, for being based on cloud memory technology, obtaining in real time includes automatic verification system
The fault diagnosis symptom vector of equipment operating data, and store and store in equipment beyond the clouds;
Fault vectors converting unit 402, for passing through fuzzy reasoning operation according to the fault diagnosis symptom vector got
Obtain fault vectors;
Failure diagnosis unit 403, for by fault vectors be input to training after the completion of ELM secondary failure diagnostic model into
The diagnosis of row secondary failure, obtains fault type diagnostic result.
Preferably, fault vectors converting unit 402 is specifically used for:
Pass through the fault vectors conversion formula in fuzzy reasoning operational model according to the fault diagnosis symptom vector got,
Fault diagnosis symptom vector is converted into corresponding fault vectors, wherein fault vectors conversion formula are as follows:
In formula, L is fault diagnosis symptom vector set, and F is fault vectors set, and R (E) is the equipment in expert knowledge library
The fuzzy reasoning matrix of environmental variance E, rij(E) the symptom vector l when environmental variance is E is indicatedjLead to failure type fi
Probability.
Preferably, failure diagnosis unit 403 specifically includes:
Input quantity obtains subelement 4031, diagnoses for fault vectors to be input to the ELM secondary failure after the completion of training
Model;
Fault diagnosis subelement 4032, for by the fault type judgment formula in ELM secondary failure diagnostic model into
Row secondary failure diagnoses, and exports fault type diagnostic result, wherein fault type judgment formula are as follows:
In formula,For (t+1) secondary fault type judging result,For (t+1) secondary input failure to
Amount, β (t) are that the t times network exports weight vector, ωiFor the weight matrix for connecting input layer and hidden layer, biFor the inclined of hidden layer
Matrix is set, g is sigmoid function, and i is prototype network hidden layer number.
Preferably, it before the training of ELM secondary failure diagnostic model is completed, is pushed away according to fault diagnosis symptom vector by fuzzy
Reason operation obtains after fault vectors further include:
Tentative diagnosis unit 404, for determining fault vectors pair by fault threshold decision procedure according to fault vectors
The fault type diagnostic result answered;
Model training unit 405, for using fault type diagnostic result and fault diagnosis symptom vector as model training
Input variable, for training ELM secondary failure diagnostic model.
Preferably, tentative diagnosis unit specifically includes:
Fixed threshold determine subelement 4041, according to the one-to-one failure fixed threshold of various fault types, will therefore
Barrier vector is compared with failure fixed threshold, and filtering out first beyond failure fixed threshold transfinites fault vectors, and obtains
Fault type preliminary judgement result;
Dynamic threshold determines subelement 4042, is used for according to fault type preliminary judgement as a result, determining in conjunction with dynamic threshold
Formula carries out secondary judgement to first fault vectors that transfinite, and filtering out second beyond dynamic threshold transfinites fault vectors, obtains
Fault type finally determines result.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
In several embodiments provided herein, it should be understood that disclosed system, device and method can be with
It realizes by another way.For example, the apparatus embodiments described above are merely exemplary, for example, the unit
It divides, only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components
It can be combined or can be integrated into another system, or some features can be ignored or not executed.Another point, it is shown or
The mutual coupling, direct-coupling or communication connection discussed can be through some interfaces, the indirect coupling of device or unit
It closes or communicates to connect, can be electrical property, mechanical or other forms.
The description of the present application and term " first " in above-mentioned attached drawing, " second ", " third ", " the 4th " etc. are (if deposited
) it is to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should be understood that use in this way
Data are interchangeable under appropriate circumstances, so that embodiments herein described herein for example can be in addition to illustrating herein
Or the sequence other than those of description is implemented.In addition, term " includes " and " having " and their any deformation, it is intended that
Cover it is non-exclusive include, for example, containing the process, method, system, product or equipment of a series of steps or units need not limit
In step or unit those of is clearly listed, but may include be not clearly listed or for these process, methods, produce
The other step or units of product or equipment inherently.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
The above, above embodiments are only to illustrate the technical solution of the application, rather than its limitations;Although referring to before
Embodiment is stated the application is described in detail, those skilled in the art should understand that: it still can be to preceding
Technical solution documented by each embodiment is stated to modify or equivalent replacement of some of the technical features;And these
It modifies or replaces, the spirit and scope of each embodiment technical solution of the application that it does not separate the essence of the corresponding technical solution.
Claims (10)
1. a kind of equipment fault diagnosis method based on cloud service characterized by comprising
S1: being based on cloud memory technology, and acquisition in real time includes the fault diagnosis disease of the equipment operating data of automatic verification system
Shape vector;
S2: fault vectors are obtained by fuzzy reasoning operation according to the fault diagnosis symptom vector got;
S3: the fault vectors are input to the ELM secondary failure diagnostic model after the completion of training and carry out secondary failure diagnosis, are obtained
To fault type diagnostic result.
2. the method according to claim 1, wherein the step S2 is specifically included:
S21: it is converted according to the fault diagnosis symptom vector got by the fault vectors in fuzzy reasoning operational model
The fault diagnosis symptom vector is converted into corresponding fault vectors, wherein the fault vectors conversion formula by formula are as follows:
In formula, L is fault diagnosis symptom vector set, and F is fault vectors set, and R (E) is the facility environment in expert knowledge library
The fuzzy reasoning matrix of variable E, rij(E) the symptom vector l when environmental variance is E is indicatedjLead to failure type fiIt is general
Rate.
3. the method according to claim 1, wherein the step S3 is specifically included:
S31: the fault vectors are input to the ELM secondary failure diagnostic model after the completion of training;
S32: secondary failure diagnosis is carried out by the fault type judgment formula in the ELM secondary failure diagnostic model, and defeated
Type diagnostic be out of order as a result, the wherein fault type judgment formula are as follows:
In formula,For (t+1) secondary fault type judging result,For the fault vectors of (t+1) secondary input, β
(t) weight vector, ω are exported for the t times networkiFor the weight matrix for connecting input layer and hidden layer, biFor the biasing square of hidden layer
Battle array, g are sigmoid function, and i is prototype network hidden layer number.
4. according to the method described in claim 2, it is characterized in that, the ELM secondary failure diagnostic model training complete before,
It is described fault vectors are obtained by fuzzy reasoning operation according to the fault diagnosis symptom vector after further include:
S4: according to the fault vectors, by fault threshold decision procedure, determine that the corresponding fault type of the fault vectors is examined
Disconnected result;
S5: it using the fault type diagnostic result and the fault diagnosis symptom vector as model training input variable, is used for
The training ELM secondary failure diagnostic model.
5. according to the method described in claim 4, it is characterized in that, the step S4 is specifically included:
S41: according to the one-to-one failure fixed threshold of various fault types, the fault vectors and the failure are fixed
Threshold value is compared, and filtering out first beyond the failure fixed threshold transfinites fault vectors, and it is preliminary to obtain fault type
Determine result;
S42: it according to the fault type preliminary judgement as a result, determining formula in conjunction with dynamic threshold, transfinites failure to described first
Vector carries out secondary judgement, and filtering out second beyond the dynamic threshold transfinites fault vectors, obtains fault type and finally sentences
Determine result.
6. a kind of equipment fault diagnosis device based on cloud service characterized by comprising
Symptom vector acquiring unit, for being based on cloud memory technology, acquisition in real time includes the equipment fortune of automatic verification system
The fault diagnosis symptom vector of row data;
Fault vectors converting unit, for being obtained according to the fault diagnosis symptom vector got by fuzzy reasoning operation
Fault vectors;
Failure diagnosis unit is carried out for the fault vectors to be input to the ELM secondary failure diagnostic model after the completion of training
Secondary failure diagnosis, obtains fault type diagnostic result.
7. device according to claim 6, which is characterized in that the fault vectors converting unit is specifically used for:
Pass through the fault vectors conversion formula in fuzzy reasoning operational model according to the fault diagnosis symptom vector got,
The fault diagnosis symptom vector is converted into corresponding fault vectors, wherein the fault vectors conversion formula are as follows:
In formula, L is fault diagnosis symptom vector set, and F is fault vectors set, and R (E) is the facility environment in expert knowledge library
The fuzzy reasoning matrix of variable E, rij(E) the symptom vector l when environmental variance is E is indicatedjLead to failure type fiIt is general
Rate.
8. device according to claim 6, which is characterized in that the failure diagnosis unit specifically includes:
Input quantity obtains subelement, the ELM secondary failure diagnostic model for being input to the fault vectors after the completion of training;
Fault diagnosis subelement, for carrying out two by the fault type judgment formula in the ELM secondary failure diagnostic model
Grade fault diagnosis, and fault type diagnostic result is exported, wherein the fault type judgment formula are as follows:
In formula,For (t+1) secondary fault type judging result,For the fault vectors of (t+1) secondary input, β
(t) weight vector, ω are exported for the t times networkiFor the weight matrix for connecting input layer and hidden layer, biFor the biasing square of hidden layer
Battle array, g are sigmoid function, and i is prototype network hidden layer number.
9. device according to claim 7, which is characterized in that before ELM secondary failure diagnostic model training is completed,
It is described fault vectors are obtained by fuzzy reasoning operation according to the fault diagnosis symptom vector after further include:
Tentative diagnosis unit, for determining the fault vectors pair by fault threshold decision procedure according to the fault vectors
The fault type diagnostic result answered;
Model training unit, for using the fault type diagnostic result and the fault diagnosis symptom vector as model training
Input variable, for training the ELM secondary failure diagnostic model.
10. device according to claim 9, which is characterized in that the tentative diagnosis unit specifically includes:
Fixed threshold determine subelement, according to the one-to-one failure fixed threshold of various fault types, by the failure to
Amount is compared with the failure fixed threshold, and filtering out first beyond the failure fixed threshold transfinites fault vectors, and
Obtain fault type preliminary judgement result;
Dynamic threshold determine subelement, for according to the fault type preliminary judgement as a result, in conjunction with dynamic threshold determine formula,
Secondary judgement is carried out to described first fault vectors that transfinite, filtering out second beyond the dynamic threshold transfinites fault vectors,
It obtains fault type and finally determines result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811626867.2A CN109711062A (en) | 2018-12-28 | 2018-12-28 | A kind of equipment fault diagnosis method and device based on cloud service |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811626867.2A CN109711062A (en) | 2018-12-28 | 2018-12-28 | A kind of equipment fault diagnosis method and device based on cloud service |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109711062A true CN109711062A (en) | 2019-05-03 |
Family
ID=66259233
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811626867.2A Pending CN109711062A (en) | 2018-12-28 | 2018-12-28 | A kind of equipment fault diagnosis method and device based on cloud service |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109711062A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110597863A (en) * | 2019-09-25 | 2019-12-20 | 上海依图网络科技有限公司 | Retrieval system and method for keeping stable performance in control library through dynamic threshold |
CN110955550A (en) * | 2019-11-24 | 2020-04-03 | 济南浪潮数据技术有限公司 | Cloud platform fault positioning method, device, equipment and storage medium |
CN111178378A (en) * | 2019-11-07 | 2020-05-19 | 腾讯科技(深圳)有限公司 | Equipment fault prediction method and device, electronic equipment and storage medium |
CN113485861A (en) * | 2021-07-08 | 2021-10-08 | 无锡江南计算技术研究所 | Basic layer fault diagnosis method and device based on resource optimization strategy |
CN114383648A (en) * | 2021-12-10 | 2022-04-22 | 国家能源集团岳阳发电有限公司 | Temperature instrument fault diagnosis method and device |
CN116723083A (en) * | 2023-08-10 | 2023-09-08 | 山东智拓大数据有限公司 | Cloud server online fault diagnosis method and device |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6636841B1 (en) * | 1997-04-01 | 2003-10-21 | Cybula Ltd. | System and method for telecommunications system fault diagnostics |
CN104133143A (en) * | 2014-07-31 | 2014-11-05 | 东北大学 | Power grid line fault diagnosis system and method based on Hadoop cloud computing platform |
CN105911975A (en) * | 2016-04-14 | 2016-08-31 | 南京航空航天大学 | Integrated diagnosis method of gas path fault of aeroengine |
CN107192951A (en) * | 2017-07-31 | 2017-09-22 | 湖南科技大学 | A kind of micro- method for diagnosing faults of wind-driven generator three-phase rotor current |
CN108062572A (en) * | 2017-12-28 | 2018-05-22 | 华中科技大学 | A kind of Fault Diagnosis Method of Hydro-generating Unit and system based on DdAE deep learning models |
-
2018
- 2018-12-28 CN CN201811626867.2A patent/CN109711062A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6636841B1 (en) * | 1997-04-01 | 2003-10-21 | Cybula Ltd. | System and method for telecommunications system fault diagnostics |
CN104133143A (en) * | 2014-07-31 | 2014-11-05 | 东北大学 | Power grid line fault diagnosis system and method based on Hadoop cloud computing platform |
CN105911975A (en) * | 2016-04-14 | 2016-08-31 | 南京航空航天大学 | Integrated diagnosis method of gas path fault of aeroengine |
CN107192951A (en) * | 2017-07-31 | 2017-09-22 | 湖南科技大学 | A kind of micro- method for diagnosing faults of wind-driven generator three-phase rotor current |
CN108062572A (en) * | 2017-12-28 | 2018-05-22 | 华中科技大学 | A kind of Fault Diagnosis Method of Hydro-generating Unit and system based on DdAE deep learning models |
Non-Patent Citations (4)
Title |
---|
杨东升等: "电能表自动检定流水线模糊故障诊断专家系统", 《电测与仪表》 * |
肖忠宝: "基于多尺度核独立元分析与核极限学习机的柴油机故障诊断", 《车用发动机》 * |
赵肖敏: "基于模糊神经网络的带式输送机故障诊断研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
魏嘉: "大数据下风电机组齿轮箱故障诊断方法研究", 《万方数据》 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110597863A (en) * | 2019-09-25 | 2019-12-20 | 上海依图网络科技有限公司 | Retrieval system and method for keeping stable performance in control library through dynamic threshold |
CN110597863B (en) * | 2019-09-25 | 2023-01-24 | 上海依图网络科技有限公司 | Retrieval system and method for keeping stable performance in control library through dynamic threshold |
CN111178378A (en) * | 2019-11-07 | 2020-05-19 | 腾讯科技(深圳)有限公司 | Equipment fault prediction method and device, electronic equipment and storage medium |
CN111178378B (en) * | 2019-11-07 | 2023-05-16 | 腾讯科技(深圳)有限公司 | Equipment fault prediction method and device, electronic equipment and storage medium |
CN110955550A (en) * | 2019-11-24 | 2020-04-03 | 济南浪潮数据技术有限公司 | Cloud platform fault positioning method, device, equipment and storage medium |
CN110955550B (en) * | 2019-11-24 | 2022-07-08 | 济南浪潮数据技术有限公司 | Cloud platform fault positioning method, device, equipment and storage medium |
CN113485861A (en) * | 2021-07-08 | 2021-10-08 | 无锡江南计算技术研究所 | Basic layer fault diagnosis method and device based on resource optimization strategy |
CN114383648A (en) * | 2021-12-10 | 2022-04-22 | 国家能源集团岳阳发电有限公司 | Temperature instrument fault diagnosis method and device |
CN116723083A (en) * | 2023-08-10 | 2023-09-08 | 山东智拓大数据有限公司 | Cloud server online fault diagnosis method and device |
CN116723083B (en) * | 2023-08-10 | 2024-01-26 | 山东智拓大数据有限公司 | Cloud server online fault diagnosis method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109711062A (en) | A kind of equipment fault diagnosis method and device based on cloud service | |
CN109086889B (en) | Terminal fault diagnosis method, device and system based on neural network | |
WO2022037068A1 (en) | Method for diagnosis of fault in machine tool bearing | |
CN106769052B (en) | A kind of mechanical system rolling bearing intelligent failure diagnosis method based on clustering | |
CN108418841B (en) | Next-generation key message infrastructure network Security Situation Awareness Systems based on AI | |
CN107578104B (en) | A kind of Chinese Traditional Medicine knowledge system | |
CN109141847B (en) | Aircraft system fault diagnosis method based on MSCNN deep learning | |
CN111596604A (en) | Intelligent fault diagnosis and self-healing control system and method for engineering equipment based on digital twinning | |
CN109655259A (en) | Combined failure diagnostic method and device based on depth decoupling convolutional neural networks | |
CN111967486A (en) | Complex equipment fault diagnosis method based on multi-sensor fusion | |
CN110334764A (en) | Rotating machinery intelligent failure diagnosis method based on integrated depth self-encoding encoder | |
CN112765890B (en) | Dynamic domain adaptive network-based multi-working-condition rotating machine residual life prediction method | |
CN110766143A (en) | Equipment fault intelligent diagnosis method based on artificial neural network | |
CN108228716A (en) | SMOTE_Bagging integrated sewage handling failure diagnostic methods based on weighting extreme learning machine | |
JP2022524244A (en) | Predictive classification of future behavior | |
CN105487009A (en) | Motor fault diagnosis method based on k-means RBF neural network algorithm | |
CN106628097A (en) | Ship equipment fault diagnosis method based on improved radial basis function neutral network | |
CN112308147A (en) | Rotating machinery fault diagnosis method based on integrated migration of multi-source domain anchor adapter | |
CN114386312A (en) | Equipment fault diagnosis method | |
CN109670584A (en) | A kind of fault diagnosis method and system based on big data | |
CN113159113B (en) | Smart power grid fault diagnosis method capable of repairing remote measurement under information malicious tampering | |
CN113902710A (en) | Method and system for detecting surface defects of industrial parts based on anomaly detection algorithm | |
CN108827671A (en) | A kind of Trouble Diagnostic Method of Machinery Equipment | |
CN107016440A (en) | The multiresolution deep neural network intelligent diagnosing method of machine driving failure | |
Chopra et al. | Classification of faults in damadics benchmark process control system using self organizing maps |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190503 |