Keeping in mind the proposed FAST experiment and aiming to meet basic requirements such as a modular and distributed architecture, where different control subsystems can be easily integrated at different times, and can operate either independently or in cooperation with other subsystems, at the end of last year we planned to upgrade the architecture of the FTU real-time system, improving in such a way its flexibility and modularity. We decided to adopt an available packages to reach our goal: MARTe. We report on the state of the art of the MARTe migration process, the difficulties dealt with, the benefits and advantages achieved, the progress made from our last report and, in particular, we describe the integration of the ODIN equilibrium reconstruction system in the real-time environment. The ODIN algorithm was already coded in previous works, but its integration in the real-time system has never been carried out at FTU. We illustrate how the MARTe architecture and the RTNet level allows for a first level of parallelization, distributing the data and/or time among nodes.
Progress in the migration towards the real-time framework MARTe at the FTU tokamak
Zaccarian, Luca;
2011-01-01
Abstract
Keeping in mind the proposed FAST experiment and aiming to meet basic requirements such as a modular and distributed architecture, where different control subsystems can be easily integrated at different times, and can operate either independently or in cooperation with other subsystems, at the end of last year we planned to upgrade the architecture of the FTU real-time system, improving in such a way its flexibility and modularity. We decided to adopt an available packages to reach our goal: MARTe. We report on the state of the art of the MARTe migration process, the difficulties dealt with, the benefits and advantages achieved, the progress made from our last report and, in particular, we describe the integration of the ODIN equilibrium reconstruction system in the real-time environment. The ODIN algorithm was already coded in previous works, but its integration in the real-time system has never been carried out at FTU. We illustrate how the MARTe architecture and the RTNet level allows for a first level of parallelization, distributing the data and/or time among nodes.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione