CN107276816B - A kind of long-range monitoring and fault diagnosis system and method for diagnosing faults based on cloud service - Google Patents
A kind of long-range monitoring and fault diagnosis system and method for diagnosing faults based on cloud service Download PDFInfo
- Publication number
- CN107276816B CN107276816B CN201710533956.1A CN201710533956A CN107276816B CN 107276816 B CN107276816 B CN 107276816B CN 201710533956 A CN201710533956 A CN 201710533956A CN 107276816 B CN107276816 B CN 107276816B
- Authority
- CN
- China
- Prior art keywords
- data
- service
- unit
- module
- cloud
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/02—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP]
- H04L67/025—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP] for remote control or remote monitoring of applications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Environmental & Geological Engineering (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses a kind of long-range monitoring and fault diagnosis system and method for diagnosing faults based on cloud service, are related to intelligence manufacture and cloud diagnostic field, including data acquisition unit, telecommunication gateway unit, cloud storage administrative unit, cloud service center unit.Data acquisition unit is deployed on data collection station;Telecommunication gateway unit, deployment is beyond the clouds on preposition gateway server;Cloud storage administrative unit, deployment is beyond the clouds on data server;Cloud service center unit, is deployed in cloud application server.BP neural network method for diagnosing faults based on genetic algorithm optimization is applied the long-range monitoring and fault diagnosis service that can be easily provided under suitable cross-region environment in long-range monitoring and fault diagnosis system by the invention;Mechanical equipment remote online monitoring of working condition can be provided for the technical staff of device manufacturer, provide quick, accurate, the efficient diagnosis of complex fault for the mechanical equipment that manufacturing enterprise uses.
Description
Technical field
The present invention relates to intelligence manufactures and cloud diagnostic field, particularly relate to a kind of long-range monitoring and failure based on cloud service
Diagnostic system and method for diagnosing faults.
Background technique
With computer, data acquisition, sensor and network technology fast development and extensive use, manufacturing technology also to
Networking and intelligent direction development.Mechanical equipment constantly uses modern advanced industrial technology, so that its complexity and dimension
Shield difficulty continues to increase, while also higher and higher to requirements such as the real-times, reliability, integrality of data collection and analysis, passes
The equipment monitoring based on single machine and live mode and fault diagnosis system of system have been unable to meet the maintenance requirement of modernization.How
Going on business for technical staff, particularly high-level R&D personnel is reduced, the Site Service workload of factory product is reduced;How
How the remote maintenance response speed for improving product finds potential problems in time, and then provides on the other hand for the transformation of product, design
Data foundation, be the realistic problem that device manufacturer faces.
Summary of the invention
The purpose of the present invention is to provide a kind of maintenance that product can be improved, maintenance response speed, timely discovering device is latent
In the long-range monitoring and fault diagnosis system and method for diagnosing faults based on cloud service of problem.
In order to achieve the above objectives, solution of the invention is:
A kind of long-range monitoring and fault diagnosis system based on cloud service, comprising: data acquisition unit, telecommunication gateway
Unit, cloud storage administrative unit and cloud service center unit;The data acquisition unit is deployed in the scene of mechanical equipment, data
Telecommunication gateway unit and the cloud storage management between acquisition unit and the telecommunication gateway unit, described
Cloud storage management between telecommunication gateway unit and the cloud service center unit between unit, described, described is single
It is first to be all made of network between the cloud service center unit and be attached.
Between the data acquisition unit and the telecommunication gateway unit, it is attached using 4G network;
Telecommunication gateway list between the telecommunication gateway unit and the cloud storage administrative unit, described
First cloud storage administrative unit between the cloud service center unit, described and between the cloud service center unit,
It is attached using Internet/Intranet network;
The cloud service center unit provides a user the interface WEB/APP/WAP using Internet or 4G network
Cloud service.
The data acquisition unit, including sensor module, controller module, remote communication module, are deployed in data
On acquisition terminal;
The sensor module acquires industrial site various kinds of sensors institute according to preset frequency acquisition in real time
The working condition signal data of perception, send remote communication module to;
The controller module, the control data that real-time reception remote communication module is transmitted, according to control data pair
The frequency acquisition of sensor module, remote communication module report the parameters such as frequency to be configured and modify;
The remote communication module, on the one hand, the downlink command that real-time reception telecommunication gateway unit issues, it then follows
Telecommunication agreement analyzes the instruction data, and by control data obtained after parsing, sends controller module to, on the other hand,
The working condition signal data that real-time reception sensor module is transmitted, it then follows telecommunication agreement encapsulates working condition signal data, and will
The director data obtained after encapsulation reports frequency according to preset, sends up-on command, real-time report gives telecommunication gateway list
Member.
