A Web-based control of a real laboratory for process engineering

A Web-based control of a real laboratory for
process engineering education
Maher Chaabene *, Kamel Mkaouar, Rachid Souissi
Applied Computer Science Research Group: The high institute of technological studies - Sfax, Tunisia.
*
Corresponding author: Address: Rte Mahdia Km 2, BP88A, ElBustan 3099- Sfax, Tunisia.
Tel.: +216 74 431 425; fax: +216 74 431 386
E-mail address: [email protected]
Abstract
Our time is branded by the orientation towards e-learning in several education fields. Yet, the
remote teaching means satisfying the needs of specific engineering studies remain limited to
simulated applications or virtual laboratories which involve a reduced assimilation of the taught
material. This paper introduces a real laboratory control based on a Web embedded system and
an interactive Web application. The laboratory is designed for process engineering education. In
order to conceive a Web-based process management for learning features, five essential design
issues have been investigated: requirement specification, architecture selection, system
implementation, design of a Web-based human-computer interface, and an access control system
for the interactive learning environment and work validation. A didactic lift is used as an
engineering educational process to demonstrate our design methodology. A set of software
modules is embedded in the local control system in order to be shared by multiple
communicating users. Time delay due to Internet traffic has been overcome by using miniature
Web server dedicated to the Web-based laboratory supervision and control. The experimental
results have shown that the Internet-based real laboratory offers similar behaviour to a local
laboratory.
Keywords: Embedded systems; Internet Supervision and control; e-learning; web-server.
1. Introduction
These last years, impressive advances in Internet have made this network the most successful.
Standards, technologies and Web-implementations do not cease improving and diversifying.
Nowadays, the use of a Web navigator for information exchange has become well-known by the
large public. The idea consists of exploiting this infrastructure with its standards of
communication to supervise and control educational laboratory instruments and features over the
Web. Thus, it will be possible to reach these distant apparatuses by using a standard Web
navigator.
Most researches in this area focus on developing virtual features for distant learning purposes
including Web-based virtual control laboratories (Aklan 1996, Shin 1998 and Dongil 2002) and
Web-based learning resources libraries (Hough 2000, Jackson 2000, Chiaming 2003). However,
few publications present real Web-control laboratory. The encountered difficulties reside in
limitation, caused by Internet environment features, such as the shortage of Web-control devices,
Internet time delay, multiple users’ collaboration and Web-related safety (Yang 2003).
Other research fields can give supports to our theme and can be directed into the same objective.
Researches interested in online acquisition and control systems as well as in Web-embedded
systems offer many design issues and implementation for internet-based process control (Shor
1998, Overstreet 1999, KO 2001, Han 2001, Shengwei 2002 and Mahalik 2005). These
researches have essentially proposed solutions to ensure real-time control and safety to systems
managed over the Web.
This paper aims to develop a new methodology for Web-based real laboratory designed to teach
interactively engineering materials. In a first stage we start by connecting a mini-lift to the Web
as a test bed in order to demonstrate the effectiveness of our designed system. The proposed
methodology is based on the control of the mini-lift by a Program Logic Controller (PLC)
connected to a miniature Web Embedded Server (WES). Since this later is configured to manage
only our educational process, the above quoted problems have to be overcome.
The various sections of this paper are organised as follows. Section 2 fixes requirement
specification of the Internet-based laboratory. The architecture discussion and selection of the
Web-based supervision and control system are investigated in section 3. Section 4 shows the
system implementation. The Web user interface and system validation design are described in
section 5. The access control system is presented in section 6. Section 7 provides perspectives
and conclusion.
2. Requirement specification
The first task to be accomplished when designing a control system is specifying the requirements.
This should determine the right control structure. In our case, the requirements should include
process monitoring and control objectives that are entirely achievable through the Internet control
level. Thus, the conflicting tradeoffs between goals and constraints of the system have to be
identified and resolved.
2.1 Requirement analysis
Several specification methods are used in research work: System Analysis Design Technique
(SADT), Unified Modelling Language (UML), and Structured Analysis for Real Time (SART),
etc. These methods depend on the nature and constraints of the system to develop. As we
consider the aimed system has a Web-related traffic delay constraint, we adopt the SART method
in order to take into account the real time behaviour. This method consists of three stages. First, a
context schema analyses the fundamental and the secondary system tasks to be accomplished and
their interactions (fig.1). In a second stage, a preliminary schema gives details for each task so as
to define all system components with their interactivity and behaviour (fig.2). Finally the
ordering schema (fig. 3) shows all lift orders details. Table 1 gives the data dictionary that
contains the whole system data base definition.
Orde
rs
ts
Even
Fig. 2: The preliminary transformation schema.
rs
de
or
E
D nab
isa le
bl /
e
Ev
ent
s
rs
C
p le y c le ’
me s
n ta
tio
Or
de
im
n
ft
Li
/
n n
tio io
ec ct
nn nne
Co sco
i
D
’s
cle
t cy m
Lif rogra
p
A
Cr cco
ea un
tio t
n
Lift state
g
in
ok
Bo
Fig. 1: The context schema.
DC1 / DC2 / DC3 / DC4
LF1 / LF2 / LF3 / LF4
RF2 / RF3 / RF4
DO1 / DO2 / DO3 / DO4
Cycle’s program *
LST
SON1 / SON2 / SON3 / SON4 / SONL
CYPR
Present floor
Raise lift
Swich on
lamp
2.5
Display lift
information
er
Trigg
2.6
r
Control lift
order
Closed door * Opened door *
er
Trigg
Trigger
2.1
er
gg
Tr i
2.8
er
gg
Opened
door signal
Tr i
Door state
GDLF1 / GDLF2 / GDLF3
Trigger
2.7
Lamp state
2.3
SOF1 / SOF2 / SOF3 / SOF4 / SOFL
Asked floor *
Get lift
down
2.2
Get cycle‘s
implementation
2.9
Cycle’s program
Opened door
Swich off
lamp
Disconnection
Connection
RLF2 / RLF3 / RLF4
Closed door
signal
Tick
Trigger
ge
ig
Tr
Present floor *
2.4
Asked floor
Cycle‘s
implementation
GDF1 / GDF2 / GDF3
Lift cycle’s
program
Fig. 3: The Order lift diagram.
2.2 Discussion on control constraints
The main constraints of a Web-control system are the uncertain Internet time-delay, the multiuser access and conflict resolution, and the system safety cover.
The multi-user access and the system safety constraints have been largely discussed through
bibliography. Multiple solutions have been suggested in order to overcome these constraints.
These solutions are essentially based on software applications ensuring users’ priority
management and system safety checking. As for time delay constraint, it is still under study. In
what follows we discuss this constraint then we give our adopted approach.
Data flow between Web-based process and Internet clients fixes the efficiency of supervision
and control operations. One of the difficulties in Internet-based process control is the dynamic
delay caused by the Internet traffic. Cycles’ times to a process control operation over the Internet
are shown by figure 4.
Fig. 4: Web control system constraints.
The total time needed to perform a Web process control operation (Luo 2000 and Yang 2003) is:
T=t1+t2+t3+t4, where:
t1: time delay in making control decision by a remote operator.
t2: time delay in transmitting a control command from the remote operator to the local system.
t3: execution time of the local system to perform the control action.
t4: time delay in transmitting the information from the local system to the remote operator.
Many approaches have been used in order to minimize the amount T by reducing its time delay
terms t1, t2, t3 and t4.
First, since the amount t1 depends only on the operator reaction, it cannot be minimized.
Second, the Internet time delays t2 and t4 increase with distance and depend essentially on the
number of nodes traversed. Han 2001 demonstrates that the Internet time delays td (k) (t2 or t4) at
instant k can be calculated as follows:
n
n
l
M n l
M
t d (k ) = ∑  i + t iR + tiL (k ) +  =∑ ( i + tiR + ) + ∑ t iL (k ) = d N + d L (k ).
bi  i=0 C
bi
i =0  C
i =0
Where li is the ith length of link, C the speed of light, tiR the routing speed of the ith node, tiL(k)
the delay caused by the ith node's load, M the amount of data, and bi the bandwidth of the ith link.
