TY - JOUR
T1 - Upgrade of the automatic analysis system in the TJ-II Thomson Scattering diagnostic
T2 - New image recognition classifier and fault condition detection
AU - Makili, L.
AU - Vega, J.
AU - Dormido-Canto, S.
AU - Pastor, I.
AU - Pereira, A.
AU - Farias, G.
AU - Portas, A.
AU - Pérez-Risco, D.
AU - Rodríguez-Fernández, M. C.
AU - Busch, P.
N1 - Funding Information:
This work was partially funded by the Spanish Ministry of Science and Innovation under the Project No. ENE2008-02894/FTN .
Funding Information:
This work, supported by the European Communities under the contract of Association between EURATOM/CIEMAT, was carried out within the framework of the European Fusion Development Agreement. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
PY - 2010/7
Y1 - 2010/7
N2 - An automatic image classification system based on support vector machines (SVM) has been in operation for years in the TJ-II Thomson Scattering diagnostic. It recognizes five different types of images: CCD camera background, measurement of stray light without plasma or in a collapsed discharge, image during ECH phase, image during NBI phase and image after reaching the cut off density during ECH heating. Each kind of image implies the execution of different application software. Due to the fact that the recognition system is based on a learning system and major modifications have been carried out in both the diagnostic (optics) and TJ-II plasmas (injected power), the classifier model is no longer valid. A new SVM model has been developed with the current conditions. Also, specific error conditions in the data acquisition process can automatically be detected and managed now. The recovering process has been automated, thereby avoiding the loss of data in ensuing discharges.
AB - An automatic image classification system based on support vector machines (SVM) has been in operation for years in the TJ-II Thomson Scattering diagnostic. It recognizes five different types of images: CCD camera background, measurement of stray light without plasma or in a collapsed discharge, image during ECH phase, image during NBI phase and image after reaching the cut off density during ECH heating. Each kind of image implies the execution of different application software. Due to the fact that the recognition system is based on a learning system and major modifications have been carried out in both the diagnostic (optics) and TJ-II plasmas (injected power), the classifier model is no longer valid. A new SVM model has been developed with the current conditions. Also, specific error conditions in the data acquisition process can automatically be detected and managed now. The recovering process has been automated, thereby avoiding the loss of data in ensuing discharges.
KW - Classifier
KW - Multi-class
KW - Support vector machines
KW - Wavelet
UR - http://www.scopus.com/inward/record.url?scp=78049342884&partnerID=8YFLogxK
U2 - 10.1016/j.fusengdes.2009.10.004
DO - 10.1016/j.fusengdes.2009.10.004
M3 - Article
AN - SCOPUS:78049342884
SN - 0920-3796
VL - 85
SP - 415
EP - 418
JO - Fusion Engineering and Design
JF - Fusion Engineering and Design
IS - 3-4
ER -