- Email: [email protected]
Quality Instrumentation Data Integrated Standard System in Cold Rolling Mill
IFAC Workshop Automation in MMM Industry
20-22 September 2006
Cracow, Poland
QUALITY INSTRUMENTATION DATA INTEGRATED STANDARD SYSTEM IN COLD ROLLING MILL
Yeongbok Park, Jaehwa Hong, Wonho Hwang, Kiwon Kim Instulllentation Research Group, Technial Lab. POSCO 1 goidong, namgu, pohang, Korea Tri [email protected] Authors' affiliations, Italics, upper and lower case, ce Ilf red, 127 mill (5 in) width (maximum)
Abstract: This paper proposed a standardization of instrumentation system aimed at standardizing the system architecture and the design methods on quality instrumentation of coil rolling mill at steel works. In first, as developing instrumentation controller, we proposed the WW speci tication and standard I/O library to be used on various device driver and common components such as communication and device handler. In second, we proposed Quality Instrumentation Data Integrated Server that the system can integrate the data about product and instrumentation and users can access and analyse information through web browser in all plan area. Copyright © 2005 IFAC Keywords : cold rolling mill, quality instrumentation, standardization, component, system integration
as communication and device handler. In second, we propose Quality Instrumentation Data Integrated Server that can integrate the data about product and instrumentation and access and analyse information through web browser in all plan area.
1. INTRODUCTION
In the major application of instrumentation and quality management on steel product, the architecture of the instrumentation system is an essential factor because it directly affects product quality work such as the range and time of quality analysis. For this reason, a number of instrumentation systems have been developed, especially in the field of steel plants. The legacy system had been developed by each developer and raised as to the degradation of system stability, the complexity of application maintenance and the usability of data. A great deal of effort has been made on standardization of process computer and component development. What seems to be lacking, however, is on quality instrumentation system.
2. SYSTEM STARDARDlZATION FOR QUALITY INSTRUMENTATION
2.1 Legacy Instrwnenation System 2Dl
This paper proposes the integrated system aimed at standardizing the system architecture and the design methods on quality instrumentation of coil rolling mill at steel works. In first, as developing instrumentation controller, we explain the HIW specification and standard I/O library to be used on various device driver and common components such
Fig . 1 Legacy Instrumentation System Configuration The legacy controller was the standalone system that includes the controller that could acquire and manipulate data and execute internal algorithm and 253
the HMI that could monitor and control sensors/actuators. Developer had built in one system that could acquire and monitor instrumentation data. The system had risen as to the degradation of system stability, the complexity of application maintenance and the usability of data.
and maintenance time decrease), as well as deployments. Greater efficiency is achieved through specialization, and instrumentation developer is free to focus on solving the business problem at hand.
3. SYSTEM STARDARDIZATION FOR For high quality production, it is necessary to retain and integrate data on product and instrumentation . It is also important to have the stability and reliability on the related system. The motives have combined to make a proposal on standardization and integration on quality instrumentation system.
QUALITY INSTRUMENTATION
3. J Instrumenatioll Controller Architecture Instrumentation controller receives data from sensor, controls actuators and communicates with the other systems in the plant. The controller should be mainly differentiated into embedded system, pc based system, VME system. We decided to use pc based system in developing the controller. We proved the fact that pc based system met the requirements for the stability and the performance through the adaptation of the system . PC based system has many advantages that we can have an abundant spare part and lean how to use and repair the system . We can various kinds of connect the system easily with sensors and devices.
2.2 Standard Instrumentation System
lnS"Jul':"I~:tatJ :o~~o;lt!
Fig. 2. Standard Configuration
Instrumentation
System
2.2 Standard Instrumentation System The improvement of standard instrumentation system has shown the importance of three features. First, we could divide the system into in strumentation controller and Quality Instrumentation Data Integrated System. The controller can acquire the data from sensors, control actuators and monitor the data. QIDIS can be used to store, access and analyse the data to be acquired by the controller. This approach could improve performance and stapility of the systems and make an upgrade of the system easy.
Application Program
Std . A
Std D
110
1/0
;-
In second, we standardized specifications of the internal device and the systems . We let the developer access use device driver of board through our standard library. We made common components such as communication module, data access module that can use commonly in the instrumentation controllers. It made an upgrade and modification of application easy. It could also reduce time and money that needs to develop the system end improved the stability of the system using the verified components. The developer could focus on the development of the algorithm on instrumentation.
~
-I
A I/ O
D I/ O
nr i v~r
nrivpr
Analog I/ O
Degital I/ O
'-
11
Std Spnc;nr
Sensor
Std ETC
-
110
ETC
nr i \lpr
nri\l".r
Sensor I/ O
ETC Board
Fig. 3. Stand 10 interface Layer We had implemented application program directly in legacy system using by the device driver to be supported by the device maker. When these codes including device handling functions were modified or changed, they spread so wide that we could not change easily with new device handling modules. In new standard system, we proposed standard 10 API in interface layer. We recommended the application program should be coded using the standard API. It makes it possible not to be seriously affected by changing devices . Whenever we introduce new device, it is enough to develop new standard 10 interface module for the device.
In last, QIDIS can be broken down into three layers data store layer, application layer, presentation layer. The modern approach to serverside Internet computing partitions the task of constructing a deployment into separate roles. Rather than build an entire deployment from scratch, a corporation can purchase a server-side platform from a middleware vendor, re-usable components from an application vendor, and development and management tools from a tools vendor. This results in a speedier time to user satisfaction (development
3.3 Common Components Instrumentation controller executes the data transportation with each devices and systems and processes the received data, stores the results of the algorithm or sends them to the related systems. In development of the legacy system, our developers 254
had cooed not only the program related to instrumentation algorithm, but also the communication programs and various kinds of interface programs. These techniques that used to program the interfaces were so vary and difficult that it had invoked many problems such system down and communication breakdown. It was very serious in the plant automation. The new corner could not under stand and update rapidly the program.
• •
We proposed the second approach that uses GUI programming tool and graphic handling component. Instrumentation controller application. The elements of HMI in instrumentation application consist of mainly the data display items and the trend graphs to display micro data in real-time. We could develop the application using the development tool such as visual studio .NET except for trend graph display. The graphic display component provided developers advanced charting features with minimal integration efforts. This feature allowed developers to focus their attention on the functional aspect of the application rather than visuals or other platform specific issues.
~~
~~ Instrumentation Algorltiun
lJ B
I/F
Da ta Processi ng
I
' -_ _ _--' 1 Device I/F I 1 Comm
Using GUI programming tool and commercial component Using GUI programming tool
4. QUALITY INSTRUMENTATION DATA INTEGRATED SYSTEM (QIDIS)
I
Fig . 4. Common Components
4.1 System Architecture
The common modules such as Database handling, Communication handling, device handling was used commonly in many instrumentation applications. Whenever the application is built, it is unnecessary to develop these newly common modules.
Operators in our plants want to inquire directly of the data for quality information and know from the informaJion whether there are problems in products. They also want to access the information at all areas in the plant ( operator room , office ) and improve product quality using various analytic tools such as statistics, OLAP and data mining. The server stores all data that received from controllers and the related control systems. The server is connected to intranet so that many users can access data and monitor graph through web browser in all plant area.
Developers that have much expericnce in system program extracted common module from instrumentation application and built common components to be reused in many application . It was robust component that had tested in many applications and was stable. As using common components, we could focus on developing in new instrumentation algorithm and could reduce time and cost to be needed .
3.4 Humall Mall Intelface
The most important feature of HMI in the controller is the real-time data display and real-time control such as emergency stop function. We restricted the function of the HMI system that it could monitor the real-time data and display real-time trend graphs and control the importance device setting . The access and analysis of the previous instrumentation data was supported through QIDIS. It made possible the realtime monitoring to be stabilized and the structure of the application to be simplified. We detached the algorithmic cOOes as separated files from the HMI application so that the application can be easily understoOO and mOOified.
Fig. 6. System Architecture of QIDIS.
There are three kinds of methods that develop a HMI application in PC based system. • Using commercial package such as intouch or citect SCADA tools.
The basic architecture of QIDIS is depicted in Fig. 6. QIDIS consists of DB server, application server and web server but we have constructed in a single server because we are in the test stage of prototype system.
255
It has modul e structure to be expanded efficiently for a later time . It may be divided into tlu'ee layers presentation layer, applicati on logic layer, data store. •
Fi g.7 .Data Storing Flow in QIDIS Whcnever tlle processing of a coil is completed, web server receives micro data from in strumentation control ler. The data file is stored in buffer tllrough socket of TCP/lP. The server calculates tlle representative values fr om tlle data and stores it to micro data summary table. The micro data files are classified by date and named witll coil id. They are mainly used in trend graph di splay.
Presentation layer: Thi s layer is responsibl e for the presentation of data, receiving user events and controlling tlle user interface . Users can access the server tl1Tough we browser in tlleir desktop. We developed web HMI using ASP.net and web form in .NET and established on lIS. As controller HMI to be explained in previous section, we also included the web form graphic di spl ay components.
•
Application Logic Layer: Thi s layer processes tlle data to be received and send result s to other layers. Thi s tier protects tlle data from direct access by the clients. It mainly includes inquiring tlle summary result, di spl ayi ng trend graphs; di splay tlle re,ult of stati stics analysis. It was implemented using C# on .NET framework and otller tools.
•
Data Store: Thi s layer is responsible for data storage. Bcsides tlle widespread relational database systems, exi stin g legacy system s databases are often reused here.
4.3 Web Server
In legacy instrumentati on controller, we had stored micro data to each controller's storage and let tllem be maintained by each system. As quality management gets more important and user nced to do complex data analysis such as stati stical anal ysis, user have wanted micro data to be integrated to one system and analysed in tllere. We have built up a web server tllat support quality analysis using simple product and data inquiry and trend data graph The functi ons of web server are as foll ow: •
4.2 Data Store
•
Data store consists of product setup data, product result data, micro summary data, and micro data fil es extra ck1.ta.
•
• •
•
• •
•
Product Setup Data Table: Record data such as coil id, target tllickness and widtll, speed is stored on per a coil. Product Results Table: Record data such as coil id, final average tllickness and widtll, average tempera ture is stored on per a coil. Micro Data Sununary Table: Micro Data Files: It stores micro data to be acquired by instrumentati on controller in time period rates. The files are created in text type such CSV format. Etc: tllis is used in server .for internal informati on handling.
•
QIDIS introducti on • System Information, configuration and functi on Basic Informati on Inquiry • Inquirin g into product setup and result information Trend Data Inquiry • Makin g graph of micro instrumentation data such as temperature, XRD. Trend Data Basic Statistics Inquiry • Makin g graph of micro instrumentation data and di splaying ba sic stati stic informati on. Summary Data Statistic Information Inquiry • Displaying daily/weekly/montllly summary data using control chart
4.4 Web based HMI
-----
4.3 Data Storillg Flow 0 1hl60Qf lllftdWftbSerw r
,' "
..
v,
.,.
,..
If.
-~--,-!-,
..
~, .
Fig.8 .Trend graph Display on Micro Data.
256
.
(
Trend graph on temperature micro data received from controller is depicted in Fig 8. First, the coil id was selected and the item to be shown among more 50 items was selected. User could detect abnormal temperature condition and compare the relation between graph data and product quality. User could also display graph with basic statistics information.
REFERENCES
Szyperski, Clemensl Gruntz, Dominikl Murer, Stephan( ) Component Software:Beyond ObjectOriented Programming, Addison- Wesley Esposito, Dino (2002) : Building Web Solutions With Asp.Net and Ado. Net, Microsoft Press
Fig.9. Control Chart Example
Control chart on temperature summary data is depicted in Fig 9. Making a control chart of summary data let us monitor abnormal data to be generated in some prexiuct easily. As investigating the cause of abnormal data in detail, we could understand the relation of production and the cause of abnormality. As comparing l,'Taphs In top and bottom simultaneously, user could investigate the relation of the variables in top and bottom and add statistical relational analysis.
5. Conclusion
This paper proposed the integrated system aimed at standardizing the system architecture and the design methods on quality instrumentation of coil rolling mill at steel works. We proposed the HIW specification, standard VO library and common components. An evaluation study showed the time of quality analysis in QIDIS system could be reduced and the information to be supported in the new system was more enough than legacy system . Based on the test result, we will extend the analytic functions and adapt the system to cold rolling mill in near future.
257
20-22 September 2006
Cracow, Poland
QUALITY INSTRUMENTATION DATA INTEGRATED STANDARD SYSTEM IN COLD ROLLING MILL
Yeongbok Park, Jaehwa Hong, Wonho Hwang, Kiwon Kim Instulllentation Research Group, Technial Lab. POSCO 1 goidong, namgu, pohang, Korea Tri [email protected] Authors' affiliations, Italics, upper and lower case, ce Ilf red, 127 mill (5 in) width (maximum)
Abstract: This paper proposed a standardization of instrumentation system aimed at standardizing the system architecture and the design methods on quality instrumentation of coil rolling mill at steel works. In first, as developing instrumentation controller, we proposed the WW speci tication and standard I/O library to be used on various device driver and common components such as communication and device handler. In second, we proposed Quality Instrumentation Data Integrated Server that the system can integrate the data about product and instrumentation and users can access and analyse information through web browser in all plan area. Copyright © 2005 IFAC Keywords : cold rolling mill, quality instrumentation, standardization, component, system integration
as communication and device handler. In second, we propose Quality Instrumentation Data Integrated Server that can integrate the data about product and instrumentation and access and analyse information through web browser in all plan area.
1. INTRODUCTION
In the major application of instrumentation and quality management on steel product, the architecture of the instrumentation system is an essential factor because it directly affects product quality work such as the range and time of quality analysis. For this reason, a number of instrumentation systems have been developed, especially in the field of steel plants. The legacy system had been developed by each developer and raised as to the degradation of system stability, the complexity of application maintenance and the usability of data. A great deal of effort has been made on standardization of process computer and component development. What seems to be lacking, however, is on quality instrumentation system.
2. SYSTEM STARDARDlZATION FOR QUALITY INSTRUMENTATION
2.1 Legacy Instrwnenation System 2Dl
This paper proposes the integrated system aimed at standardizing the system architecture and the design methods on quality instrumentation of coil rolling mill at steel works. In first, as developing instrumentation controller, we explain the HIW specification and standard I/O library to be used on various device driver and common components such
Fig . 1 Legacy Instrumentation System Configuration The legacy controller was the standalone system that includes the controller that could acquire and manipulate data and execute internal algorithm and 253
the HMI that could monitor and control sensors/actuators. Developer had built in one system that could acquire and monitor instrumentation data. The system had risen as to the degradation of system stability, the complexity of application maintenance and the usability of data.
and maintenance time decrease), as well as deployments. Greater efficiency is achieved through specialization, and instrumentation developer is free to focus on solving the business problem at hand.
3. SYSTEM STARDARDIZATION FOR For high quality production, it is necessary to retain and integrate data on product and instrumentation . It is also important to have the stability and reliability on the related system. The motives have combined to make a proposal on standardization and integration on quality instrumentation system.
QUALITY INSTRUMENTATION
3. J Instrumenatioll Controller Architecture Instrumentation controller receives data from sensor, controls actuators and communicates with the other systems in the plant. The controller should be mainly differentiated into embedded system, pc based system, VME system. We decided to use pc based system in developing the controller. We proved the fact that pc based system met the requirements for the stability and the performance through the adaptation of the system . PC based system has many advantages that we can have an abundant spare part and lean how to use and repair the system . We can various kinds of connect the system easily with sensors and devices.
2.2 Standard Instrumentation System
lnS"Jul':"I~:tatJ :o~~o;lt!
Fig. 2. Standard Configuration
Instrumentation
System
2.2 Standard Instrumentation System The improvement of standard instrumentation system has shown the importance of three features. First, we could divide the system into in strumentation controller and Quality Instrumentation Data Integrated System. The controller can acquire the data from sensors, control actuators and monitor the data. QIDIS can be used to store, access and analyse the data to be acquired by the controller. This approach could improve performance and stapility of the systems and make an upgrade of the system easy.
Application Program
Std . A
Std D
110
1/0
;-
In second, we standardized specifications of the internal device and the systems . We let the developer access use device driver of board through our standard library. We made common components such as communication module, data access module that can use commonly in the instrumentation controllers. It made an upgrade and modification of application easy. It could also reduce time and money that needs to develop the system end improved the stability of the system using the verified components. The developer could focus on the development of the algorithm on instrumentation.
~
-I
A I/ O
D I/ O
nr i v~r
nrivpr
Analog I/ O
Degital I/ O
'-
11
Std Spnc;nr
Sensor
Std ETC
-
110
ETC
nr i \lpr
nri\l".r
Sensor I/ O
ETC Board
Fig. 3. Stand 10 interface Layer We had implemented application program directly in legacy system using by the device driver to be supported by the device maker. When these codes including device handling functions were modified or changed, they spread so wide that we could not change easily with new device handling modules. In new standard system, we proposed standard 10 API in interface layer. We recommended the application program should be coded using the standard API. It makes it possible not to be seriously affected by changing devices . Whenever we introduce new device, it is enough to develop new standard 10 interface module for the device.
In last, QIDIS can be broken down into three layers data store layer, application layer, presentation layer. The modern approach to serverside Internet computing partitions the task of constructing a deployment into separate roles. Rather than build an entire deployment from scratch, a corporation can purchase a server-side platform from a middleware vendor, re-usable components from an application vendor, and development and management tools from a tools vendor. This results in a speedier time to user satisfaction (development
3.3 Common Components Instrumentation controller executes the data transportation with each devices and systems and processes the received data, stores the results of the algorithm or sends them to the related systems. In development of the legacy system, our developers 254
had cooed not only the program related to instrumentation algorithm, but also the communication programs and various kinds of interface programs. These techniques that used to program the interfaces were so vary and difficult that it had invoked many problems such system down and communication breakdown. It was very serious in the plant automation. The new corner could not under stand and update rapidly the program.
• •
We proposed the second approach that uses GUI programming tool and graphic handling component. Instrumentation controller application. The elements of HMI in instrumentation application consist of mainly the data display items and the trend graphs to display micro data in real-time. We could develop the application using the development tool such as visual studio .NET except for trend graph display. The graphic display component provided developers advanced charting features with minimal integration efforts. This feature allowed developers to focus their attention on the functional aspect of the application rather than visuals or other platform specific issues.
~~
~~ Instrumentation Algorltiun
lJ B
I/F
Da ta Processi ng
I
' -_ _ _--' 1 Device I/F I 1 Comm
Using GUI programming tool and commercial component Using GUI programming tool
4. QUALITY INSTRUMENTATION DATA INTEGRATED SYSTEM (QIDIS)
I
Fig . 4. Common Components
4.1 System Architecture
The common modules such as Database handling, Communication handling, device handling was used commonly in many instrumentation applications. Whenever the application is built, it is unnecessary to develop these newly common modules.
Operators in our plants want to inquire directly of the data for quality information and know from the informaJion whether there are problems in products. They also want to access the information at all areas in the plant ( operator room , office ) and improve product quality using various analytic tools such as statistics, OLAP and data mining. The server stores all data that received from controllers and the related control systems. The server is connected to intranet so that many users can access data and monitor graph through web browser in all plant area.
Developers that have much expericnce in system program extracted common module from instrumentation application and built common components to be reused in many application . It was robust component that had tested in many applications and was stable. As using common components, we could focus on developing in new instrumentation algorithm and could reduce time and cost to be needed .
3.4 Humall Mall Intelface
The most important feature of HMI in the controller is the real-time data display and real-time control such as emergency stop function. We restricted the function of the HMI system that it could monitor the real-time data and display real-time trend graphs and control the importance device setting . The access and analysis of the previous instrumentation data was supported through QIDIS. It made possible the realtime monitoring to be stabilized and the structure of the application to be simplified. We detached the algorithmic cOOes as separated files from the HMI application so that the application can be easily understoOO and mOOified.
Fig. 6. System Architecture of QIDIS.
There are three kinds of methods that develop a HMI application in PC based system. • Using commercial package such as intouch or citect SCADA tools.
The basic architecture of QIDIS is depicted in Fig. 6. QIDIS consists of DB server, application server and web server but we have constructed in a single server because we are in the test stage of prototype system.
255
It has modul e structure to be expanded efficiently for a later time . It may be divided into tlu'ee layers presentation layer, applicati on logic layer, data store. •
Fi g.7 .Data Storing Flow in QIDIS Whcnever tlle processing of a coil is completed, web server receives micro data from in strumentation control ler. The data file is stored in buffer tllrough socket of TCP/lP. The server calculates tlle representative values fr om tlle data and stores it to micro data summary table. The micro data files are classified by date and named witll coil id. They are mainly used in trend graph di splay.
Presentation layer: Thi s layer is responsibl e for the presentation of data, receiving user events and controlling tlle user interface . Users can access the server tl1Tough we browser in tlleir desktop. We developed web HMI using ASP.net and web form in .NET and established on lIS. As controller HMI to be explained in previous section, we also included the web form graphic di spl ay components.
•
Application Logic Layer: Thi s layer processes tlle data to be received and send result s to other layers. Thi s tier protects tlle data from direct access by the clients. It mainly includes inquiring tlle summary result, di spl ayi ng trend graphs; di splay tlle re,ult of stati stics analysis. It was implemented using C# on .NET framework and otller tools.
•
Data Store: Thi s layer is responsible for data storage. Bcsides tlle widespread relational database systems, exi stin g legacy system s databases are often reused here.
4.3 Web Server
In legacy instrumentati on controller, we had stored micro data to each controller's storage and let tllem be maintained by each system. As quality management gets more important and user nced to do complex data analysis such as stati stical anal ysis, user have wanted micro data to be integrated to one system and analysed in tllere. We have built up a web server tllat support quality analysis using simple product and data inquiry and trend data graph The functi ons of web server are as foll ow: •
4.2 Data Store
•
Data store consists of product setup data, product result data, micro summary data, and micro data fil es extra ck1.ta.
•
• •
•
• •
•
Product Setup Data Table: Record data such as coil id, target tllickness and widtll, speed is stored on per a coil. Product Results Table: Record data such as coil id, final average tllickness and widtll, average tempera ture is stored on per a coil. Micro Data Sununary Table: Micro Data Files: It stores micro data to be acquired by instrumentati on controller in time period rates. The files are created in text type such CSV format. Etc: tllis is used in server .for internal informati on handling.
•
QIDIS introducti on • System Information, configuration and functi on Basic Informati on Inquiry • Inquirin g into product setup and result information Trend Data Inquiry • Makin g graph of micro instrumentation data such as temperature, XRD. Trend Data Basic Statistics Inquiry • Makin g graph of micro instrumentation data and di splaying ba sic stati stic informati on. Summary Data Statistic Information Inquiry • Displaying daily/weekly/montllly summary data using control chart
4.4 Web based HMI
-----
4.3 Data Storillg Flow 0 1hl60Qf lllftdWftbSerw r
,' "
..
v,
.,.
,..
If.
-~--,-!-,
..
~, .
Fig.8 .Trend graph Display on Micro Data.
256
.
(
Trend graph on temperature micro data received from controller is depicted in Fig 8. First, the coil id was selected and the item to be shown among more 50 items was selected. User could detect abnormal temperature condition and compare the relation between graph data and product quality. User could also display graph with basic statistics information.
REFERENCES
Szyperski, Clemensl Gruntz, Dominikl Murer, Stephan( ) Component Software:Beyond ObjectOriented Programming, Addison- Wesley Esposito, Dino (2002) : Building Web Solutions With Asp.Net and Ado. Net, Microsoft Press
Fig.9. Control Chart Example
Control chart on temperature summary data is depicted in Fig 9. Making a control chart of summary data let us monitor abnormal data to be generated in some prexiuct easily. As investigating the cause of abnormal data in detail, we could understand the relation of production and the cause of abnormality. As comparing l,'Taphs In top and bottom simultaneously, user could investigate the relation of the variables in top and bottom and add statistical relational analysis.
5. Conclusion
This paper proposed the integrated system aimed at standardizing the system architecture and the design methods on quality instrumentation of coil rolling mill at steel works. We proposed the HIW specification, standard VO library and common components. An evaluation study showed the time of quality analysis in QIDIS system could be reduced and the information to be supported in the new system was more enough than legacy system . Based on the test result, we will extend the analytic functions and adapt the system to cold rolling mill in near future.
257