Docstoc

ENTER TITLE HERE _14 PT TYPE SIZE_ UPPERCASED_ BOLD AND CENTERED

Document Sample
ENTER TITLE HERE _14 PT TYPE SIZE_ UPPERCASED_ BOLD AND CENTERED Powered By Docstoc
					    AN APPLICATION OF REMOTELY CONTROLLED EXPERIMENTS TO
    PERFORM FEEDBACK-DAMPING CONTROL OF A VIBRATING BEAM

                                          M.C.Levesley1, P.Culmer1 P.Cripton2
                            1
                            School of Mechanical Engineering, University of Leeds, Leeds, U.K.
               2
                 Department of Mechanical Engineering, University of British Columbia, Vancouver, Canada,
                                               m.c.levesley@leeds.ac.uk


Abstract                                                        of a group of 20-30 students, working in pairs or small
This paper presents a system developed at the University        groups within a laboratory environment. To ensure each
of Leeds in the UK which allows remote access to real           of the students in a group has adequate access to facilities,
experimental equipment (ReLOAD). The system, initially          multiple copies of experimental equipment are often
developed in 2000 has been in use since 2001 to provide         required. The need for multiple copies of equipment
improved access to experimental equipment used as part          inevitably places limits on the quality of the equipment
of an undergraduate engineering degree programme.               that can be deployed. Even with group sizes of 20-30
Recently improvements which allow very simple                   students, laboratory sessions often need be repeated many
expansion have prompted an increase in the number of            times in a given year to ensure all students on a particular
remote experiments now in use. Most recently, the system        programme are exposed to the session. Even with multiple
has been used as an integral part of a post graduate            copies of equipment and duplicate laboratory sessions,
distance delivered module in Dynamic Simulation,                students can find that access to laboratory equipment is
offered at the University of Leeds. Key to its operation is     limited. For example, they cannot be easily repeated by
the ability of the system to serve video clips that record      students who may wish to collect more data, after
the experiment in progress, providing essential                 reflecting on their initial results, or may have been ill at
visualisation plus an improved sense of realism for the         the time the laboratory session was being run. They can
user. Both student and system evaluation has proved             also be seen as being inefficient, since although expensive
positive, with student feedback being particularly positive     equipment is used intensively for a period, it may remain
regarding the improvement in access to real data it offers.     unused for most of the year.
The system is shortly to be used to deliver experiments
hosted at Leeds to over 200 3rd and 4th year mechanical                  LAN
engineering undergraduates at the University of British                                        Internet
Columbia (UBC), Vancouver. The system shows huge
potential to allow more efficient and collaborative use of
laboratory equipment across the global engineering                               ReLOAD                   Client Computer
education community.                                                            Web Server


Key Words                                                                                    Hard Wired
                                                                                             Connection
E-Learning, Remotely Operated Experiments


                                                                                                          Equipment 1.1
1. Introduction                                                                 Experiment
                                                                                 Server 1