The telecommunication gateway unit includes Communications service module, data prediction service module and intelligent adaptation clothes
Business module, is disposed respectively on preposition gateway server beyond the clouds;
The Communications service module, on the one hand, the up-on command that real-time reception data acquisition unit reports, it then follows long-range
Communications protocol analyzes the instruction data, and by working condition signal data obtained after parsing, sends data prediction service module to,
On the other hand, control data transmitted by real-time reception cloud service center unit, it then follows telecommunication agreement, encapsulation control number
According to, and the director data that will be obtained after encapsulation, downlink command is sent, real time down is to data acquisition unit;
The data prediction service module, the working condition signal data that real-time reception Communications service module is transmitted, into
Line number Data preprocess, and it is sent to cloud storage administrative unit in real time;
The intelligent adaptation service module, provides mode adapter, for corresponding model mechanical equipment specify it is corresponding
Telecommunication agreement.
The cloud storage administrative unit, including data storage service module, data retrieval service module and data encryption
Service module, each module deployment is beyond the clouds on data server;
The data storage service module, characteristic vector data transmitted by real-time reception telecommunication gateway unit,
It stores in corresponding database;
The data retrieval service module, inquiry request transmitted by real-time reception cloud service center unit, using number
According to retrieval technique, target data is obtained from database, and target data is returned into cloud service center unit;
The data encryption services module, using the mature close encryption technology of quotient, for certain sensitive or crucial features
The storage of vector data provides reliable encryption, decryption service.
The cloud service center unit, including monitoring service module, diagnostic service module, Warning Service module, meeting
Service module, library service module, training service module, are deployed in cloud application server;
The monitoring service module, can operating status to mechanical equipment, live scene etc., provide figure, chart,
Process in the real-time visual monitoring service and designated time period of the ways of presentation such as image, video recalls service;
The diagnostic service module is different difficulty, different levels by built-in knowledge base, model library, state repository etc.
Diagnostic requirements provide expert diagnosis service;
The expert diagnosis service is calculated using the BP neural network method for diagnosing faults based on genetic algorithm optimization
Steps are as follows for method:
Step 1: determine input, output vector:
It according to the aufbauprinciple of Boolean matrix, is located in fault diagnosis, characteristic parameter has m, i.e. input feature value P=
(s1,s2,…,sm), fault type to be identified has n, i.e. output feature vector Q=(r1,r2,…,rn);According to fuzzy clustering
Analysis, rj(j ∈ [1, n]) value between (0,1) determines rjMiddle degree of membership the maximum is the reason of component failure occurs;
Step 2: choose the network number of plies:
Using three layers of BP neural network, respectively input layer, hidden layer, output layer.It is the input according to step 1, defeated
Outgoing vector determines that input layer number is a, wherein a=m, and output layer neuron number is b, wherein b=n;
Step 3: calculate hidden layer neuron number:
Hidden layer neuron number is by formulaDetermine, x is a constant, value [1,10] it
Between;
Step 4: setting initial weight:
Set random number of the initial weight as between [- 1,1];
Step 5: setting learning rate:
Set random number of the learning rate as between [0.01,0.8];
Step 6: being optimized, avoided subsequent using initial weight and learning rate of the genetic algorithm to BP neural network
E-learning falls into local minimum, comprising the following steps:
Step 6.1: genetic algorithm is determined according to the input layer of BP neural network, hidden layer and output layer neuron number
Code length L
L=a*b+b*h+h*a;
Step 6.2: determining the fitness function of genetic algorithm;
Step 6.3: new population at individual is generated by the selection of genetic algorithm, intersection and mutation operation;
Step 6.4: according to code length and fitness function, the fitness value of population at individual is calculated, if the fitness value
Meet adaptive optimal control degree, then step 6.3 is obtained into population at individual as optimal individual and be output to BP neural network as initial
Weight and learning rate, enter step 6.5, otherwise continue the operation of step 6.3;
Step 6.5: judging whether genetic algorithm has reached the maximum evolutionary generation of setting, optimal solution work is exported if reaching
For the initial weight and learning rate of BP neural network, 7 are entered step, step 6.3 is otherwise gone to;
Step 7: feature vector grouping:
It is two groups that input vector P, which is divided to, and one group, as learning sample data progress e-learning, is used XPIt indicates, another group
As diagnostic analysis data, Y is usedPIt indicates;
Step 8: e-learning, comprising the following steps:
Step 8.1: initial weight, learning rate and the learning sample data X that step 6.4 and step 6.5 are obtainedPInput
The input layer of BP neural network calculates the output of hidden layer, each neuron of output layer;
Step 8.2: calculating the deviation E of output layer desired output and real output valueP;
Step 8.3: if EPMeet training error condition, then e-learning terminates, and 9 is entered step, conversely, then adjusting defeated
The weight of layer and hidden layer out, return step 8.1 continue to learn, and so on, until deviation EpIt is eligible;
Step 8.4: the correspondence weight that the final weight that e-learning obtains is analyzed as follow-up diagnosis, and diagnosed
The algorithm model of analysis;
Step 9: diagnostic analysis:
By diagnostic analysis data YPThe algorithm model that input step 8.3 obtains calculates real output value (degree of membership), if being subordinate to
It is to think that equipment working condition exists for failure that category degree, which is more than 0.8, on the contrary, then it is assumed that equipment working condition is normal.
