BeamlinesStatus_MeetingTango200307
Download
Report
Transcript BeamlinesStatus_MeetingTango200307
Overview architecture of
SOLEIL technical solutions
for Beam Lines Control
« A control system is a sea of network objects,
every part of the control system should be a
network object whether it is a low-level, highlevel or application oriented frameworks »
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
1
Plan
Introduction
Devices
Hardware
Equipment and sub-system devices
Process and Calculation devices
High level applications
Tango generic tools
GlobalScreen
Salsa
Passerelle
Status of the control of a beamline: SAMBA
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
2
Software Architecture
Process Driving Applications
Supervision
Applications
Archiving
Applications
Process & calculation devices
Device
TANGO Software bus
Process & calculation devices
Device
Device
Equipment and subsystem devices
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
3
Devices
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
4
Hardware devices
The electronics group has standardized many hardware systems for
Machine and BeamLines with commercial solutions:
Vacuum control done via PLC and Profibus -> Siemens
Motor Control -> Galil
All cPCI acquisition boards :
Digital I/O boards -> ADLink
Counting boards -> National Instruments
Analog I/O boards -> ADLink
Video acquisition -> National Instruments
…
So all the « generic » devices developed for these systems are used
on Machine and BeamLines :
Vacuum: VacuumValve, Thermocouples, IonPump …
Motor: ControlBox, GalilAxis, GalilSlit…
CPCI acquisition board: ImgGrabber, SingleShotDIO, EventCounting…
…
For non standardized systems such as detectors:
ICALEPS 2005
Specific devices were developed.
20th March 2007 – Tango Meeting – G.Abeillé
5
Equipment and sub system devices
For beamline systems such as :
Mirrors
Monochromators
Diffractometers
Some libraries were developed based on a long term specification work
done with beamline scientists :
MonochromatorLib
HKL lib…
The devices are then developed on top of these libraries:
If possible a generic device is developed: (GenericBender,
DiffractometerEulerian4C, …)
If not possible: a device per beamline (MonochromatorDiffabs,
MonochromatorSamba, …)
These libraries allow:
Reduced the device development cycle (One week per Monochromator)
Standardized optics devices control software interface
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
6
Process And Calculation devices
As for other devices, we tried to standardize:
ScanServer
DataFitter
TangoParser
Java Generic Devices (statecomposers,
attributelistewriters, ..)
BeamImgAnalyzer
Archiving devices
DataRecorder: data storage in Nexus Format (see next
presentation of S.Poirier)
…
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
7
High Level Applications
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
8
Generic Tango Applications
Intensively used for tests and commisionning
by beamline teams:
Jive
ATKPanel
Astor
ATKTuning
ATKTrend
DeviceTree
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
9
Supervision: GlobalScreen
As for the Machine, we provide
the GlobalScreen IDE which
allows non software developers
to realize their own Supervision
Applications:
A work group in Experiment
Division has defined « graphical
views » for common equipments:
Front End
Mirror, Monochromator, …
Each beamline has
developed its own
application using these elementary
components.
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
10
Process Driving: Salsa
Salsa is the application on top of the Scan Server
device to realize:
TimeScan
1D Scan
2D Scan
Deployed and will
be used on all
Beamlines
Used also on the
Machine
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
11
Process Driving: Passerelle
PASSERELLE
allow to
graphically
design
sequences by drag
and drop
and execute them
PASSERELLE is provided by a company called ISENCIA
That realized process driving for industrial companies.
PASSERELLE is based on an environment for scientific
modelisation and simulation: PTOLEMY (developed by the
Berkeley University)
ICALEPS 2005
http://ptolemy.eecs.berkeley.edu/ptolemyII/
20th March 2007
– Tango Meeting – G.Abeillé
12
Principle
We connect « boxes » and « wires »:
The « boxes » are called ACTORS, they execute an action.
The « wires » are called MESSAGES, they transfer data.
The graphic language for editing sequences provided all
functionnalities to build complex logics:
Loops
Comparison
Error management
Parameterization
Every actor is configurable with parameters.
Our team is developing actors that control Tango devices.
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
13
Passerelle users’ environments: IDE
IDE: development, configuration, execution,
debugging:
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
14
Passerelle users’ environments: Generic HMI
Graphical panel on top of any Passerelle sequence.
Possibility for users to customize the configuration panel :
Selection of the visible parameters.
Parameters’ labels.
Panels organization.
Allow to launch it,
and visualize
User logs
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
15
Passerelle users’ environments: Specific HMI
Principle:
Some beamlines were not happy with the generic HMI
For complex sequences the parameterization with the generic HMI is too
complicated.
Allow end users to do acquisition on the beamline without any
knowledge about Tango or Passerelle
These GUI appear to be « the beamline acquisition application »
But it is only possible for fixed range of sequences:
Because sequences modification may imply the evolution of the
configuration interface.
For the moment, these Specific HMI are developed by us
(Java):
But our aim is to provide « something » that could be
developed/customized by beamline teams (as for GlobalScreen)
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
16
Passerelle users’ environments: Specific HMI
At the moment, 4 Specific HMI have been developed:
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
17
Passerelle users’ environments:
Bossa Nova, a prototype
An application that allows to
create « a batch »
of many passerelle
sequences:
A table where
the user gives
the list of
sequences to
execute
The Passerelle
Generic HMI to edit
the parameters of each
sequence of the table
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
18
SAMBA
Presentation of the control of a
Beamline: SAMBA
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
19
Optical Hutch
Main equipment
Ctrl
Installed
Generic
devices
Absorber
Specific
device
Library used
I/O boards :ASL
Primary Slit
Mirror 1 (TPP+Bender)
Mirror +
Generic Bender
Beam Imager 1
VSL + ISL
Monochomator 1 (DCM)
Monochromator
Mono 2 (Quick EXAFS:
2008)
Monochromator
Beam Imager 2
VSL + ISL
Secondary Slit
Mirror 2 (TPP+Bender)
Mirror +
Generic Bender
Beam Imager 3
VSL + ISL
Shutter (+ support)
20th March 2007 – Tango Meeting – G.Abeillé
ICALEPS 2005
20
Experimental Hutch 1: EXAFS
Main equipment
Ctrl
Installed
Generic
devices
Specific
device
Library used
Exafs Slit
support Exafs table
Ionisation Chamber I0 (Alim
Iseg) + Femto + counters
Counting board
library
Ionisation Chamber I1 (Alim
Iseg) + Femto + counters
Counting board
library
Ionisation Chamber I2 (Alim
Iseg) + Femto + counters
Counting board
library
Sample support table
Cryostat, fluo detector
Counting board
library
Combined Experiment:
RAMAN, UV-Visible, DSC
Shutter
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
21
Experimental Hutch 2: SURFAS
Main equipment
Ctrl
Installed
Generic
devices
Specific
device
Library used
Surfas Slit
support SurfAs table
Event counting
Counting board
library
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
22
Samba: High level application progress
Tango Generic Tools: used intensively by beamline team
GlobalScreen: used intensively by beamline team.
BeamLine is currently developing a new version.
Salsa: Beginning to use it (tomorrow with Hercules
students).
Passerelle: A specific HMI is installed and is in test by us.
DataStorage system: installed and ready to be used.
SnapShot system: installed and ready to be used.
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
23
SAMBA
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
24
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
25
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
26
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
27
Conclusion
A big effort has been made by both the electronics and computing
teams to standardize as much as possible:
We spend many time in specifications meetings with our client before
starting coding something.
We have adopt the iterative cycle for development of user tools:
We now provide versions devices/applications in cycle of 1-3 weeks so that
client can test it and precise its specifications.
There is still many work to do for us:
Quality, e.g.:
many documentation still to been done
It is not clear what version of devices are deployed.
Integration: we provide many graphical tools the users see as small bricks.
At the end, it should be integrated in one/few applications.
Because everything is done in Java, there are no technical difficulties to do it.
The problem is that the clients do not have a clear vision of how they want to
have !!
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
28
ICALEPS 2005
20th March 2007 – Tango Meeting – G.Abeillé
29