The term dN depends on network performances and can only be minimized by improving the
Internet load: the processing speed of nodes, the connection bandwidth type, the transmission
speed, etc. As for dL(k), it is a time-dependent term which can be reduced by choosing a control
architecture which is insensitive to the time delay.
Finally, reducing the time delay t3 requires an increase of the local system speed to perform an
action. Thus, the most useful approach consists in using a virtual control strategy. Although it is
the best solution since it is not sensitive to the time delay t3, this method presents the handicap of
virtual applications and cannot convey real information between the remote control operator and
the local control system.
2.3 Our approach
The approach consists in configuring a Web-Embedded Server (WES) connected to a
Programmable Logic Controller (PLC). The PLC ensures the process supervision and control. At
the same time it communicates with the WES by receiving instructions and sending reports. As
for the WES, it communicates through the Web with distant users. This approach must offer a
dialogue with a real process.
Since it deals only with the network management and a limited number of users, the WES
performs a much reduced task compared with that of usual servers. As for the PLC, it is
completely dedicated to the process supervision and control. Consequently, this strategy
generates a remarkable reduction in access time (t3) to the process.
The remote control operator sends his requests (process control) to the WES which answers
regularly (process acquisition) thanks to a local program pointing PLC data base. This task is
accomplished by using the servlet and socket concepts (network library allowing interface
between the programs and the network layer) available in the JAVA Script programming
language. Since it is an object oriented programming, this language offers several functionalities
so as to generate applications of great flexibility.
3. Architecture selection
Two different architectures are often used in process control via Internet.
In the first architecture (figure 5) the process is directly connected to the WES via the PLC. As
the process is very close to the WES, this involves a significant decrease of the access time t3.
Nevertheless this architecture might cause an additional load to WES because this later has to
manage the network communication as well as the dialogue with PLC.
Fig. 5: The process is directly connected to the WES
As for the second architecture, (figure 6) the process is connected via an application terminal and
a PLC to Internet network. The communication between the WES and the process is guaranteed
by the Web. This architecture offers a portable application since the process does not need to be
near the WES. In addition, several processes in different places could be managed by the same
WES. Yet, a permanent connection between the application terminal and the WES is required.
Moreover, the time delay t3 increases considerably compared to the first architecture because it
becomes equal to (t’3 + t”3).
As our principal objective is the minimization of access time t3, we have selected the first
architecture in order to develop our application.
Fig. 6: The process is connected to the WES via an application terminal
4. System implementation
The WES is an e-Device that permits to manage and control remotely any equipment from a
standard Internet navigator (Internet Explorer, Netscape, Mozilla, etc.). Thus, an Internet site
allows full management of the equipment simply by using a terminal connected to Internet or
Ethernet. The WES solution integrates all necessary components to transform remotely a noncommunicating standard machine into controllable equipment from a standard Web navigator.
Equipped with a supervision and control program, the PLC is connected to a four stage mini-lift
through its input/output lines. This later is protected against operators’ wrong movements by
hardware system. The data convey between the PLC and the WES through a serial connection
RS232C using MEWTOCOL protocol (protocol of Matsushita). As for the WES, it ensures the
communication with users through an Ethernet (10 base T) connection. Figure 7 gives the system
implementation details.
Fig. 7: System implementation
5. Access control system
The access control system supports two types of users:
Ordinary users who are subscribers having login and password. They are authorised to
benefit only from supervision services (visualisation of the lift state).
A privileged user who is an ordinary unique user possessing access rights that allow him to
handle direct orders on the lift as well as to implement a cycle program in PLC. The
program commands can be entered in the form of: Ladder, schema of functional context,
list of instructions, functional block diagram or structured text. This type of user is
identified by an access code available during a fixed time period and obtained after a
preliminary booking.
Figure 8 gives flowchart details of the users’ access management. As for figure 9, it shows an
illustration of the subscription and the booking Web page.
Super user
Web connection
Privileged user
Web connection
Link to subscription page
Link to booking page
Identification
Ordinary user
Web connection
Privilege
code and
password
Subscription
code and
password
Priority
management
Users’ access
management and
account creation
Straightforward
commands
Cycle’s
implementation
Fig. 8: flowchart details of users’ access.
Supervision of
lift status
Fig. 9: Links to subscription and booking.
6. Web user interface and system validation
This phase consists to establish a human-computer interactive Web page in HTML format that
ensures the software system configuration (WES and PLC), the communication with users and
the system access control. The work validation is conducted on the local network of the high
institute of technological studies - Sfax, Tunisia (ISET).
6.1 WES configuration
The WES, which is an advanced system, was acquired from NAIS Control Company. Its
configuration requires appropriate software called FP Web-Server Configurator. The main
menu parameters are fixed so as to minimize the time delay t3 and to guarantee a high access
control to the process (figure 10):
The WES manages only one PLC.
Information is returned to users by EMAIL.
The WES access is controlled by password and is reserved to subscribed users.
The WES accepts three subscribed users at the same time. Only the privileged user is
authorized to handle the PLC.
Fig. 10: WES configuration
6.2 Home page design
The remote control home page is designed so as to offer maximum flexibility, clarity and
effectiveness to the user. It puts forward three control icons: supervision, control orders and
cycle programming. Nevertheless, some rights are reserved for safety checking. Page
animation uses many multimedia tools that provide information in textual, vocal and graphic
forms.
An APACHE server is installed to provide, through supervision icons, home page images and
all information about lift state: stage number, emergency key, alarm, door sensor, etc. In
addition, control icons allows access, via PLC, to lift straightforward commands such as light
switching of the lift box, alarm sound and stage reaching. Finally, cycle programming icons
allows conception, adjustment, implementation and testing of an operating cycle algorithm by
means of FP Win Pro software.
Figure 11 shows the remote control home page: icons and design.
Fig. 11: Remote control home page
6.3 system validation
The current prototype system is implemented at ISET of Sfax which counts about 3000 students.
More than 30% of them are asked to handle the same work on the mini-lift. In the traditional
teaching method, each student first elaborates a flowchart allowing the mini-lift function in
different control modes by applying standard methods and tools (GRAFCET, Petri Network,
etc.). Second, the established diagram must be programmed on a P.C. micro-computer connected
to the mini-lift through the serial port, then tested by verifying the cycle’s stages. This lab course
is included in the studying plan of all profiles as process control and automation materials. The
teaching method has involved time access barrier when using the system (mini-lift) and a non
homogenous student’s knowledge evaluation.
By adopting the new teaching method, many facilities and favours have been offered to the
student.
• Browser applets provide interactive demonstrations on the operation manner of the system.
• Preliminary tasks (system study, simulation, functioning diagram implementation, etc.) are
developed on line without system occupation.
• Students’ access to the system for testing their work during all times after booking.
• Work evaluation is ensured on line by a commission of experts and returned to students by
E-mail.
Since our main objective is the technological improvement of the lab control, we have chosen to
apply the habitual learning content and evaluation methodology which presented no failure. The
effectiveness of the proposed system appears essentially in the optimal time sharing of the
system, the time free given to students to handle the work and the judicious evaluation of
students’ applications.
7. Conclusion
The teaching of engineering matters with practical aspect involves problems essentially in
features’ availability. In fact the student should apply theoretical studied concepts by leading
experiments in laboratory equipped with real processes. But several preventions can be met: the
features’ cost, the required large space to install laboratory, the maintenance and the profitability
of the devices, etc. Consequently, many educational institutes were directed towards remote
teaching. The generally followed strategy consists in designing virtual laboratories so as to
overcome all problems arising from processes supervision and control throughout the Internet
network: the process access time, the process safety, and the users’ conflicts resolution. However,
by adopting this procedure, the student can in no way lead experiments on real processes.
Our approach comes to propose the supervision and the control of a Web-based real laboratory.
In order to minimise the process access time obstacle when performing the online laboratory
operations, a Web Embedded Server is installed and connected to a PLC using an adequate
architecture for the system management.
The proposed system has been conceived on the basis of real time specification method which
makes the lab accessible over the Web at any time and from any place. We have designed and
implemented a Web interactive page that offers a satisfactory access control and guarantees
users’ conflict resolution. Users are classified into two distinct categories: privileged user and
ordinary users. Each category has its own access rights to the controlled system.
The actual work was carried out on a didactic lift and is being extended in the aim to connect
other features to the conceived lab: acquisition data system, robot, etc. The lab system is expected
to contribute to increasing students’ adaptability to working in real process plants via Internet
simply by using a Web browser.
Table and figures legends
Figure 1
The context schema.
Figure 2
The preliminary transformation schema.
Figure 3
The order lift diagram.
Figure 4
Web control system constraints.
Figure 5
The process is directly connected to the WES
Figure 6
The process is connected to the WES via an application
terminal
Figure 7
System implementation
Figure 8
Flowchart details of users’ access.
Figure 9
Links to subscription and booking
Figure 10
WES configuration
Figure 11
. Remote control home page
Table 1
Data dictionary
Trigger
Account creation
Booking
Enable
Disable
Events
DO1 / DO2 / DO3 / DO4
DC1 / DC2 / DC3 / DC4
Lift orders
GDLF1 / GDLF2 / GDLF3
RLF2 / RLF3 / RLF4
Lift cycle’s program
Cycle’s implementation
Connection
Disconnection
Orders
RF2 / RF3 / RF4
GDF1 / GDF2 / GDF3
SON1 / SON2 / SON3 / SON4
SONL
SOF1 / SOF2 / SOF3 / SOF4
Lift state
LF1 / LF2 / LF3 / LF4
LST
CYPR
Present floor
Asked floor
Lamp State
Cycle’s Program
Door state
Closed door
Opened door
Tick
= * To activate a discrete process *
= * To make an account for each user *
= * To reserve the lab for a time period *
= * To activate a periodical process *
= * To deactivate a periodical process *
= * To get some information that describe the state of the lift*
[DO1 | DO2 | DO3 | DO4 | DC1 | DC2 | DC3 | DC4]
= * The door of the 1st / 2nd / 3rd / 4th floor is opened *
= * The door of the 1st / 2nd / 3rd / 4th floor is closed *
= * Give the lift the different orders to take down and to bring up *
[GDLF1 | GDLF2 | GDLF3 | RLF2 | RLF3 | RLF4]
= * Get down the lift to the 1st / 2nd / 3rd floor *
= * Raise the lift to the 2nd / 3rd / 4th floor *
= * The user sends a program to the management system in order
to control the lift *
= * The management system implements the lift cycle’s program
in the PLC *
= * A user requests a connection to the system *
= * A user leaves the lift *
= * To give commands to the lift *
[RF2 | RF3 |RF4 | GDF1 | GDF2 | GDF3 | SON1 | SON2 | SON3 |
SON4 | SONL | SOF1 | SOF2 | SOF3 | SOF4]
= * The system sends the order to the lift to raise to the 2nd / 3rd / 4th
*
= * The system sends the order to the lift to get down the to the 1st
/ 2nd / 3rd floor *
= * To switch on the lamp of the 1st / 2nd / 3rd / 4th floor *
= * To switch on the lamp of the lift *
= * To switch off the lamp of the 1st / 2nd / 3rd / 4th floor *
= * To display the hole information of the lift *
[LF1 | LF2 | LF3 | LF4 | LST | CYPR]
= * To display the state of the lamp of the 1st / 2nd / 3rd / 4th floor *
= * The lift state *
= * To display the cycle’s program *
= * To memorise the actual floor *
= * To memorise the asked floor *
= * To memorise the lamp state *
= * To memorise the cycle’s program *
= * To memorise the door state *
= * To memorise the door closed *
= * To memorise the door opened *
= * The clock tick *
Table 1: Data dictionary
References
Aktan, B., Bohus, C. A., Crowl, L. A., & Shor, M. H. (1996). Distance learning applied to
control engineering laboratories. IEEE Transactions on Education, 39(3), 320–326.
Atherton, R. (1998). Java object technology can be next process control wave. Control
Engineering, 45(13), 81–85.
Baresi, L., Garzotto, F., Mainetti, L. and Paolini, P. Meta-modelling Techniques Meet Web
Application Design Tools. Lecture Notes in Computer Science 2306, Springer-Verlag, Berlin,
2002.
Chiaming Yen and Wu-Jeng Li (2003). Web-based learning and instruction support system for
pneumatics. Computers & Education Volume 41, Issue 2, Pages 107-120
David J. Hutches and Douglas Dankel, II SADT -- An Overview and Its Relationship to the
Intelligent Programming Environment, SERC-TR-22-F, 1988
Dongil Shin, EN Sup Yoon, Kyung Yong Lee, Euy Soo Lee (2002). A web-based interactive
virtual laboratory system for unit operations and process systems engineering education: issues,
design and implementation. Computers & Chemical Engineering, 26, 319-330.
Halley, A., & Gauld, P. (1999). Integration where Java meets process control. Control and
Instrumentation, 31(4), 57–58.
Han, K. H., Kim, S., Kim, Y. J., & Kim, J. H. (2001). Internet control architecture for Internetbased personal robot. Autonomous Robots, 10, 135–147.
Hatley, D.and Pirbhai, I. Strategies for Real-Time System Specification. Dorset Horse
Publishing, New York, 1988.
Hori, S., & Shimizu, Y. (1999). Designing methods of human interface for supervisory control
systems. Control Engineering Practice, 7,1413–1419.
Hough, M., & Marlin, T. (2000)..Web-based interactive learning modules for process control.
Computers and Chemical Engineering,24, 1485 – 1490.
http://62.180.233.51/
http://tbox.biz
http://www.edevice.com
http://www.iosoft.co.uk
http://www.nais-e.com/plc/plc_support/index.html
http://www.spidercontrol.net
Ko, C. C., Chen, B. M., Chen, J., Zhuang, Y., & Tan, K. C. (2001). Development of a webbased laboratory for control experiments on a coupled tank apparatus. IEEE Transactions on
Education, 44(1),76–86.
Levson, N. G., Heimdahl, M. P. E., Hildreth, H., & Reese, J. D. (1994). Requirements
specification for process-control systems. IEEE Transaction on Software Engineering, 20(9),
684–707.
Luo, R.C., & Chen, T.M. (2000). Development of a multi-behaviour based mobile robot for
remote supervisory control through the Internet. IEEE Transaction on Software Engineering,
20(9), 684-707.
M. Lissandre, Maîtriser SADT, Armand-Colin, 1990.
Michel Lai, Penser objet avec UML et Java, Dunod, 1998.
N.P. Mahalik, J. Ryu, B.H. Ahn (2005). Simulation integrated management layer for real-time
embedded DCN. Computer standards and interfaces (Article in press).
Overstreet, J. W., & Tzes, A. (1999). An Internet-based real-time control engineering
laboratory. IEEE Control Systems Magazine, 19(5), 320–326.
Pascal Roques, Frank Vallée. UML en action de l’analyse des besoins à la conception en
JAVA. EYROLLES 2003.
Paul T. Ward, Stephen J. Mellor. Structured Development for Real-Time Systems: Essential
Modeling Techniques. Prentice Hall 1986.
S.H. Yang, X. Chen, J.L. Alty (2003). Design issues and implementation of internet-based
process control systems. Control engineering practice, 11, 709-720.
Shaheen, M., Loparo, K. A., & Buchner, M. R. (1998). Remote laboratory experimentation.
American control conference, Philadel- phia, PA (pp. 1326–1329).
Shengwei Wang, Junlong Xie (2002). Integrating building management system and facilities
management on the Internet. Automation in construction, 11, 707-715.
Sheryl Duggins, SADT -- An Overview of the Methodology, SERC-TR-21-F,1988.
Shin, D., Lee, T. J., Kang, T., Park, S., & Lee, E. S. (1998). Virtual laboratories on the Internet
for unit operations experiments. Proceedings of KIChE Fall Meeting, 4 (2), 2321.
Shor, M., & Bhandari, A. (1998). Access to an instructional control laboratory experiment
through the World Wide Web. Proceedings of the American control conference, Philadelphia,
PA (pp. 1319–1325).
Ward, P.and Mellor, S. Structured Development for Real-Time Systems. Addison Wesley, New
Jersey, 1985.
Westhead, M. (1999). Use of Web and Internet technology in teaching and learning. Edinburgh
Parallel Computing Centre Report, Ver- sion 0.2, University of Edinburgh.
Yang, S. H., & Alty, J. L. (2002). Development of a distributed simulator for control
experiments through the Internet. Future Generation Computer Systems, 18(5), 595–611.