CN108322548A - A kind of industrial process data analyzing platform based on cloud computing - Google Patents
A kind of industrial process data analyzing platform based on cloud computing Download PDFInfo
- Publication number
- CN108322548A CN108322548A CN201810185964.6A CN201810185964A CN108322548A CN 108322548 A CN108322548 A CN 108322548A CN 201810185964 A CN201810185964 A CN 201810185964A CN 108322548 A CN108322548 A CN 108322548A
- Authority
- CN
- China
- Prior art keywords
- cluster
- sub
- cloud server
- server terminal
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- 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
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
-
- 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]
-
- 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
- H04L67/50—Network services
- H04L67/75—Indicating network or usage conditions on the user display
Abstract
The present invention discloses a kind of industrial process data analyzing platform based on cloud computing, it includes cloud server terminal, industry spot end, client and monitoring client, cloud server terminal is made of system host and n sub-cluster two parts, each sub-cluster is including sub-cluster host and the multiple slaves being connected with host, the system host controls multiple sub-cluster hosts, sub-cluster host computer control slave connected to it;The data of industrial object are read at industry spot end in real time, and are sent it to cloud server terminal and carried out analyzing processing;Client reads and shows cloud server terminal result of calculation, and instruction can be remotely sent to cloud server terminal in real time for remotely connecting cloud server terminal;Monitoring client sends control instruction to industry spot end in real time for monitoring cloud server terminal, managing user information and industrial object information, and by cloud server terminal.The platform can be realized to industrial process remote monitoring, and operation efficiency is high, can carry large-scale consumer while use.
Description
Technical field
The invention belongs to industrial process control fields, and in particular to a kind of industrial process data parsing based on cloud computing is flat
Platform.
Background technology
In industrial stokehold, industrial process data analysis software platform often plays the role of highly important, this is flat
Platform can mainly acquire in real time with storing process data, mainly realize the work based on data-driven using corresponding modeling algorithm
Industry process monitoring and hard measurement to process variable and quality variable, to remain able to even running, Improving The Quality of Products.
Wherein the Industrial Process Monitoring based on data-driven establishes mathematical model by surveying process data, to estimate
And the operating status of acquisition process, it finally identifies the abnormal behaviour of process, realizes the timely diagnosis of failure, ensure that system can
The operation of safety;And process variable and quality variable hard measurement are to use certain appropriate model, using being relatively easy to measure
Process variable data predict to be difficult to using sensor process variable measured directly and quality variable and then can preferably control
Product quality processed improves production efficiency.
Industrial process data analysis software platform is in the local computer of industry spot mostly at present, client and calculation
Method program is mutually nested, and algorithm is calculated on one computer, and data storage method is counted also with certain
Calculation machine is stored as database, therefore the operation efficiency of its algorithm, and the storage capacity of data can be restricted.However with
The continuous development of modern industry, the industrial data that can be stored is more and more, has formd a series of industrial big datas and has asked
Topic, while with the continuous rise of internet, Internet of Things, realizing network-based remote monitoring, melted using mobile client etc.
The pattern for closing " internet+industry " is also the inexorable trend of modern industry.
However traditional industrial process data analyzing platform can not expire under internet and ultra-large data-driven
The demand of sufficient modern industry, first legacy system can not store ultra-large data, if still using less number
Larger deviation is will produce according to operation is carried out;Secondly because data volume is excessive, traditional uniprocessor algorithm operational mode can not
Meet the real-time and stability calculated;The client of simultaneity factor is in the form of a single, client and algorithm routine at the scene without
Method realizes long-range monitoring and control;Algorithm in last legacy system does not have sharing and durability, it is necessary to be each work
Industry object individually writes a set of process monitoring and soft measurement algorithm, and algorithm utilization rate is relatively low, and is not easy to the management to algorithm.It is comprehensive
Upper traditional industrial process data analyzing platform is required for being improved in terms of algorithm and software platform.
Invention content
In view of the deficiencies of the prior art, the present invention proposes a kind of industrial process data analyzing platform based on cloud computing, should
Platform can realize that various algorithm height are independent, data shared resources, and operation efficiency is high, and specific technical solution is as follows:
A kind of industrial process data analyzing platform based on cloud computing, which is characterized in that the platform includes cloud server terminal, work
Industry scene end, client and monitoring client;
The cloud server terminal is made of system host and n sub-cluster two parts, the system host and each height
Cluster is in same LAN, wherein is installed as monitoring client in the system host and is provided the Web server of service, drives
The load equalizer of the data server of dynamic system host database and connection and each sub-cluster of management;Each sub-cluster is equal
The multiple slaves being connected including sub-cluster host and with host, the system host controls multiple sub-cluster hosts, described
Sub-cluster host computer control slave connected to it;
The industry spot end includes multiple data processing modules, and each data processing module binds corresponding industry
The sensor of object reads the data of industrial object during the work time, and sends it to cloud server terminal and analyzed in real time
Processing;The data processing module is additionally operable to remotely receive the control instruction sent by client and monitoring client, described in realization
Data processing module gathered data start and stop, the data processing module can also set corresponding according to actual field situation
Industrial object control operation;
The client is for remotely connecting cloud server terminal, reading the result of calculation of cloud server terminal in real time and carrying out visual
Change display, and remotely send instruction to cloud server terminal as needed, is adopted by the data at cloud server terminal remote control industry spot end
Collect the start and stop of module data collection work and the control of industrial object is operated under conditions of allowing at the scene;
The monitoring client for monitoring cloud server terminal working condition and industrial object operation data, managing user information and
Industrial object information, and control instruction is sent to industry spot end by cloud server terminal in real time.
Further, the interactive process of the industry spot end, cloud server terminal and client is as follows:
The industry spot end logs in the system host of cloud server terminal, and system host calculates subset for its distribution is corresponding
Group;Industry spot end obtains variable data in real time from industry spot local data base, and it is reached according to certain period
The corresponding sub-cluster of cloud server terminal carries out Algorithm Analysis, and data results are stored in the database of system host by cloud server terminal,
Client reads the data results, and visualized by being polled to system host database;
Client sends to cloud server terminal and instructs, and changes the flag bit of respective field in system host database, and industry is existing
Field end is polled system host database respective field, reads the flag bit, starts corresponding control operation.
Further, the interaction of the monitoring client, cloud server terminal and industry spot end includes:
The monitoring client is polled by the system host database to cloud server terminal, and by data visualization, into
And monitor the operating status at industry spot end;Monitoring client accesses the system host database in cloud server terminal, reads user information,
Realization, which changes the additions and deletions of user information, looks into;The running parameter Mobile state of going forward side by side that monitoring client reads system host and sub-cluster in real time is aobvious
Show;
Monitoring client sends to cloud server terminal and instructs, and changes the flag bit of system host database respective field, industry spot
End is polled system host database respective field, reads the flag bit, starts corresponding control operation.
Further, the process of the load equalizer equally loaded is as follows:
Load equalizer reads the work state information for the sub-cluster being attached thereto, including each subset every same time
The static configuration of group node and dynamic property during the work time, the dynamic property includes cpu frequency, EMS memory occupation
Situation, magnetic disc i/o reading rate and communication network bandwidth usage, are dynamically calculated using load-balancing algorithm after reading
It should be at this time the number of users of each sub-cluster distribution, and then change each client connexon stored in system host database
The field of clustered node IP address, client read the IP of distributed calculate node by being polled to the field
Address, and be connected with the node, it is that each sub-cluster allocates user again to realize.
Further, the load equalizer uses the Dynamic Load-balancing Algorithm based on load weights probability to control institute
The load for the sub-cluster stated distributes, and the algorithm includes the following steps:
Step 1:The Integrated Static performance indicator C (s of calculate node in definition description sub-clusterij) and real time and dynamic energy
Index L (sij):
C(sij)=k1×nij×C(cij)+k2×C(mij)+k3×C(dij)+k4×C(pij)
L(sij)=k1×L(cij)+k2×L(mij)+k3×L(dij)+k4×L(pij)
Wherein, i=1 ..., N, j=1 ..., Mi, the sub-cluster number that N is connected by system host, MiFor i-th of subset
Lower the connected calculate node number of group, sijFor j-th of calculate node under i-th of sub-cluster, nijFor the calculate node
CPU core number, cijRepresent the service condition of node cpu, C (cij) represent preconfigured static frequency, L (cij) when representing work
Real-time dynamic cpu frequency;mijRepresent the service condition of node memory, wherein C (mij) represent preconfigured memory size, L
(mij) represent real-time Dram occupancy when work;dijRepresent the service condition of node magnetic disc i/o, C (dij) represent in advance
The magnetic disc i/o reading rate of configuration, L (dij) represent real-time dynamic disk I/O occupancies when work;pijRepresent each node net
Network service condition, wherein C (pij) represent preconfigured meshed network handling capacity, L (pij) represent the real-time of meshed network bandwidth
Occupancy, wherein k1,k2,k3,k4For weight coefficient, and k1+k2+k3+k4=1;
Step 2:By the static performance index of calculate node and the load weights W (s of dynamic performance index definition nodeij)
Step 3:The load weights summation of each node is taken into mean value, obtains the load weights W (S of each sub-clusteri)
Step 4:It incites somebody to action maximum load weights in sometime each sub-cluster and is expressed as Wmax(S), by itself and institute at this time
There are the load weights of sub-cluster to make poor, obtains load difference Δ Q (Si)
ΔQ(Si)=Wmax(S)-W(Si)
Step 5:Define probability P (S of the system host to each sub-cluster distributing useri)
Step 6:The number of users of the currently used platform is multiplied with user's allocation probability, obtaining each sub-cluster should
The number of users of distribution
R(S1)=floor (P (S1)×Num)
R(S2)=floor (P (S2)×Num)
R(Sk)=0
Wherein Num is the number of the currently used platform, and floor is downward rounding, R (Si) each sub-cluster distributed
Number of users.
Further, the system host database is for storing bound in essential information and user of user
The control instruction that industrial object information, client and monitoring client are sent and the analysis data obtained from sub-cluster.
Further, the subset group configuration distributed document storage architecture and based on Map-Reduce it is distributed simultaneously
Row Computational frame, the sub-cluster host be scheduling node, the sub-cluster host connection slave be back end and
Calculate node, all process monitorings and variable soft measurement algorithm are programmed in each sub-cluster in the form of Map-Reduce
In slave calculate node on, and based on MVC programming models be each algorithm formed a functional interface be stored in each height
In cluster system, while the data also distributed storage in the database of each calculate node.