The Warning Service module is provided by technologies such as early stage small fault detection, time prediction, qualitative analyses
Fault pre-alarming service to mechanical equipment, and warning information can be passed through sound alarm, mail notification, short massage notice, automatic language
The forms such as sound phone, inform related personnel in time;
It is white to provide video conference, voice conferencing, electronics for device manufacturer and manufacturing enterprise for the conference service module
Plate, file-sharing, Desktop Share, collaborative browse, electronic voting etc. are remotely linked up and collaboration services;
The library service module, the users at different levels for device manufacturer and manufacturing enterprise provide conveniently information
Or the service such as document uploads, is shared, inquiry;
The training service module, facilitates on-line teaching, and device manufacturer can provide remote training for manufacturing enterprise
With interactive teaching service.
As prioritization scheme of the invention, the telecommunication gateway unit, using server cluster and load balancing
Technology deployment, to support high concurrent to access;
As prioritization scheme of the invention, the cloud storage administrative unit, using distributed data base, server cluster
It is disposed with load-balancing technique, to support the management of magnanimity characteristic vector data and high concurrent to access;
As prioritization scheme of the invention, the cloud service center unit, using server cluster and load balancing skill
Art deployment, to support high concurrent to access.
It advantages of the present invention and has the active effect that
(1) the present invention is based on the cloud service building mode of loose coupling so that the reusable Du Genggao of functional module, be easy to be
The extension of system is realized and is effectively integrated to the resource of dispersion, isomery, can easily be provided under suitable cross-region environment
Long-range monitoring and fault diagnosis service;
(2) present invention creatively applies the BP neural network method for diagnosing faults based on genetic algorithm optimization in machinery
Equipment fault diagnosis effectively improves the diagnostic accuracy of diagnostic system.
(3) present invention can provide mechanical equipment remote online monitoring of working condition for the technical staff of device manufacturer, improve
The remote fault diagnosis of mechanical equipment is horizontal, reduces going on business for technical staff, reduces Site Service workload;It can be production enterprise
The mechanical equipment that industry uses provides quick, accurate, the efficient diagnosis of complex fault, shortens the response time of maintenance request, improves
Maintenance efficiency is reduced because losing caused by maintenance shut-downs;Device manufacturer passes through the machinery that telecommunication network is that manufacturing enterprise uses
Equipment completes periodic maintenance, data monitoring, system upgrade, failure consulting, coordinates diagnosis and maintenance service, reduces product maintenance
Expense reaches the target of green manufacturing, production and maintenance;
(4) present invention can real-time monitoring equipment state and parameter, find the problem immediately, realize the transparence of equipment operation
Management, possesses complete device history data, analysis failure can be traced, by monitoring the working order of production equipment in real time, no
Disconnected optimizing process improves production efficiency to improve product quality;By remotely monitoring and the maintenance function of fault diagnosis
The behaviour in service of equipment and development trend are fed back into design and manufacture department, can constantly improve and improve equipment, facilitate reality
Show equipment from design, manufacture, installation, operation, superseded lifecycle management.
Detailed description of the invention
Fig. 1 is the structural diagram of the present invention;
Fig. 2 is data acquisition unit structure chart of the invention;
Fig. 3 is intelligent communication gateway unit structure chart of the invention;
Fig. 4 is cloud storage administrative unit structure chart of the invention;
Fig. 5 is cloud service center cellular construction figure of the invention;
Fig. 6 is remote diagnosis data flowchart of the invention;
Fig. 7 is diagnosis algorithm flow chart of the invention.
Below in conjunction with the drawings and specific embodiments, the invention will be further described.
Specific embodiment
The present embodiment is using the long-range monitoring and fault diagnosis system of RT-RMDS as prototype, detailed description of the present invention embodiment party
Formula.The long-range monitoring and fault diagnosis system of RT-RMDS is based on cloud service framework, for large size, complexity, the high skill in manufacturing industry
The characteristic distributions of art equipment will be distributed independent equipment connection by data acquisition, network communication and fault diagnosis for mutually association
The organism of work timely responds to status monitoring and failure disposition with realizing, and has resource-sharing, remote collaboration, data
The functions such as exchange.
As shown in Figure 1, a kind of long-range monitoring and fault diagnosis system based on cloud service, including data acquisition unit 100,
Telecommunication gateway unit 200, cloud storage administrative unit 300, cloud service center unit 400;
Between the data acquisition unit 100 and the telecommunication gateway unit 200, connected using 4G network
It connects;
Between the telecommunication gateway unit 200 and the cloud storage administrative unit 300, the telecommunication
Cloud storage administrative unit 300 and the cloud between gateway unit 200 and the cloud service center unit 400, described take
It is engaged between center cell 400, is attached using Internet/Intranet network;
The cloud service center unit 400 provides a user WEB/APP/WAP circle using Internet/4G network
The cloud service in face;
As shown in Fig. 2, the data acquisition unit 100, including it is sensor module 101, controller module 102, long-range
Communication module 103, is deployed on data collection station;
The sensor module 101 acquires industrial site various kinds of sensors according to preset frequency acquisition in real time
The working condition signal data perceived, send remote communication module 103 to;
The controller module 102, the control data that real-time reception remote communication module 103 is transmitted, according to control
The frequency acquisition of data mutual transmission sensor module 101, remote communication module 103 report the parameters such as frequency to be configured and modify;
Cooperate shown in Fig. 6, the remote communication module 103, on the one hand, real-time reception telecommunication gateway unit 200
The downlink command issued, it then follows telecommunication agreement analyzes the instruction data, and by control data obtained after parsing, sends to
Controller module 102, on the other hand, the working condition signal data that real-time reception sensor module 101 is transmitted, it then follows telecommunication
Agreement encapsulates working condition signal data, and the director data that will be obtained after encapsulation, reports frequency according to preset, sends uplink and refer to
It enables, real-time report is to telecommunication gateway unit 200;
As shown in figure 3, the telecommunication gateway unit 200, including Communications service module 201, data prediction clothes
Business module 202, intelligent adaptation service module 203, deployment is beyond the clouds on preposition gateway server;
Cooperate shown in Fig. 6, the Communications service module 201, on the one hand, real-time reception data acquisition unit 100 reports
Up-on command, it then follows telecommunication agreement analyzes the instruction data, and by working condition signal data obtained after parsing, sends to
Data prediction service module 202, on the other hand, control data transmitted by real-time reception cloud service center unit 400, it then follows
Telecommunication agreement, encapsulation control data, and the director data that will be obtained after encapsulation, send downlink command, real time down is to number
According to acquisition unit 100;
The data prediction service module 202, the working condition signal number that real-time reception Communications service module 201 is transmitted
According to, progress data prediction, and it is sent to cloud storage administrative unit 300 in real time;
The intelligent adaptation service module 203, provides mode adapter, is the mechanical equipment of different model, specifies not
Same telecommunication agreement;
As shown in figure 4, the cloud storage administrative unit 300, including data storage service module 301, data retrieval clothes
Business module 302, data encryption services module 303, deployment is beyond the clouds on data server;
The data storage service module 301, feature transmitted by real-time reception telecommunication gateway unit 200 to
Data are measured, are stored into corresponding database;
The data retrieval service module 302, inquiry request transmitted by real-time reception cloud service center unit 400,
Using data retrieval technology, target data is obtained from database, and target data is returned into cloud service center unit 400;
The data encryption services module 303, using the mature close encryption technology of quotient, for certain sensitive or crucial spies
The storage for levying vector data provides reliable encryption, decryption service;
As shown in figure 5, the cloud service center unit 400, including monitoring service module 401, diagnostic service module
402, Warning Service module 403, conference service module 404, library service module 405, training service module 406, are deployed in cloud
On end application server;
The monitoring service module 401, can operating status to mechanical equipment, live scene etc., figure, figure are provided
Process in the real-time visual monitoring service and designated time period of the ways of presentation such as table, image, video recalls service;
The diagnostic service module 402 is different difficulty, difference by built-in knowledge base, model library, state repository etc.
The diagnostic requirements of level provide expert diagnosis service;
Cooperate the expert diagnosis service, using the BP neural network failure based on genetic algorithm optimization shown in Fig. 7
Diagnostic method, algorithm steps are as follows:
Step 1: determine input, output vector:
According to the aufbauprinciple of Boolean matrix, be located in fault diagnosis, characteristic parameter has a m, i.e., input vector (feature to
Amount) P=(s1,s2,…,sm), fault type to be identified has n, i.e. output vector Q=(r1,r2,…,rn);According to fuzzy poly-
Alanysis, rj(j ∈ [1, n]) value between (0,1) determines rjMiddle degree of membership the maximum is the original that component failure occurs
Cause;
Step 2: choose the network number of plies:
Using three layers of BP neural network, respectively input layer, hidden layer, output layer.It is the input according to step 1, defeated
Outgoing vector determines that input layer number is a, wherein a=m, and output layer neuron number is b, wherein b=n;
Step 3: calculate hidden layer neuron number:
Hidden layer neuron number is by formulaDetermine, x is a constant, value [1,10] it
Between;
Step 4: setting initial weight:
Set random number of the initial weight as between [- 1,1];
Step 5: setting learning rate:
Set random number of the learning rate as between [0.01,0.8];
Step 6: being optimized, avoided subsequent using initial weight and learning rate of the genetic algorithm to BP neural network
E-learning falls into local minimum, comprising the following steps:
Step 6.1: genetic algorithm is determined according to the input layer of BP neural network, hidden layer and output layer neuron number
Code length L
L=a*b+b*h+h*a;
Step 6.2: determining the fitness function of genetic algorithm;
Step 6.3: new population at individual is generated by the selection of genetic algorithm, intersection and mutation operation;
Step 6.4: according to code length and fitness function, the fitness value of population at individual is calculated, if the fitness value
Meet adaptive optimal control degree, then step 6.3 is obtained into population at individual as optimal individual and be output to BP neural network as initial
Weight and learning rate, enter step 6.5, otherwise continue the operation of step 6.3;
Step 6.5: judging whether genetic algorithm has reached the maximum evolutionary generation of setting, optimal solution work is exported if reaching
For the initial weight and learning rate of BP neural network, 7 are entered step, step 6.3 is otherwise gone to;
Step 7: feature vector grouping:
It is two groups that input vector P, which is divided to, and one group, as learning sample data progress e-learning, is used XPIt indicates, another group
As diagnostic analysis data, Y is usedPIt indicates;
Step 8: e-learning, comprising the following steps:
Step 8.1: initial weight, learning rate and the learning sample data X that step 6.4 and step 6.5 are obtainedPInput
The input layer of BP neural network calculates the output of hidden layer, each neuron of output layer;
Step 8.2: calculating the deviation E of output layer desired output and real output valueP;
Step 8.3: if EPMeet training error condition, then e-learning terminates, and 9 is entered step, conversely, then adjusting defeated
The weight of layer and hidden layer out, return step 8.1 continue to learn, and so on, until deviation EpIt is eligible;
Step 8.4: the correspondence weight that the final weight that e-learning obtains is analyzed as follow-up diagnosis, and diagnosed
The algorithm model of analysis;
Step 9: diagnostic analysis:
By diagnostic analysis data YPThe algorithm model that input step 8.3 obtains calculates real output value, that is, degree of membership, if being subordinate to
It is to think that equipment working condition exists for failure that category degree, which is more than 0.8, on the contrary, then it is assumed that equipment working condition is normal.
The Warning Service module 403 is mentioned by technologies such as early stage small fault detection, time prediction, qualitative analyses
For the fault pre-alarming service to mechanical equipment, and warning information can be passed through sound alarm, mail notification, short massage notice, automatic
The forms such as voice call, inform related personnel in time;
The conference service module 404 provides video conference, voice conferencing, electricity for device manufacturer and manufacturing enterprise
Sub- blank, file-sharing, Desktop Share, collaborative browse, electronic voting etc. are remotely linked up and collaboration services;
The library service module 405, the users at different levels for device manufacturer and manufacturing enterprise provide conveniently
Information or the service such as document upload, shared, inquiry;
The training service module 406, facilitates on-line teaching, and device manufacturer can provide long-range training for manufacturing enterprise
Instruction and interactive teaching service.
As prioritization scheme of the invention, the telecommunication gateway unit 200 is equal using server cluster and load
Weighing apparatus technology deployment, to support high concurrent to access;
As prioritization scheme of the invention, the cloud storage administrative unit 300, using distributed data base, server
Cluster and load-balancing technique deployment, to support the management of magnanimity characteristic vector data and high concurrent to access;
As prioritization scheme of the invention, the cloud service center unit 400, using server cluster and load balancing
Technology deployment, to support high concurrent to access.
The above is only the preferable specific embodiments of the present invention, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art the variation that can readily occur in or replaces in the technical scope that the embodiment of the present invention discloses
It changes, should all be included within the scope of the present invention.
Claims (6)
1. a kind of method for diagnosing faults of the long-range monitoring and fault diagnosis system based on cloud service, it is characterised in that: this is based on
The long-range monitoring and fault diagnosis system of cloud service includes data acquisition unit, telecommunication gateway unit, cloud storage management list
Member, cloud service center unit;The data acquisition unit is deployed in the scene of mechanical equipment, data acquisition unit with it is described remote
Between telecommunication gateway unit and the cloud storage administrative unit between journey Communication Gateway unit, described, it is described remote
In cloud storage administrative unit and the cloud service between journey Communication Gateway unit and the cloud service center unit, described
Network is all made of between heart unit to be attached;The cloud service center unit, including monitoring service module, diagnostic service mould
Block and Warning Service module, monitoring service module, diagnostic service module, Warning Service module dispose application service beyond the clouds respectively
On device;The monitoring service module provides the mistake in real-time visual monitoring service and designated time period to mechanical equipment
Journey backtracking service;The diagnostic service module is different difficulty, the diagnostic requirements of different levels provide expert diagnosis service;Institute
The Warning Service module stated provides the fault pre-alarming service to mechanical equipment, and warning information is informed related personnel in time;Its
Diagnostic method, comprising the following steps:
Step 1: determine input, output vector:
It according to the aufbauprinciple of Boolean matrix, is defined in fault diagnosis, characteristic parameter has m, i.e. input feature value P=(s1,
s2,…,sm), fault type to be identified has n, i.e. output feature vector Q=(r1,r2,…,rn);According to fuzzy cluster analysis,
rj(j ∈ [1, n]) value between (0,1) determines rjMiddle degree of membership the maximum is the reason of component failure occurs;
Step 2: choose the network number of plies:
Using three layers of BP neural network, respectively input layer, hidden layer, output layer;The input feature vector according to step 1 to
Amount, output feature vector determine that input layer number is a, wherein a=m, and output layer neuron number is b, wherein b=n;
Step 3: calculate hidden layer neuron number:
Hidden layer neuron number is by formula+ x determines that x is a constant, and value is between [1,10];
Step 4: setting initial weight:
Set random number of the initial weight as between [- 1,1];
Step 5: setting learning rate:
Set random number of the learning rate as between [0.01,0.8];
Step 6: being optimized using initial weight and learning rate of the genetic algorithm to BP neural network, avoid subsequent net
Network study falls into local minimum, comprising the following steps:
Step 6.1: the volume of genetic algorithm is determined according to the input layer of BP neural network, hidden layer and output layer neuron number
Code length L
L=a*b+b*h+h*a;
Step 6.2: determining the fitness function of genetic algorithm;
Step 6.3: new population at individual is generated by the selection of genetic algorithm, intersection and mutation operation;
Step 6.4: according to code length and fitness function, the fitness value of population at individual is calculated, if the fitness value meets
Step 6.3 is then obtained population at individual as optimal individual and is output to BP neural network as initial weight by adaptive optimal control degree
And learning rate, 6.5 are entered step, the operation of step 6.3 is otherwise continued;
Step 6.5: judging whether genetic algorithm has reached the maximum evolutionary generation of setting, optimal solution is exported if reaching as BP
The initial weight and learning rate of neural network, enter step 7, otherwise go to step 6.3;
Step 7: feature vector grouping:
It is two groups that input feature value P, which is divided to, and one group, as learning sample data progress e-learning, is used XPIt indicates, another group of work
For diagnostic analysis data, Y is usedPIt indicates;
Step 8: e-learning, comprising the following steps:
Step 8.1: initial weight, learning rate and the learning sample data X that step 6.4 and step 6.5 are obtainedPInput BP mind
Input layer through network calculates the output of hidden layer, each neuron of output layer;
Step 8.2: calculating the deviation E of output layer desired output and real output valueP;
Step 8.3: if EPMeet training error condition, then e-learning terminates, enter step 9, conversely, then adjust output layer and
The weight of hidden layer, return step 8.1 continue to learn, and so on, until deviation EpIt is eligible;
Step 8.4: the correspondence weight that the final weight that e-learning obtains is analyzed as follow-up diagnosis, and obtain diagnostic analysis
Algorithm model;
Step 9: diagnostic analysis:
By diagnostic analysis data YPThe algorithm model that input step 8.3 obtains calculates real output value, that is, degree of membership, if degree of membership
Think that equipment working condition exists for failure more than 0.8, it is on the contrary, then it is assumed that equipment working condition is normal.
2. a kind of fault diagnosis side of long-range monitoring and fault diagnosis system based on cloud service according to claim 1
Method, it is characterised in that:
Between the data acquisition unit and the telecommunication gateway unit, it is attached using 4G network;
Between the telecommunication gateway unit and the cloud storage administrative unit, the telecommunication gateway unit with
Between cloud storage administrative unit and the cloud service center unit between the cloud service center unit, described, use
Internet/Intranet network is attached;
The cloud service center unit provides a user the cloud clothes at the interface WEB/APP/WAP using Internet or 4G network
Business.
3. a kind of fault diagnosis side of long-range monitoring and fault diagnosis system based on cloud service according to claim 1
Method, which is characterized in that the data acquisition unit includes: sensor module, controller module and remote communication module, data
Each module of acquisition unit is separately positioned on data collection station;
The working condition signal data transmission that the sensor module acquisition sensor is perceived is to remote communication module;
The controller module receives the control data that remote communication module is transmitted;
The remote communication module receives the downlink command of telecommunication gateway unit, sends control data and gives controller mould
Block;The working condition signal data that the remote communication module also receiving sensor module is transmitted, and up-on command is sent to remote
Journey Communication Gateway unit.
4. a kind of fault diagnosis side of long-range monitoring and fault diagnosis system based on cloud service according to claim 1
Method, which is characterized in that the telecommunication gateway unit includes: Communications service module, data prediction service module, intelligence
Adaptation services module, each module deployment of telecommunication gateway unit is beyond the clouds on preposition gateway server;
The Communications service module receives the up-on command of data acquisition unit, sends working condition signal data to data prediction
Service module, the Communications service module also receives control data transmitted by cloud service center unit, and sends downlink command
To data acquisition unit;
The data prediction service module receives the working condition signal data that Communications service module is transmitted, after sending pretreatment
Obtained characteristic vector data gives cloud storage administrative unit;
The intelligent adaptation service module, provides mode adapter, for the mechanical equipment of corresponding model, specifies corresponding long-range
Communications protocol.
5. a kind of fault diagnosis side of long-range monitoring and fault diagnosis system based on cloud service according to claim 1
Method, which is characterized in that the cloud storage administrative unit includes: data storage service module, data retrieval service module and number
According to cryptographic service module, each module deployment of cloud storage administrative unit is beyond the clouds on data server;
The data storage service module receives characteristic vector data transmitted by telecommunication gateway unit and stores to corresponding
Database in;
The data retrieval service module receives inquiry request transmitted by cloud service center unit, and mesh is obtained from database
Mark data return to cloud service center unit.
6. a kind of fault diagnosis side of long-range monitoring and fault diagnosis system based on cloud service according to claim 1
Method, which is characterized in that the cloud service center unit further includes conference service module, library service module, training service mould
Block, each module are deployed in cloud application server;
The conference service module provides long-range communication and collaboration services for device manufacturer and manufacturing enterprise;
The library service module, the users at different levels for device manufacturer and manufacturing enterprise provide information or document uploads, altogether
It enjoys, query service;
The training service module, facilitates on-line teaching, and device manufacturer provides remote training and interaction religion for manufacturing enterprise
Learn service.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2016211823350 | 2016-11-03 | ||
CN201621182335 | 2016-11-03 |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107276816A CN107276816A (en) | 2017-10-20 |
CN107276816B true CN107276816B (en) | 2019-06-25 |
Family
ID=60071324
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710533956.1A Active CN107276816B (en) | 2016-11-03 | 2017-07-03 | A kind of long-range monitoring and fault diagnosis system and method for diagnosing faults based on cloud service |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107276816B (en) |
Families Citing this family (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107943002B (en) * | 2017-11-28 | 2019-10-18 | 湖南酷陆网络科技有限公司 | Sanitation equipment fault diagnosis method and system |
CN109900436A (en) * | 2017-12-08 | 2019-06-18 | 中国石油化工股份有限公司 | Valves leakage in-circuit diagnostic system and method based on cloud computing |
JP7074542B2 (en) * | 2018-04-06 | 2022-05-24 | ファナック株式会社 | Diagnostic service system and diagnostic method using network |
CN108871434B (en) * | 2018-05-30 | 2024-05-31 | 北京必创科技股份有限公司 | Online monitoring system and method for rotating equipment |
CN109102001A (en) * | 2018-07-16 | 2018-12-28 | 东南大学 | A kind of gene improve the rotor on-line fault diagnosis method of neural network |
CN109269556A (en) * | 2018-09-06 | 2019-01-25 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of equipment Risk method for early warning, device, terminal device and storage medium |
CN109144014B (en) * | 2018-10-10 | 2021-06-25 | 北京交通大学 | System and method for detecting operation condition of industrial equipment |
CN111077851A (en) * | 2018-10-22 | 2020-04-28 | 中国科学院沈阳自动化研究所 | Chemical process fault diagnosis system based on gas chromatography fog calculation framework |
CN109460828A (en) * | 2018-10-26 | 2019-03-12 | 湖北大学 | A kind of artificial intelligence deep learning method based on network cloud collaboration |
CN109129574B (en) * | 2018-11-08 | 2021-05-04 | 山东大学 | Cloud fault diagnosis system and method for service robot motion system |
CN109801364B (en) * | 2019-01-23 | 2022-09-16 | 厦门嵘拓物联科技有限公司 | Three-dimensional digital modeling method and digital workshop management system |
CN109978190A (en) * | 2019-03-29 | 2019-07-05 | 中国原子能科学研究院 | A kind of medical cyclotron remote failure diagnosis system based on artificial intelligence |
CN110430128B (en) * | 2019-06-24 | 2021-08-27 | 上海展湾信息科技有限公司 | Edge computing gateway |
CN110554657B (en) * | 2019-10-16 | 2023-03-03 | 河北工业大学 | Health diagnosis system and diagnosis method for operation state of numerical control machine tool |
CN110794799A (en) * | 2019-11-28 | 2020-02-14 | 桂林电子科技大学 | Big data system with fault diagnosis function applied to industrial production |
CN111277656A (en) * | 2020-01-21 | 2020-06-12 | 南京兴丞智能制造研究院有限公司 | Industrial big data acquisition network system and application method thereof |
CN111814991B (en) * | 2020-02-22 | 2024-07-19 | 中国原子能科学研究院 | Medical cyclotron remote fault diagnosis system based on artificial intelligence |
CN111556032A (en) * | 2020-04-14 | 2020-08-18 | 江苏天人工业互联网研究院有限公司 | Industrial big data processing system based on artificial intelligence algorithm |
CN111639742A (en) * | 2020-05-22 | 2020-09-08 | 安徽科技学院 | System and method for diagnosing state fault of desulfurization and denitrification circulating pump |