Engineering is by its nature a very practical subject and
extensive use is made of laboratory based teaching                                                        Equipment 1.2
sessions, particularly in the first few years of a typical
undergraduate degree programme. These sessions are
ideally suited to develop both cognitive skills (knowledge,
information analysis etc.) and psychomotor skills                               Experiment
(physical skills, such as the use of tools and equipment).                       Server 2                 Equipment 2.1
In addition, well designed laboratory sessions can
enthuse, motivate and inspire students, as well as having
the potential to allow them to work at their own pace and                                                  Equipment 2.2
take ownership of their learning material. In the UK, a
typical face-to-face session may last 2-3 hours and consist       Figure 1 ReLOAD Remote Experiment Structure
 In 2000, to overcome some of these limitations, a system       3. The ReLOAD Web Server
called ReLOAD (Real Labs Operated at Distance) was
developed within the School of Mechanical Engineering           The ReLOAD web sever is set-up to receive and process
at the University of Leeds, which allows access to              requests from the client, and since the web server is a
experimental equipment via a web interface. Initially it        single computer, custom software can be readily created
was used by level 2 undergraduates studying Vibration           and installed to monitor internet requests. The client
and Control using the school’s intranet but was soon            computer may be one of hundreds of different PC's and
adapted for more widespread access via the internet.            the manner in which data is transmitted and received from
Although the system cannot be used to develop                   the client computer has significant impact on the overall
psychomotor skills, it does allow students to develop           functionality and versatility of the final system. Several
higher cognitive skills such as planning experiments,           alternative methods have been proposed. An Active-X
collecting and analysing data, and conducting further           controls panel can be installed on the client computer to
experiments if required. It offers greater accessibility, can   send and receive information [2] through an Active-X
be run 24 hours a day, 365 days a year and allows many          enabled browser. Writing an Active-X control gives
students to access a single piece of high quality               maximum flexibility to place and position controls to give
equipment. More recently the ReLOAD system has been             the feel of some physical system, however this restricts
used successfully as a key part of a distance delivered         the client to using Microsoft Internet Explorer. In
Masters Level module, allowing experimental validation          addition, security settings may prevent these Active-X
where none would otherwise have been possible. Student          controls from executing. This results in a large amount of
feedback has so far been very positive, with students           time configuring multiple machines and inconvenience
particularly appreciating the virtually unlimited access        for a single user. Furthermore it is likely that users
and the ability to easily visualize the output via the use of   requiring a one off trial will be unable or unwilling to
the video clips simultaneously with near real-time              install custom software.
analogue data and the reinforcement of the sense that they      A Java applet can be created to run on a standard Internet
are accessing real experimental data these clips provide.       browser to control the experiment [3,4,5]. Java has the
                                                                advantage of cross-browser and cross-platform support
                                                                and the majority of today’s computers are java enabled.
2. System Overview                                              However, writing custom java applets for experiments is
                                                                not trivial and could result in an overly complex solution
The essential idea of remote access to experiments is that      that is difficult to maintain.
the physical experimental equipment does not need to be         The approach taken here was to simplify all aspects of the
in the same location as the person performing the               remote experiment to create a reliable, maintainable
experiments. Indeed, the experimental equipment may be          system. A standard web form is used at the client end to
in another room, building or even country. This                 request experiments and retrieve and view the associated
detachment from the physical system enables greater             results. At the server end, LabVIEW software performs
flexibility in performing experiments, but comes at the         the experiments while monitoring, receiving and sending
risk that the user may doubt the realism of the data or feel    data across the internet to the client computer. The client
detached from the experiment being performed no matter          only requires standard java support to view video clips.
how good the interface [1]. The Internet or LAN network
is ideal for the transfer of data in such cases and though
several different methods can be employed to deliver            4. Experiment Server Computer
remote experiments, the basic structure adopted here is
shown in figure 1.                                              The Experiment Servers perform the experiments and
A client computer sends a request to the central ReLOAD         communicate across a LAN and in turn pass the data on
web server for experimental data across the Internet, from      via the internet to the client PC. LabVIEW is an ideal
any location with an Internet connection. Depending on          platform to control experimental equipment though a data
the type of experiment requested, the ReLOAD web server         acquisition system. It enables intuitive design of control
redirects this request via a LAN to one of several              systems through a graphical user interface. Indeed, the
Experiment Servers, physically connected to either a            majority of undergraduate experiments within the School
single or multiple pieces of experimental equipment. The        of Mechanical Engineering are constructed using this
Experiment Server converts the request into a series of         system and it is also taught to students in several modules
commands appropriate for the specific equipment being           within the school. It is also a suitable and convenient
run. Upon completion of the experiment, the Experiment          option with which to develop the communications
Server sends results in the form of data images and video       functionality of the system. Using a single programming
clips back to the user on the client computer via the           environment and language simplifies the development
ReLOAD web server, providing virtually the same level of        process since no inter-language interfaces are required
information as would be obtained by performing the              making the system easier to adapt.
experiment locally.                                             The front panel of the Experiment Server computer is
                                                                designed for use as a demonstration display, allowing
staff, students and visitors at the hosting institute to         The user can view the front-panel or request to control it.
observe the experiments being requested remotely.                When viewing, the user sees a ‘live’ representation of the
Therefore all parameters used by the experiment are              front panel. When controlling, the user can do everything
displayed, together with a preview window for the video          that would be possible if using the actual VI, for example
camera and the results graph. Below this a tabbed panel          altering controls, pressing buttons, etc.
displays either a log of the requests submitted to the           Using this method a continuous stream of information is
system or the setup panel. The logs of requests shows the        passed back and forth in order to synchronise the
parameters submitted by users before they have been limit        embedded panel with the actual VI. The number of users
checked so any attempted misuse of the system will be            allowed to connect to the LabVIEW Web Server is
highlighted. The shutdown button should be used to close         restricted by a license and as standard this is set at a single
the system down. It ensures that any requests being              remote user. To increase this a new license must be
processed are first dealt with before the ‘shutdown’ state.      purchased to cater for the number of people expected to
                                                                 connect at a single time. When the number of users
                                                                 attempting to connect to the Web Server exceeds that on
5. System Communication                                          the license a customisable error message is returned.

LabVIEW allows two alternative methods, LabVIEW                  Comparison of Methods : The two methods discussed
HTTP Server with CGI Scripts or embedded ActiveX                 above are suited to different applications. The Web Server
control panels, to communicate the required information          and embedded front panel allow continuous interaction
between computers to perform remote experiments.                 between the user and experiment whilst using the HTTP
The LabVIEW Internet Toolkit is available as an add-on           Server and CGI script allows only a single interaction in
to the standard LabVIEW package. It provides an HTTP             which parameters are sent and results received.
server and a set of related VI’s (virtual instrument, the        The LabVIEW Web Server and embedded panel method
term used for a LabVIEW function or program). Of                 has several shortcomings. First is the need to purchase as
particular relevance to this application are the CGI             many licenses as the number of users that may be
(common gateway interface) script VI’s. CGI scripts are          expected on the experiment. Secondly is the provision of
typically used in association with a form on a web-page.         video; since continuous experimental interaction is
Forms are defined using standard HTML elements and               possible it would be consistent to use streaming video
allow a user to enter information into a number of fields        footage, giving a ‘live’ indication to the state of the
and post it to a server. The server receives the post and        experiment. However, video quality is dependant on
invokes the relevant CGI script. The script then processes       connection speed. The final factor is the reliance on
the information and sends back a response (typically a           proprietary plug-ins in order to embed the VI’s panel in a
web-page) through the server to the user (once the user          web-page.
has posted a form the web-browser will await a response          The HTTP Server with CGI scripts method does not
from the server and display it once received).                   encounter these problems. The HTTP server does not
By combining the LabVIEW HTTP Server and the CGI                 require a license dependant upon the number of users
script VI’s it is possible to achieve the required interaction   expected and only a widely supported, multi-platform
between the user and the experiment. The experimental            java plugin is required, in this case to display the video.
parameters can be entered into a form on a web-page and          This makes it the more attractive choice for this
posted to the LabVIEW HTTP Server. This will then                application. However, it is important to note that only
invoke a CGI script VI that interprets these parameters          experiments not requiring continuous interaction are
and uses them to conduct the experiment. When the                possible using this method.
experiment is complete the results can be returned in the
form of a web-page to the user. It should be noted that the
CGI script VI’s can only be used in conjunction with the         6. Experiment Request Control Software
LabVIEW HTTP Server.
The communication between the web browser and HTTP               The server can potentially receive multiple requests from
Server can be over any network supporting HTTP traffic,          clients. Each experiment requested must be performed
for example a departmental LAN or the Internet. The              and since each experiment requires a finite amount of
latest version of LabVIEW contains tools aimed at                time, a queuing system is used to ensure all experiments
making it easier to control VI’s using a web-browser. The        are performed in an orderly manner.
front panel of a VI can be embedded into a web-page. It          Two queues store and order incoming experiments and
takes the form of an ActiveX component if Internet               outgoing results. When the server receives data posted
Explorer is being used, or a plug-in for Netscape. The           from the ‘input parameters’ web-page it is added to the
computer running the VI must also run the LabVIEW                back of the request queue. When the experimental
Web Server which controls communication between the              apparatus is ready to conduct an experiment it removes
embedded front panel in the web-page and the actual front        the oldest request from the queue and uses the parameters
panel of the VI.                                                 given. When the experiment concludes the results are
                                                                 placed at the back of the results queue. Both queues
operate on a first in first out basis. The queue creator        equipment, shown in figure 3, was designed and built by
constantly waits for entries in the results queue and sends     staff and students of the School of Mechanical
them back to the user as soon as they are received.             Engineering at Leeds and is based on equipment used in
Each item in the request queue contains a number of             face-to-face laboratory sessions in the second year of an
elements. Firstly are the IP Address that the request was       undergraduate degree programme.
submitted from and the CGI connection information. The          A pulse sent to the electromagnet is used to commence
CGI connection information is required in order to send         the experiment by displacing the free end of a cantilever
the results back to the user. The experiment name is sent       beam. The resulting vibrations are measured using strain
as a text string to provide for the possibility of running      gauges mounted to the beam surface. A signal from the
multiple experiments on the same machine. Lastly, the           strain gauge is used to derive a velocity signal, which is
parameters sent from the web-page form are included as a        fed back into an electromagnetic shaker to provide active
keyed array. This is a cluster of two arrays, the first         damping of the beam. The user is free to vary the
containing the parameters names (called keys) and the           magnitude of the feedback signal within limits, so that the
second their values. An example for the vibrating beam          effect of increased damping can be observed. The user
experiment is given in figure 2.                                can also adjust the magnitude of the pulse signal sent to
                                                                the electromagnet so that tests at various amplitudes of
                  LabVIEW HTTP Server                           vibration can be undertaken. In addition the user can
                                                                select the period of time over which the experimental data
        Form POST Data           HMTL and reply Data            is collected and whether or not they wish to see a video
                                                                clip showing the movement of the beam.
                    Queue creator                                                                                               Beam Displacement ( D= 0 )

         CGI Read Request      CGI Write Reply                                                  0.5


                                                                                                0.4


                                                                                                0.3                                  Low Damping
                                                                                                0.2


                                                                 D is p la c e m e n t          0.1


                                                                                                 0




                                                                                                                                                                                        Strain Guages
                                                                                         -0.1

         Request Queue           Results Queue                                           -0.2


                                                                                         -0.3



                            Beam experiment                                              -0.4


                                                                                         -0.5




                                                                                                                                                                                                             Cantilever Beam
                                                                                                         0    0.5    1    1.5    2          2.5        3      3.5   4   4.5   5
                                                                                                                                       Time, seconds




                                                                                                                                Beam Displacement ( D= 50 )
     Figure 2 The Queue Communication Process                                              0.5


                                                                                           0.4


                                                                                           0.3
                                                                                                                         Medium Damping
It should be noted that the ‘keys’ are simply the names                                    0.2




given to the individual fields in the web-page form. Each                                  0.1
                                                                D is p la c e m e n t




                                                                                                                                                                                                    Shaker
                                                                                                 0


item in the Results queue consists of two elements. Firstly                              -0.1




the CGI connection information required to send the                                      -0.2


                                                                                         -0.3


results back to the user. Secondly, a text stream                                        -0.4




containing the results web-page location.                                                -0.5




                                                                                                                                                                                  Electro Magnet
                                                                                                      0       0.5    1    1.5    2          2.5        3      3.5   4   4.5   5
                                                                                                                                      Time, seconds




                                                                                                                                                                                                                          Webcam
To prevent large numbers of requests amassing under                                                                             Beam Displacement ( D= 0 )

heavy demand the queue size is limited. The maximum                                             0.5


                                                                                                0.4


number of requests allowed in the queue is set by a                                             0.3
                                                                                                                                 High Damping
default value to 10, judged to be an acceptable                                                 0.2


                                                                                                0.1
                                                                        D is p la c e m e n t




compromise between allowing multiple users and                                                       0



minimizing the delay before results are returned. If the                                        -0.1


                                                                                                -0.2

queue is full when a request is submitted, the user is sent a                                   -0.3



‘busy’ web-page in reply, informing them that the system                                        -0.4


                                                                                                -0.5

is busy. The request is then discarded. A further check can                                               0    0.5   1    1.5     2         2.5
                                                                                                                                       Time, seconds
                                                                                                                                                       3      3.5   4   4.5   5




be performed to check if the IP address of a submitted
request is the same as one currently in the queue. If this is                                                 Figure 3 Experimental Equipment and Results
the case the request is discarded and the user is again sent
the ‘busy’ web-page. This prevents a single user from           In this particular application, the equipment is used as a
dominating the queue system.                                    key part of a laboratory session for a level 2
                                                                undergraduate module in Vibration and Control. At this
                                                                level the beam is treated as a 1 degree of freedom, lumped
7. Experimental Equipment                                       parameter, mass/spring/damper dynamic system. The
                                                                session is used to reinforce principals taught during the
The example chosen here to demonstrate the principle of a       lectures which relate the position of the roots of the
typical ReLOAD enabled remote experiment is one which           characteristic equation on a root locus diagram, to the
implements negative velocity feedback to apply active           type of free vibration response that is seen in practice.
damping to a vibrating beam. This purpose made                  Hence visualisation of the beams motion is a crucial
component, which is captured using the webcam. As                the experiment and a set of standard controls allowing
damping is increased students see the time taken to settle       replays, rewinding, fast-forwarding etc. The video is
is reduced and the frequency of oscillation is decreased.        displayed using a java applet providing a consistent
Careful examination of the data will however reveal to the       experience across different browsers and does not require
more observant students, that the roots do not follow a          the user to download any special plugins. A link to the
classic semicircular locus as damping is increased, since        results data file is provided allowing it to be imported into
the system is actually a continuous system and not a             spreadsheet applications such as Excel and a copy to
lumped parameter system. If the damping is increased             clipboard function is also included, which serves a similar
beyond a particular level the point of attachment of the         purpose. The results web-page is shown in figure 5.
damper becomes locked, effectively preventing the
damper dissipating energy and the system reverts to a
lightly damped system with an effectively shortened
beam. This apparently contradictory phenomena of
increasing damping resulting in a more lightly damped
system is returned to in later modules, when multi degree
of freedom systems are introduced.


8. The Client Computer
The client computer can be any computer with
appropriate Internet browser software installed and is
only required to send a relatively small amount of data to
the ReLOAD web server. Upon entering the experiment
internet address, the user is presented with the Parameters
Input web page as shown in figure 4.




                                                                             Figure 5 The Results Web-Page


                                                                 9. Evaluation
                                                                 The particular remote experiment described in this paper
       Figure 4 The Parameter Input Web-Page                     has been used successfully since 2001 as an alternative to
                                                                 an almost identical face-to-face version of the beam
A standard HTML form is used to submit information to            experiment. Students are offered the option of taking the
the ReLOAD Web server. Parameters are entered into the           ReLOAD enabled version, in situations where they find it
text boxes. In order to post the form to the server a submit     impossible to attend the face-to-face session, due to
button is used. Usually a submit button will directly post       illness or other unavoidable circumstances. Since it was
the information in a form to the defined address. However        introduced, student feedback has been very positive, with
in this case a java-script function first checks to see if the   most appreciating the alternative approach available.
user has already pressed the submit button. This is to help      Many overseas students have also be able to perform the
prevent the user from repeatedly clicking the submit             laboratory session even when personal circumstances
button, causing spurious experimental requests. The form         have prevented them from being in the UK at the time of
will be submitted only if the function returns a Boolean         the timetabled face-to-face session. In end of session
‘true’. Note it would be possible to remove this java script     examinations, the principals which are reinforced during
and use the server computer to delete duplicate                  this laboratory session are examined. Though there is only
experiment requests.                                             a relatively small sample (around 5 students per year)
The results web-page presents results from the experiment        analysis of their scores suggests they perform equally well
to the user in a number of ways. A graph is displayed            as those who take the face-to-face version.
showing the beam’s displacement and the control signal           System performance evaluation has been encouraging,
of the damper. Alongside are the embedded video clip of          with tests being performed by multiple users at various
global locations showing minimal delay in results being       Students are familiar with the webcam, accelerometers
served to the client computer, usually within less than 20    and other hardware associated with the ReLOAD
seconds, depending on the queue size and internet speed.      experiment but they stills consider this approach to
                                                              delivering an experiment to be "high tech" (even though
                                                              the actual equipment is exactly equivalent to that students
10. Delivering Additional Experiments                         have used in past years in these courses). As a way to
                                                              augment the face-to-face and hands-on experiments then,
The solution to performing remote experiments outlined        the ReLOAD system appears ideal. Students in Mech 364
here uses a vibrating beam experiment as an example. The      and 465 will cycle the through a version of the ReLOAD
system provides a method of performing any remote             beam experiment during March 2006 and will be polled at
experiment not requiring real-time video feedback or data     the end of the term regarding the effectiveness of
transfer. The system allows for multiple pieces of            ReLOAD to deliver the experiment.
equipment to be connected to a single Experiment Server
and numerous variants of an experiment to run on each         11. Conclusions
piece of equipment. In addition the ReLOAD web server
supports access to multiple Experiment Server computers       The solution presented in this paper allows remote control
allowing a highly flexible and reconfigurable system. This    of an experiment using only a typical java enabled web
option is especially useful if few experiments are            browser. The solution does not offer continuous control;
performed, or if demand for a particular experiment is        once a set of parameters have been submitted to the
focused around a specific time period. For example, a two     experiment the user must wait until the experiment
week time slot to complete module course work. It is now      terminates before resubmitting revised values. Continuous
straightforward to create new ReLOAD enabled                  control is possible at the current time but was not
experiments, since each experiment is controlled using a      considered necessary or viable due to the limitations of
standard LabVIEW VI. A set of tools is used to create the     the streaming video that would be needed to support it.
necessary webpages with minimal intervention required.        However, improvements in video streaming technology,
This has resulted in a relatively straightforward and rapid   coupled with inevitable increases in network bandwidth
expansion of the system. Numerous new experiments are         and processing power may soon make this a more realistic
currently under development, one being in collaboration       option. The current solution offers a simple, easily
with the University of British Columbia (UBC), this being     adaptable system based upon well-established Internet
an excellent transatlantic test case for the ReLOAD           technology and makes most use of available time, while
system.                                                       one person is analysing their data numerous other
UBC currently maintains two vibrations courses in the         experiments can be performed for other users. Both
undergraduate Mechanical Engineering curriculum. One          student and system evaluation has so far proved very
is at the third year level (Mech 364) and one is at the       encouraging, and based on initial success additional
fourth year level (Mech 465). Both courses incorporate        experiments are currently being commissioned.
laboratory experiments and both courses are large with
more than 90 students registered in each. Classes this size
render classic face-to-face experiments with small group      References
sizes and adequate instruction, time-consuming and
logistically challenging. The acquisition of psychomotor      [1] A. Selmer, M. Goodson, M. Kraft, S. Sen, V.
skills and "mechanical intuition" associated with face-to-    McNeill, B. S. Johnston, C.K. Colton, Process Dynamics
face experiments is a priority at UBC but the requirement     and Control Experiments Performed Across the Atlantic
for small group sizes and adequate supervision combined       Workshop on Internet Accessible Laboratory Experiments
with the large class sizes means that providing more than     in Chemical Engineering Ed., Cambridge, UK 2005.
two face-to-face experiments in a four month semester is      [2] M. Hites, Creating and running undergraduate
not plausible. Ideally, it would be possible to augment       experiments controlled through the internet, Proc. IL/IN
these experiments with additional experiments that could      Conf. of American Soc. of Eng. Ed., 2002, 192-198
be performed individually by each student and at each         [3] C. Rohrig, A. Jochheim. The virtual Lab for
student's own pace. The ReLOAD system offers an ideal         controlling real experiments via Internet, Proc. IEEE
method to provide extra experiments that can be done          International symposium on computer aided control
individually and as part of the students' own timetable.      system design. 1999 279-284.
The same detailed lab reports and documentation of            [4] J. Sanchez, F. Morilla, S. Dormido, J. Aranda, P.
methods and results that is provide for face-to-face          Ruiperez. Virtual and remote control Labs using Java: A
experiments will be required from the ReLOAD                  qualitative approach, IEEE Control systems magazine,
experiments. In addition, the ReLOAD experiments can be       April 2000, 8-20.
demonstrated in a class or tutorial session with the same     [5] C. Rohrig, A.Jochheim, Java-based framework for
procedures and software as the students themselves will       remote access to laboratory experiments, Proc. IFAC/
use. In-classroom laboratory instruction of this type is      IEEE Symp. on Advances in Control Ed., Australia, 2000.
often not plausible with the face-to-face experiments.

				
DOCUMENT INFO