Further, the client operates on PC and/or on mobile terminal, and the monitoring client is in the form of web pages
Operation.
Further, the process monitoring and the soft survey of variable configured in the calculate node of the sub-cluster of the cloud server terminal
Quantity algorithm all has distributed parallel pattern and single cpu mode.
Further, the client can also directly upload local data and carry out data analysis to cloud server terminal.
Beneficial effects of the present invention are as follows:
The industrial process data analyzing platform based on cloud computing of the present invention, by configuring multiple collection in Cloud Server
Group, each cluster are mounted on distributive parallel computation framework, and the monitoring of all distributed industrials and soft measurement algorithm are taken in cloud
It is write in business device, and forms the calling interface of each algorithm, while realizing that all algorithms can be using multithreading
The same time is used;Industry spot end real-time data collection is simultaneously remotely sent to high in the clouds and is calculated, and client takes from cloud again
Result of calculation is read in business device and is visualized.Realize that process monitoring and soft measurement algorithm height are independent, shared resources, and promote fortune
Calculate efficiency.
Description of the drawings
Fig. 1 is the integrated stand composition of the industrial process data analyzing platform based on cloud computing of the present invention;
Fig. 2 is the configuration diagram of cloud server terminal;
Fig. 3 is the fundamental diagram of the load equalizer of cloud server terminal;
Fig. 4 is file distributed storage architecture schematic diagram in sub-cluster;
Fig. 5 is the configuration diagram of distributed computing framework in sub-cluster;
Fig. 6 is the interactive process schematic diagram at industry spot end and cloud server terminal, client;
Fig. 7 is client and cloud server terminal, the interactive process schematic diagram at industry spot end;
Fig. 8 is monitoring client and cloud server terminal, the interactive process schematic diagram at industry spot end.
Specific implementation mode
Below according to attached drawing and the preferred embodiment detailed description present invention, the objects and effects of the present invention will become brighter
In vain, below in conjunction with drawings and examples, the present invention will be described in further detail.It should be appreciated that described herein specific
Embodiment is only used to explain the present invention, is not intended to limit the present invention.
As shown in Figure 1, a kind of industrial process data analyzing platform based on cloud computing, the platform include cloud server terminal, work
Industry scene end, client and monitoring client;
Wherein cloud server terminal is made of system host and n sub-cluster two parts, the system host and each subset
Group is in same LAN, wherein is installed as monitoring client in the system host and provides the Web server of service, driving
The load equalizer of the data server of system host database and connection and each sub-cluster of management;Each sub-cluster is wrapped
Enclosed tool cluster system and the multiple slaves being connected with host, the system host control multiple sub-cluster hosts, the son
Cluster system controls slave connected to it;Subset group configuration distributed document storage architecture and the distribution based on Map-Reduce
Formula parallel computation frame, the sub-cluster host are scheduling node, and the slave of sub-cluster host connection is data section
Point and calculate node, all process monitorings and variable soft measurement algorithm are programmed in each height in the form of Map-Reduce
In the calculate node of slave in cluster, and it is that each algorithm one functional interface of formation is stored in respectively based on MVC programming models
In a sub- cluster system, while the data also distributed storage in the database of each calculate node.
Industry spot end is to present in a software form, including multiple data processing modules, each data processing module are tied up
The sensor of fixed corresponding industrial object reads the data of industrial object during the work time in real time, and sends it to cloud clothes
Business end carries out analyzing processing;The data processing module is additionally operable to remotely receive to be referred to by the control of client and monitoring client transmission
It enables, realizes the start and stop of the data processing module gathered data, the data processing module can also be according to actual field feelings
Condition sets corresponding industrial object control operation;
Client is used to remotely connect cloud server terminal, reads the result of calculation of cloud server terminal in real time and visualize aobvious
Show, and remotely send instruction to cloud server terminal as needed, by the data acquisition module at cloud server terminal remote control industry spot end
The start and stop of block data collection task and the control of industrial object is operated under conditions of allowing at the scene;Client operates in PC
On upper and/or mobile terminal.
Monitoring client uses form web page, and for monitoring cloud server terminal working condition and industrial object operation data, management is used
Family information and industrial object information include the log-on message of editor user, increase and logging off users, check that current cloud server terminal is tied up
Fixed industrial object information, and control instruction is sent to industry spot end by cloud server terminal in real time.Administrator can be any
When and where logs in monitoring client webpage and is managed.
The specific works for specifically introducing platform cloud server terminal, industry spot end, client and monitoring client separately below are former
Reason and function.
1. cloud server terminal
The specific framework of cloud server terminal is as shown in Fig. 2, lower mask body introduces the concrete function of modules in cloud server terminal
And the course of work.
1.1 system host
1.1.1 data server
Data server is mainly used for drive system host data base, and obtains in sub-cluster and be stored in distributed data
Algorithm Analysis result in library.User information tables of data wherein in system host database is used for storing stepping on for user first
Land, log-on message, using information such as duration, time buyings, next stores industrial object information, client bound in the user
And the control instruction that sends of monitoring client and the analysis data that are obtained from sub-cluster, (such as with PCA (pivot analysis) progress processes
Monitoring, then the analysis data obtained from sub-cluster are T2Control with SPE statistics and the two limits), so as to client
It reads data using data server with monitoring client and sends and instruct.