CN112235154A (en) * | 2020-09-09 | 2021-01-15 | 广州安食通信息科技有限公司 | Data processing method, system, device and medium based on Internet of things |
CN112926257A (en) * | 2020-09-25 | 2021-06-08 | 中国石油天然气集团有限公司 | Reciprocating natural gas compressor fault diagnosis system and diagnosis method |
CN112346787B (en) * | 2020-09-30 | 2022-04-19 | 广州明珞装备股份有限公司 | Non-standard equipment test system, method and storage medium |
CN112187942B (en) * | 2020-09-30 | 2022-01-11 | 武汉理工大学 | Edge computing system serving intelligent engine room |
CN114603598B (en) * | 2020-12-09 | 2024-06-21 | 炬星科技(深圳)有限公司 | Robot fault detection method, equipment and storage medium |
CN112529320A (en) * | 2020-12-18 | 2021-03-19 | 上海应用技术大学 | Intelligent diagnosis system for air compressor cluster |
CN112737829A (en) * | 2020-12-23 | 2021-04-30 | 大连理工大学人工智能大连研究院 | Method and system for integrating fault diagnosis system of excavating equipment |
CN113697424B (en) * | 2021-09-03 | 2022-11-11 | 中煤科工集团上海有限公司 | Belt conveyor monitoring and fault diagnosis system and method based on cloud technology |
CN114906383A (en) * | 2022-06-15 | 2022-08-16 | 江苏自立新材料科技有限公司 | Full-automatic film coiled material packaging integrated production process |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101893875A (en) * | 2009-05-18 | 2010-11-24 | 中国石化集团南京化学工业有限公司 | State-based machine integrated comprehensive management system |
CN102819684A (en) * | 2012-08-15 | 2012-12-12 | 西安建筑科技大学 | Remote cooperative diagnosis task allocation method |
-
2017
- 2017-07-03 CN CN201710533956.1A patent/CN107276816B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101893875A (en) * | 2009-05-18 | 2010-11-24 | 中国石化集团南京化学工业有限公司 | State-based machine integrated comprehensive management system |
CN102819684A (en) * | 2012-08-15 | 2012-12-12 | 西安建筑科技大学 | Remote cooperative diagnosis task allocation method |
Non-Patent Citations (3)
Title |
---|
一种基于改进遗传算法的神经网络优化算法研究;刘浩然,赵翠香,李轩,王艳霞,郭长江;《仪器仪表学报》;20160731;全文 |
制造装备远程监控故障诊断***研究;王海;《中国博士学位论文全文数据库》;20150715;第17页倒数第1行-第18页第3行、第20页倒数第2-8行、第21页第1行-第33页倒数第1行 |
基于遗传BP神经网络的电控发动机故障诊断研究;谢春丽,张继洲;《湖北汽车工业学院学报》;20160630;全文 |
Also Published As
Publication number | Publication date |
---|---|
CN107276816A (en) | 2017-10-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107276816B (en) | A kind of long-range monitoring and fault diagnosis system and method for diagnosing faults based on cloud service | |
US11409251B2 (en) | Modeling of an industrial automation environment in the cloud | |
US11159771B2 (en) | Virtual reality and augmented reality for industrial automation | |
CN108089696B (en) | Virtual reality and augmented reality for industrial automation | |
CN1936751B (en) | Use of a really simple syndication communication format in process control | |
CN1297926C (en) | Affair type data communication of procedue control system | |
CN106993059A (en) | A kind of agriculture feelings monitoring system based on NB IoT | |
CN109088785A (en) | A kind of IDC data center apparatus management system | |
CN107194565B (en) | Power grid scheduling optimization method and system based on cloud decision | |
CN113179190B (en) | Edge controller, edge computing system and configuration method thereof | |
CN105721194A (en) | Intelligent positioning system of faults and hidden dangers of mobile network | |
TWI751387B (en) | Software defined driven ict service provider system based on end to end orchestration | |
TW201525710A (en) | Cloud based flexible assembly system of service and method suitable for telecommunication applications | |
CN109313841A (en) | For realizing the method and system of self-adaption cluster in sensor network | |
TWI676148B (en) | A system of virtual and physical integrated network service fulfillment and monitor based on artificial intelligence | |
CN110460454A (en) | Network equipment port intelligent fault prediction technique and principle based on deep learning | |
CN110175648B (en) | Non-invasive information communication method for equipment by applying artificial intelligent cloud computing | |
Levy et al. | Emerging trends in data center management automation | |
TWI651667B (en) | Interface system of virtual and physical integrated network with resources topology | |
CN106354015B (en) | Long-range monitoring and the on-line debugging method of Diagonal Recurrent Neural Network control system | |
TWI608442B (en) | Software definition driven cloud computing network component service assembly system | |
CN111404742A (en) | Industry interconnection framework towards wisdom supply chain | |
CN103780884A (en) | Video monitoring engineering project supervision system and method | |
CN105227492A (en) | Control information is separated with information on services and the system of fusion treatment | |
CN110362614A (en) | A kind of Information Visualization System based on data-driven |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |