information coming from the images of the plasma facing components. - PDF document

FINAL WORKSHOP OF GRID PROJECTS “PON RICERCA 2000-2006, AVVISO 1575” 1 Applications of imaging analysis to tokamak fusion plasma by ENEA-GRID technologies M. Chinnici 1 , S. Migliori 2 , R. De Angelis 3 , S. Borioni 4 , S. Pierattini 5 1,4 FIM-INFOPPQ, Casaccia Research Center, Santa Maria di Galeria, Roma, Italy marta.chinnici@enea.it 2 ENEA-FIM, Enea-Sede, Roma, Italy 3 ENEA, Frascati Research Center, Associazione EURATOM-ENEA sulla Fusione, Italy 5 ENEA- FIM-INFOGER, Bologna Research Center, Bologna, Italy Abstract— Today a number of applications Index Te1ms — Tokamak, IDL, images analysis, Grid. in scientific fields (such as medical industry, astronomy, physics, chemistry, forensics, remote I. INTRODUCTION sensing, manufacturing, defense) rely upon images to store, display, and provide information The global demand for energy continues to about the world around us. The challenge to grow and there is an immediate need to find scientists, engineers and business people is to new sources of energy. Controlled thermo quickly extract valuable information from raw nuclear fusion offers significant potential image data. The primary purpose of our work advantages as a clear future source of energy in (within CRESCO Project) - i. e. converting a scenario where there is increasing concern images of a nuclear fusion plasma coming from the experiments ( shots ) of Frascati Tokamak that the emission of green house gases from Upgrade (FTU) into information by ENEA-GRID burning fossil fuels is producing tremendous infrastructures– is related to such issue. In climatic change by contaminating the particular, we use IDL (Interactive Data environment. Fusion is a clean energy source. Language) in order to quickly access image data In order to effectively use fusion as energy and to process them. IDL is a high-level source, it is of pivotal importance to monitor programming language that contains an extensive library of image processing and plasma evolution during fusion experiments. analysis routines. Real-Time Imaging analysis is one of the important non destructive, non-invasive, non- perturbative Frascati Tokamak Upgrade plasma diagnostics. This contribution (within the CRESCO Project) concerns the retrieval of

FINAL WORKSHOP OF GRID PROJECTS “PON RICERCA 2000-2006, AVVISO 1575” 1 information coming from the images of the plasma facing components. Data analysis is Tokamak experiments. In particular, we use normally performed off-line, due to the high IDL (Interactive Data Language) in order to amount of information to be processed, making analyze images of plasma. The real-time the data acquired by the camera quantitatively imaging analysis helps to visualize and useful only for post pulse evaluations. The estimate relevant parameters of plasma main difficulty in using visible or infrared phenomenon like run-away electrons or images for plasma feedback control is the fact Marfie. We ported a number of applications that real-time image processing is challenging which analyse and elaborate images coming and heavy in terms of processing time, from the tokamak database. In details, these especially when complex tasks are required. applications allow image quality improvement At the beginning, the visualization of FTU (noise reduction, contrast enhancement, images has been done under the Videoftu distortions correction), automatic classification Project. Since FTU image database is rather huge ( 4×10 6 Frames ), we used the multicase by pattern recognition algorithms and brightness analysis, used to detect images submission with multicluster queue to achieve which present a characteristic feature (quite efficient performance in terms of elapsed time recurrent in the plasma) in the brightness and CPU time. In order to reduce the run-time distribution. Since FTU image database is of the processes, the route of multicase rather huge, we used ENEA- GRID in order to processing has been utilized. achieve efficient performance in terms of IDL and ENEA-GRID ELAPSED and CPU time: benchmark tests we carried out on different platforms with different In CRESCO Project, under the task type of queues show real and meaningful performance improvements in running jobs by “Development and Integration of the GRID opting for this scheduling solution. and 3D Graphics” we ported a number of applications which analyse and elaborate the Setting images coming from the tokamak database. Image and processing analysis of FTU data In modern tokamaks visible and infrared through IDL applications: Falsecolor, video cameras are becoming more and more Database, Volume, Brightzone . In details, the important to monitor plasma evolution during applications allow image quality improvement fusion experiments. The real-time analysis of (noise reduction, contrast enhancement, FTU images performed by IDL applications distortions correction), automatic classification ( Falsecolor, Database, Volume, Brightzone ) by pattern recognition algorithms and can really provide relevant information to brightness analysis , used to detect images with control the plasma and the safety of the a characteristic feature (quite recurrent in the plasma) in the brightness distribution. machines. In the last years video cameras have been extensively used in magnetic confinement An example is the detection of bright toroidal bands ( i.e. lying in the vessel’ s fusion experiments for both the understanding of the physics and the safety of the operation. equatorial plane), which precede the onset of Both visible and InfraRed (IR) images can be regimes of enhanced gas recycling on the wall used not only to monitor the evolution of a (a phenomenon known in tokamaks as plasma discharge but also to evaluate specific ‘Marfe’), sometimes followed by distructive parameters, from the determination of impurity events. A second example is the identification radiation to the distribution of power loads on of bright spots, characterised by typical shapes

FINAL WORKSHOP OF GRID PROJECTS “PON RICERCA 2000-2006, AVVISO 1575” 1 and localization, which are due to high energy to achieve efficient performance in terms of electrons (‘runaway electrons’), potentially ELAPSED and CPU time: benchmark tests we carried out on different platforms with dangerous for the vacuum chamber. The different type of queue show real and applications allow a large number of tokamaks meaningful performance improvements in images’s classification according to specific running jobs by opting for this scheduling events and help understanding their correlation solution. The images analysis of FTU data with other physical quantities. On the other through IDL applications are in the output. hand the achievement of event recognition on For example, in order to consider the benefits timescales shorter than those of the evolution of multicluster queue, let’s consider an FTU of unwanted events, can provide a useful input experiment where a range of 20 shots (each for the feedback control of plasma operations shot contains 109 frame ) is produced by a single job: Method : The experimental evaluation of the algorithm in IDL environment has been  CPU TIME : ≈ 12 min performed through the use of the ENEA-GRID  CPU TIME (DATABASE) : ≈ 160 infrastructures for the submission and the hours execution of jobs ( Fig. 1-2 ). We used the multicluster queue to achieve efficient Experimental Tests with distributed run: performance in terms of elapsed time. Hence, the experimental evaluation of the algorithm in  ELAPSED TIME (10 parallel jobs ): IDL environment has been performed through ≈ 17 hours the use of the ENEA-GRID infrastructures for  ELAPSED TIME (30 parallel jobs ): the submission and the execution of job. ≈ 6.2 hours  ELAPSED TIME (60 parallel jobs ): ≈ 3.5 hours Fig.1 Example of Application. FTU images are input database ( 4×10 6 Frames ). Fig.2 IDL and ENEA-GRID submission: We have used the IDL resources in ENEA- Experimental result with multicluster queue GRID infrastructures for the submission and the execution of jobs. In details, we used the multicase submission with multicluster queue II. C ONCLUSION that run applications simultaneously on the 6 ENEA-GRID clusters ( Portici, Frascati, Experiments are conducted on ENEA-GRID Bologna, Trisaia, Brindisi, Casaccia ) in order

FINAL WORKSHOP OF GRID PROJECTS “PON RICERCA 2000-2006, AVVISO 1575” 1 facilities, on which we submitted the IDL applications in order to analyze images of FTU plasma. We compare the results of the submission and the execution of jobs with multicase submission and without, and show that the utilization of the ENEA-GRID technology is an efficient solution to reduced the run-time required to execute the simulations. In details, we use the multicluster queue to achieve efficient performance in terms of elapsed time. R EFERENCES [1] R. De Angelis, S. Migliori, S. Borioni, G. Bracco, S. Pierattini, A. Perozziello, Analysis of images from videocameras in the FTU tokamak , Review of scientific instruments, Vol. 75, N. 10 [2] http://ftu.frascati.enea.it Idl reference guide , Vol. 1 and Vol. 2 [3] [4] cresco/LA1/cresco_sp12_graf3d [5] M. Chinnici, S. Migliori, R. De Angelis, S. Borioni, S. Pierattini, Image analysis of a nuclear plasma: Frascati Tokamak Upgrade using IDL and ENEA- GRID technologies, (Poster) E-Science Conference, Naples, 27-29 May, 2008.