1.1.2Web server
Web server is used as the server of driving monitoring client webpage, i.e. all programs of monitoring client web-site exist
It is executed in Web server in system host.It is instructed since monitoring client needs to send to high in the clouds in real time, reads and manage storage
User information in system host database, industrial object information, data etc. are analyzed in high in the clouds, therefore monitoring client needs pass through
The database that Web server accesses system host carries out the above-mentioned dynamic operation to system host database.
1.1.3 load equalizer
It is to solve a large number of users while accessing that the purpose of each sub-cluster load balancing is realized in system host
Caused load is uneven when some sub-cluster.When the live load of some sub-cluster is very big, operational capability and fortune
Calculating speed can substantially reduce, and the load of other sub-clusters may be less, it is therefore desirable to (use the load of big load sub-cluster
Family) other sub-clusters are distributed to, the working condition to make each sub-cluster keep best ensure that Cloud Server entirety
Arithmetic speed.
The designed load equalizer operation principle in this platform is as shown in figure 3, load equalizer is mounted on system master
In machine, load equalizer reads the work state information for the sub-cluster being attached thereto, including each sub-cluster every same time
The static configuration of node and dynamic property during the work time, the dynamic property includes cpu frequency, EMS memory occupation feelings
Condition, magnetic disc i/o reading rate and communication network bandwidth usage, dynamically calculate this using load-balancing algorithm after reading
When should be number of users of each sub-cluster distribution, then change each client connected dominating set for storing in system host database
The field of group node IP address, client is by being polled the field, with reading the IP of distributed calculate node
Location, and be connected with the node, it is that each sub-cluster allocates user again to realize.
And load equalizer designed in this platform, internal load balancing are based on load weights probability
The principle of Dynamic Load-balancing Algorithm, the algorithm is:
Step 1:The Integrated Static performance indicator C (s of calculate node in definition description sub-clusterij) and real time and dynamic energy
Index L (sij):
C(sij)=k1×nij×C(cij)+k2×C(mij)+k3×C(dij)+k4×C(pij)
L(sij)=k1×L(cij)+k2×L(mij)+k3×L(dij)+k4×L(pij)
Wherein, i=1 ..., N, j=1 ..., Mi, the sub-cluster number that N is connected by system host, MiFor i-th of subset
Lower the connected calculate node number of group, sijFor j-th of calculate node under i-th of sub-cluster, nijFor the calculate node
CPU core number, cijRepresent the service condition of node cpu, C (cij) represent preconfigured static frequency, L (cij) when representing work
Real-time dynamic cpu frequency;mijRepresent the service condition of node memory, wherein C (mij) represent preconfigured memory size, L
(mij) represent real-time Dram occupancy when work;dijRepresent the service condition of node magnetic disc i/o, C (dij) represent in advance
The magnetic disc i/o reading rate of configuration, L (dij) represent real-time dynamic disk I/O occupancies when work;pijRepresent each node net
Network service condition, wherein C (pij) represent preconfigured meshed network handling capacity, L (pij) represent the real-time of meshed network bandwidth
Occupancy, wherein k1,k2,k3,k4For weight coefficient, and k1+k2+k3+k4=1;
Step 2:By the static performance index of calculate node and the load weights W (s of dynamic performance index definition nodeij)
Step 3:The load weights summation of each node is taken into mean value, obtains the load weights W (S of each sub-clusteri)
Step 4:It incites somebody to action maximum load weights in sometime each sub-cluster and is expressed as Wmax(S), by itself and institute at this time
There are the load weights of sub-cluster to make poor, obtains load difference Δ Q (Si)
ΔQ(Si)=Wmax(S)-W(Si)
Step 5:Define probability P (S of the system host to each sub-cluster distributing useri)
Step 6:The number of users of the currently used platform is multiplied with user's allocation probability, obtaining each sub-cluster should
The number of users of distribution
R(S1)=floor (P (S1)×Num)
R(S2)=floor (P (S2)×Num)
R(Sk)=0
Wherein Num is the number of the currently used platform, and floor is downward rounding, R (Si) each sub-cluster distributed
Number of users.
1.2 sub-cluster
1.2.1 the file distribution storage architecture in sub-cluster
It can be seen from the above, in order to improve storage and the extraction rate of mass data, this platform is stored using distributed document
Framework, and in the less database being only stored in attached by system host of amount of user data, and be largely used to calculate and
The data of algorithm modeling are stored in each sub-cluster with just needing parallel distributed.
For each sub-cluster, cluster system is considered as scheduling node, and the slave connected is considered as counting
According to node.Data are carried out piecemeal by the scheduling node when carrying out file storage in cluster system according to certain size first,
The allocation result of every block number evidence is notified into the back end in slave again, finally by each back end by obtained data
In the disk of computer or server where block deposit.
When carrying out file reading, equally notify to extract required for each back end by the scheduling node in cluster system
File, corresponding data block concurrently can be extracted and be spliced into new as desired by each back end after completion notice
File is sent to client.
Above-mentioned file distribution storing process is as shown in Figure 4.
1.2.2 the distributed computing framework in sub-cluster
The part that algorithm operation is mainly completed in this platform is the calculate node in sub-cluster, is being handled to improve algorithm
Ability during big data, this platform are configured with the Distributed Parallel Computing frame based on MapReduce in each sub-cluster
Frame.
Distributive parallel computation framework is similar with file distribution storing framework, and each by cluster system management and scheduling
The operation process of a calculate node calculates data block by multiple calculate nodes under its administration to realize task or calculation simultaneously
Method it is parallel.
And MapReduce is one kind of Distributed Parallel Computing, it mainly will be stored in each meter by Map links
Data block in operator node is extracted parallel and handles and be mapped to key-value pair form, and the disk of each calculate node is then stored into
In, and Reduce links then concurrently extract the Map key-value pairs generated, and carried out according to the data with identical key assignments
Reduction, then the data after reduction are handled.
Above-mentioned MapReduce processes are completed in calculate node, and cluster system remain responsible for it is each during this
Scheduling between item process, therefore the distributed computing framework in sub-cluster is as shown in Figure 5.
1.2.3 process monitoring and soft measurement algorithm library
In each sub-cluster include process monitoring and soft measurement algorithm library, the algorithms library from the method for operation come divide including
Distributed Parallel Algorithm and uniprocessor algorithm;Divide from functional perspective including process monitoring algorithm and soft measurement algorithm.It is all
The code of algorithm is stored in the calculate node of sub-cluster, and it is incoming to form parameter and data, the calling interface as a result spread out of.
When multiple users use identical algorithms simultaneously, calculate node can handle two simultaneously with the pattern of multithreading and ask
It asks, and an exclusive file is formed for each user so that while large-scale consumer may be implemented in all algorithms
It uses.
User can also directly be uploaded local data by client and carry out data analysis to cloud server terminal.
The interactive process between each end of platform is described below.
Industry spot end is with the interactive process of cloud server terminal and client as shown in fig. 6, industry spot end logs in cloud service
The system host at end, system host calculate sub-cluster for its distribution is corresponding;Industry spot end is from industry spot local data base
In obtain variable data in real time, and it is reached into the corresponding sub-cluster of cloud server terminal according to certain period and carries out Algorithm Analysis,
Data results are stored in the database of system host by cloud server terminal, and client is by taking turns system host database
It askes, reads the data results, and visualized;
Client sends to cloud server terminal and instructs, and changes the flag bit of respective field in system host database, and industry is existing
Field end is polled system host database respective field, reads the flag bit, starts corresponding control operation.
Client is with cloud server terminal, the interactive process at industry spot end as shown in fig. 7, client reads cloud server terminal in real time
The method of data is still used to cloud server terminal system host database polling technique, i.e., every certain period to cloud system
Host data base reads an analysis result, and carries out visualization in client and show.Client is sent out to industrial object simultaneously
It is also by being updated to system host database relevant field when sending instruction, industry spot end can also be to system host number
It is polled according to library and then reads the field, and then the local data base of Let it Be update industrial monitoring and control software completes control behaviour
Make.
Monitoring client and the interaction of cloud server terminal, industry spot end are as shown in Figure 8.Monitoring client is to present in the form of a web page
, mainly there is cloud server Working Status Monitoring, subscriber information management, industrial object information management, industrial object operation
The functions such as data monitoring and the simple control instruction of transmission.
Monitoring client is polled by the system host database to cloud server terminal, and by data visualization, and then is monitored
The operating status at industry spot end;Monitoring client accesses the system host database in cloud server terminal, reads user information, realization pair
The additions and deletions of user information, which change, looks into;The running parameter Mobile state of going forward side by side that monitoring client reads system host and sub-cluster in real time shows, and
Draw real-time performance graph;
Monitoring client sends to cloud server terminal and instructs, and changes the flag bit of system host database respective field, industry spot
End is polled system host database respective field, reads the flag bit, starts corresponding control operation.
It will appreciated by the skilled person that the foregoing is merely the preferred embodiment of invention, it is not used to limit
System invention, although invention is described in detail with reference to previous examples, for those skilled in the art, still
It can modify to the technical solution of aforementioned each case history or equivalent replacement of some of the technical features.It is all
Within the spirit and principle of invention, modification, equivalent replacement for being made etc. should be included within the protection domain of invention.
Claims (10)
1. a kind of industrial process data analyzing platform based on cloud computing, which is characterized in that the platform includes cloud server terminal, industry
Live end, client and monitoring client;
The cloud server terminal is made of system host and n sub-cluster two parts, the system host and each sub-cluster
In same LAN, wherein be installed as monitoring client in the system host and the Web server of service, driving system are provided
The load equalizer of the data server of system host data base and connection and each sub-cluster of management;Each sub-cluster includes
Sub-cluster host and the multiple slaves being connected with host, the system host control multiple sub-cluster hosts, the subset
Group's host computer control slave connected to it;
The industry spot end includes multiple data processing modules, and each data processing module binds corresponding industrial object
Sensor, read industrial object data during the work time in real time, and send it to cloud server terminal and carry out analyzing processing;
The data processing module is additionally operable to remotely receive the control instruction sent by client and monitoring client, realizes the data
The start and stop of processing module gathered data, it is right that the data processing module can also set corresponding industry according to actual field situation
As control operates;
The client is used to remotely connect cloud server terminal, reads the result of calculation of cloud server terminal in real time and visualize aobvious
Show, and remotely send instruction to cloud server terminal as needed, by the data acquisition module at cloud server terminal remote control industry spot end
The start and stop of block data collection task and the control of industrial object is operated under conditions of allowing at the scene;
The monitoring client is for monitoring cloud server terminal working condition and industrial object operation data, managing user information and industry
Object information, and control instruction is sent to industry spot end by cloud server terminal in real time.
2. the industrial process data analyzing platform according to claim 1 based on cloud computing, which is characterized in that the work
The interactive process at industry scene end, cloud server terminal and client is as follows:
The industry spot end logs in the system host of cloud server terminal, and system host calculates sub-cluster for its distribution is corresponding;
Industry spot end obtains variable data in real time from industry spot local data base, and it is reached cloud clothes according to certain period
The corresponding sub-cluster in end of being engaged in carries out Algorithm Analysis, and data results are stored in the database of system host, client by cloud server terminal
The data results are read, and visualized by being polled to system host database in end;
Client sends to cloud server terminal and instructs, and changes the flag bit of respective field in system host database, industry spot end
System host database respective field is polled, the flag bit is read, starts corresponding control operation.
3. the industrial process data analyzing platform according to claim 1 based on cloud computing, which is characterized in that the prison
Control end, cloud server terminal and industry spot end interaction include:
The monitoring client is polled by the system host database to cloud server terminal, and by data visualization, Jin Erjian
Control the operating status at industry spot end;Monitoring client accesses the system host database in cloud server terminal, reads user information, realizes
The additions and deletions of user information are changed and are looked into;The running parameter Mobile state of going forward side by side that monitoring client reads system host and sub-cluster in real time is shown;
Monitoring client sends to cloud server terminal and instructs, and changes the flag bit of system host database respective field, industry spot end pair
System host database respective field is polled, and reads the flag bit, starts corresponding control operation.
4. the industrial process data analyzing platform according to claim 1 based on cloud computing, which is characterized in that described is negative
The process for carrying balanced device equally loaded is as follows:
Load equalizer reads the work state information for the sub-cluster being attached thereto, including each sub-cluster section every same time
The static configuration and dynamic property during the work time of point, the dynamic property includes cpu frequency, EMS memory occupation feelings
Condition, magnetic disc i/o reading rate and communication network bandwidth usage, dynamically calculate this using load-balancing algorithm after reading
When should be number of users of each sub-cluster distribution, then change each client connected dominating set for storing in system host database
The field of group node IP address, client is by being polled the field, with reading the IP of distributed calculate node
Location, and be connected with the node, it is that each sub-cluster allocates user again to realize.
5. the industrial process data analyzing platform according to claim 4 based on cloud computing, which is characterized in that described is negative
The load distribution that balanced device uses the Dynamic Load-balancing Algorithm control sub-cluster based on load weights probability is carried, it is described
Algorithm include the following steps:
Step 1:The Integrated Static performance indicator C (s of calculate node in definition description sub-clusterij) and real time and dynamic energy index L
(sij):
C(sij)=k1×nij×C(cij)+k2×C(mij)+k3×C(dij)+k4×C(pij)
L(sij)=k1×L(cij)+k2×L(mij)+k3×L(dij)+k4×L(pij)
Wherein, i=1 ..., N, j=1 ..., Mi, the sub-cluster number that N is connected by system host, MiFor under i-th of sub-cluster
The calculate node number connected, sijFor j-th of calculate node under i-th of sub-cluster, nijFor the CPU core of the calculate node
Number, cijRepresent the service condition of node cpu, C (cij) represent preconfigured static frequency, L (cij) represent it is real-time when work
Dynamic cpu frequency;mijRepresent the service condition of node memory, wherein C (mij) represent preconfigured memory size, L (mij) generation
Table real-time Dram occupancy when working;dijRepresent the service condition of node magnetic disc i/o, C (dij) represent it is preconfigured
Magnetic disc i/o reading rate, L (dij) represent real-time dynamic disk I/O occupancies when work;pijEach meshed network is represented to use
Situation, wherein C (pij) represent preconfigured meshed network handling capacity, L (pij) represent the real-time occupancy of meshed network bandwidth
Rate, wherein k1,k2,k3,k4For weight coefficient, and k1+k2+k3+k4=1;
Step 2:By the static performance index of calculate node and the load weights W (s of dynamic performance index definition nodeij)
Step 3:The load weights summation of each node is taken into mean value, obtains the load weights W (S of each sub-clusteri)
Step 4:It incites somebody to action maximum load weights in sometime each sub-cluster and is expressed as Wmax(S), by itself and all sons at this time
It is poor that the load weights of cluster are made, and obtains load difference Δ Q (Si)
ΔQ(Si)=Wmax(S)-W(Si)
Step 5:Define probability P (S of the system host to each sub-cluster distributing useri)
Step 6:The number of users of the currently used platform is multiplied with user's allocation probability, obtaining each sub-cluster should distribute
Number of users
Wherein Num is the number of the currently used platform, and floor is downward rounding, R (Si) user that is distributed of each sub-cluster
Number.
6. the industrial process data analyzing platform according to claim 2 based on cloud computing, which is characterized in that described is
The essential information and industrial object information, client and monitoring bound in the user that system host data base is used to store user
Hold the control instruction sent and the analysis data obtained from sub-cluster.
7. the industrial process data analyzing platform according to claim 1 based on cloud computing, which is characterized in that the son
Cluster configuration distributed document storage architecture and distributive parallel computation framework based on Map-Reduce, the sub-cluster master
Machine is scheduling node, the sub-cluster host connection slave be back end and calculate node, all process monitorings with
And variable soft measurement algorithm is programmed in the form of Map-Reduce in the calculate node of the slave in each sub-cluster, and base
It is that each algorithm one functional interface of formation is stored in each sub-cluster host, while data are also each in MVC programming models
Distributed storage in the database of a calculate node.
8. the industrial process data analyzing platform according to claim 1 based on cloud computing, which is characterized in that the visitor
Family end operates on PC and/or on mobile terminal, and the monitoring client is run in the form of web pages.
9. the industrial process data analyzing platform according to claim 1 based on cloud computing, which is characterized in that the cloud
The process monitoring and variable soft measurement algorithm configured in the calculate node of the sub-cluster of server-side all has distributed parallel mould
Formula and single cpu mode.
10. the industrial process data analyzing platform according to claim 1 based on cloud computing, which is characterized in that described
Client can also directly upload local data and carry out data analysis to cloud server terminal.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810185964.6A CN108322548B (en) | 2018-03-07 | 2018-03-07 | Industrial process data analysis platform based on cloud computing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810185964.6A CN108322548B (en) | 2018-03-07 | 2018-03-07 | Industrial process data analysis platform based on cloud computing |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108322548A true CN108322548A (en) | 2018-07-24 |
CN108322548B CN108322548B (en) | 2020-06-05 |
Family
ID=62901354
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810185964.6A Active CN108322548B (en) | 2018-03-07 | 2018-03-07 | Industrial process data analysis platform based on cloud computing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108322548B (en) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109889585A (en) * | 2019-02-01 | 2019-06-14 | 中电科仪器仪表有限公司 | A kind of the operating status remote supervision system and method for electronic testing instrument |
CN110071918A (en) * | 2019-04-19 | 2019-07-30 | 北京那镁克科技有限公司 | A kind of data communications method and device based on mixed cloud |
CN110716986A (en) * | 2019-10-17 | 2020-01-21 | 华普通用技术研究(广州)有限公司 | Big data analysis system and application method thereof |
CN111221807A (en) * | 2019-12-25 | 2020-06-02 | 中国特种设备检测研究院 | Cloud service-oriented industrial equipment big data quality testing method and architecture |
CN111638672A (en) * | 2020-06-05 | 2020-09-08 | 昆山艾控智能科技有限公司 | Automatic control system of industrial machine table |
CN111885146A (en) * | 2020-07-21 | 2020-11-03 | 合肥学院 | Industrial data cloud service platform data transmission method for new energy automobile drive motor assembly production line |
CN112051771A (en) * | 2020-09-01 | 2020-12-08 | 北京视界云天科技有限公司 | Multi-cloud data acquisition method and device, computer equipment and storage medium |
CN112217894A (en) * | 2020-10-12 | 2021-01-12 | 浙江大学 | Load balancing system based on dynamic weight |
CN112822049A (en) * | 2021-01-05 | 2021-05-18 | 浙江工业大学 | Remote monitoring system for access of multiple embedded gateway devices |
CN113156909A (en) * | 2021-05-30 | 2021-07-23 | 广东新征程科技有限公司 | Ultrashort wave signal transmission automatic production line system |
CN113518127A (en) * | 2021-07-15 | 2021-10-19 | 罗慧 | Industrial Internet of things information integration platform |
CN113566864A (en) * | 2021-09-03 | 2021-10-29 | 合肥米克光电技术有限公司 | Distributed machine vision system based on 5G and edge calculation |
CN115277704A (en) * | 2022-09-27 | 2022-11-01 | 美冠(北京)科技有限公司 | Cloud playing system based on distributed preloading |
CN116366660A (en) * | 2023-03-31 | 2023-06-30 | 广州大学 | Communication management intelligent system and method for distributed parallel simulation calculation |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102981440A (en) * | 2012-11-02 | 2013-03-20 | 武汉理工大学 | Intelligent device monitoring and managing system based on software as a service (SaaS) |
CN103546536A (en) * | 2013-08-28 | 2014-01-29 | 北京清控人居环境研究院有限公司 | Internet of things system of sewage treatment plant |
US20170238073A1 (en) * | 2016-02-16 | 2017-08-17 | Premiere Concrete Solutions, LLC | Concrete Maturity System |
-
2018
- 2018-03-07 CN CN201810185964.6A patent/CN108322548B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102981440A (en) * | 2012-11-02 | 2013-03-20 | 武汉理工大学 | Intelligent device monitoring and managing system based on software as a service (SaaS) |
CN103546536A (en) * | 2013-08-28 | 2014-01-29 | 北京清控人居环境研究院有限公司 | Internet of things system of sewage treatment plant |
US20170238073A1 (en) * | 2016-02-16 | 2017-08-17 | Premiere Concrete Solutions, LLC | Concrete Maturity System |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109889585B (en) * | 2019-02-01 | 2022-02-15 | 中电科思仪科技股份有限公司 | Remote monitoring system and method for running state of electronic test instrument |
CN109889585A (en) * | 2019-02-01 | 2019-06-14 | 中电科仪器仪表有限公司 | A kind of the operating status remote supervision system and method for electronic testing instrument |
CN110071918A (en) * | 2019-04-19 | 2019-07-30 | 北京那镁克科技有限公司 | A kind of data communications method and device based on mixed cloud |
CN110071918B (en) * | 2019-04-19 | 2020-10-09 | 北京那镁克科技有限公司 | Data communication method and device based on hybrid cloud |
CN110716986A (en) * | 2019-10-17 | 2020-01-21 | 华普通用技术研究(广州)有限公司 | Big data analysis system and application method thereof |
CN111221807A (en) * | 2019-12-25 | 2020-06-02 | 中国特种设备检测研究院 | Cloud service-oriented industrial equipment big data quality testing method and architecture |
CN111638672A (en) * | 2020-06-05 | 2020-09-08 | 昆山艾控智能科技有限公司 | Automatic control system of industrial machine table |
CN111638672B (en) * | 2020-06-05 | 2024-01-16 | 昆山艾控智能科技有限公司 | Automatic control system of industrial machine |
CN111885146A (en) * | 2020-07-21 | 2020-11-03 | 合肥学院 | Industrial data cloud service platform data transmission method for new energy automobile drive motor assembly production line |
CN111885146B (en) * | 2020-07-21 | 2022-06-24 | 合肥学院 | Data transmission method for industrial data cloud service platform |
CN112051771A (en) * | 2020-09-01 | 2020-12-08 | 北京视界云天科技有限公司 | Multi-cloud data acquisition method and device, computer equipment and storage medium |
CN112217894A (en) * | 2020-10-12 | 2021-01-12 | 浙江大学 | Load balancing system based on dynamic weight |
CN112822049A (en) * | 2021-01-05 | 2021-05-18 | 浙江工业大学 | Remote monitoring system for access of multiple embedded gateway devices |
CN113156909A (en) * | 2021-05-30 | 2021-07-23 | 广东新征程科技有限公司 | Ultrashort wave signal transmission automatic production line system |
CN113518127A (en) * | 2021-07-15 | 2021-10-19 | 罗慧 | Industrial Internet of things information integration platform |
CN113566864A (en) * | 2021-09-03 | 2021-10-29 | 合肥米克光电技术有限公司 | Distributed machine vision system based on 5G and edge calculation |
CN115277704A (en) * | 2022-09-27 | 2022-11-01 | 美冠(北京)科技有限公司 | Cloud playing system based on distributed preloading |
CN115277704B (en) * | 2022-09-27 | 2022-12-27 | 美冠(北京)科技有限公司 | Cloud playing system based on distributed preloading |
CN116366660A (en) * | 2023-03-31 | 2023-06-30 | 广州大学 | Communication management intelligent system and method for distributed parallel simulation calculation |
Also Published As
Publication number | Publication date |
---|---|
CN108322548B (en) | 2020-06-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108322548A (en) | A kind of industrial process data analyzing platform based on cloud computing | |
CN103152393B (en) | A kind of charging method of cloud computing and charge system | |
US10922452B2 (en) | Digital simulation system of power distribution network | |
Sharma et al. | Energy-efficient resource allocation and migration in private cloud data centre | |
US8756441B1 (en) | Data center energy manager for monitoring power usage in a data storage environment having a power monitor and a monitor module for correlating associative information associated with power consumption | |
CN105677836A (en) | Big data processing and solving system simultaneously supporting offline data and real-time online data | |
CN105893628A (en) | Real-time data collection system and method | |
CN107465708A (en) | A kind of CDN bandwidth scheduling systems and method | |
CN108874640A (en) | A kind of appraisal procedure and device of clustering performance | |
US20150271023A1 (en) | Cloud estimator tool | |
CN112860695B (en) | Monitoring data query method, device, equipment, storage medium and program product | |
CN106101196B (en) | A kind of cloud rendering platform task scheduling system based on probabilistic model | |
CN103581339A (en) | Storage resource allocation monitoring and processing method based on cloud computing | |
CN113505993B (en) | Distribution center management method, device, equipment and storage medium | |
CN103561078A (en) | Telecom operation system and service implementation method | |
CN108777637A (en) | A kind of data center's total management system and method for supporting server isomery | |
CN109241096A (en) | Data processing method, device and system | |
CN107645410A (en) | A kind of virtual machine management system and method based on OpenStack cloud platforms | |
CN103595815A (en) | Method for distributing storage resources based on cloud computing | |
Parygin et al. | A convergent model for distributed processing of Big Sensor Data in urban engineering networks | |
CN112365366A (en) | Micro-grid management method and system based on intelligent 5G slice | |
Gupta et al. | Long range dependence in cloud servers: a statistical analysis based on google workload trace | |
CN106170744B (en) | Energy consumption of virtual machine determines method, physical machine and network system | |
CN107908459B (en) | Cloud computing scheduling system | |
CN116231860B (en) | Cloud edge end cooperation-based intelligent power load identification system, method and equipment |
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 |