A. Luchetta

The RFX centralized control, data acquisition and machine protection systems

Abstract RFX employs a centralized approach to the tasks of control, monitoring, data acquisition and machine protection. In this paper we describe the requirements, the structure, the components, and the operation of the corresponding systems of RFX. To guarantee a high degree of reliability, a strict subdivision has been imposed from the very start, between the control, monitoring and data acquisition system SIGMA (“sistema di gestione, monitoraggio ed acquisizione dati”) and the global fast machine protection system SGPR (“sistema generale di protezione rapida”). SIGMA has been designed for the following signals: about 5000 slow, mostly digital, signals, which provide non-shot-related information from the plant and commands to the plant, and about 1000 channels of fast (2 kHz–1 MHz), shot-related data from the plant, which produce up to 20 Mbytes of data per shot. In addition, fast plant-wide timing signals (precision better than 10 μs) have to be provided. SIGMA employs two distinct technologies: industrial-type programmable controllers (PLCs) handle the slow signals; a centralized VAX-VMS computer cluster with front end according to the CAMAC standard takes care of the fast system components. The CAMAC-based system covers both the fast data acquisition and the timing requirements. Workstations are used as operator consoles for the fast part of the system; personal computers are used as consoles to the PLCs. All components of the system communicate via a single, fibre optic Ethernet. A single PLC acts as supervisor of the entire shot sequence. The PLCs are programmed in an assembler-type language at the lower level and in a language according to the grafcet standard at the higher level. The PC-based consoles employ a commercial package for plant monitoring and control. The VAX-based systems run a purpose-developed software package, known as mds-plus , which provides integrated operation of timing and fast data acquisition. mds-plus is a joint software development with the MIT Plasma Fusion Center and the Los Alamos National Laboratory. The machine protection system SGPR has to deal with up to 50 possible requests for fast intervention and to distribute the corresponding command signals with an overall reaction time of 1 ms. SGPR is based on dedicated hardware which implements decision logic for intervention requests of four different urgency levels. It despatches the intervention commands to the corresponding protective devices in the various local units. All signal paths and the decision logic are duplicated with continuous automatic checks for integrity. Single transmission from the sensors and to the protective devices is by duplicated fibre optic lines with continuous self-test. Operation of RFX is under complete control of SIGMA from a single, central control room. It houses the supervisor console, all subsystem consoles (PCs and workstations) as well as some additional equipment (SGPR display, printers, television equipment, etc.). Shot execution follows a strict sequence, which is implemented as a unique state machine on all subsystems (PLCs and VAX computers). The supervisor console is the operator interface to the system-wide state machine which controls the shot sequence. The performance of the system is considered satisfactory with scope for further improvement. The cycle time of the PLCs is below 200 ms; the picture update time on the PC-based consoles is below 3 s. The fast system acquires all channels (at the moment around 15 Mbytes of uncompressed data per shot) within around 10 min. This time includes the display of several hundreds of measurement channels on different workstations in the central control room, and the automatic execution of a number of data analysis programs.

The data acquisition system of the RFX experiment

The paper describes the experience gained in the development and maintenance of the data acquisition system for the RFX nuclear fusion experiment. The design of the system core started in 1987 and the system was ready at the beginning of the machine operation in late 1991. Since the early design stage it was clear that a careful definition of the system architecture was crucial for allowing the system to cope with the expected improvement in the hardware performance. Not all the initial design choices proved successful for this purpose. In the paper the (positive or negative) consequences of the main design choices of the system are reported.

Update of control and data acquisition in RFX

The control and data acquisition system of the RFX experiment was designed and implemented during the second half of the 1980s, and several hardware components used in the system are currently obsolete. For these reasons an extensive redesign of the control and data acquisition system has been considered at RFX, with the aim of preserving well-tested concepts but at the same time replacing out-of-date technology. The update of the control and data acquisition system is possible due to the fact that major changes are to be accomplished in the machine power supply systems, requiring a long period of machine shutdown. In addition, important modifications of the machine load assembly are under consideration, which introduce new requirements for the control system. The introduction of modern technologies for distributed data acquisition is discussed and the planned organisation of the new system is presented.

The data acquisition system of the RFX tomographic diagnostic

The VME data acquisition system is described for a new tomographic diagnostic currently under development at the RFX nuclear fusion experiment. The system architecture, and its integration into the existing Control and Data Acquisition System of RFX is presented. A modelling approach in the definition and analysis of the distributed software architecture is then discussed. Finally the user interface requirements of the system and the implemented solutions are presented. I. INTRODUCTION RFX is a large magnetic confinement, nuclear fusion experiment currently in operation in Padova, Italy at the Istituto Gas Ionizzati of CNR [ll. Measurement of the intensity of soft X-ray and bolometric emission from the plasma in a fusion device can be related to properties such as plasma position, shape, impurity distribution, and magnetohydrodynamic phenomena. The use of tomographic methods, which provide multiple views of plasma emissivity along chords through a plasma crosssection, allows for the reconstruction of the spatial distribution of the plasma emissivity. In RFX, plasma emission is measured along up to 78 chords during the plasma pulse which can last up to 250 ms and is repeated about every 10 minutes. Due to the limited number of plasma emission views, the tomographic procedures implemented for plasma emissivity reconstructions do not use finite element techniques as used in medical tomography, rather a least square approach of analytic solutions of the Radon transform constructed using base function expansions: Fourier series for the angular components of the emissivity, various base function expansions for its radial components depending on the kind of inversion implemented [2]. The result of a tomographic inversion is therefore a set of parameters to be used in the chosen angular and radial expansions of the emissivity function. Due to the limitation in the number of angular harmonics and in the maximum order of the radial expansion functions, which are related to geomeVic factors in the chord displacements, the number of such parameters is usually much less than the number of acquired views, and the results of the tomographic inversion form a more compact representation of the information produced by the diagnostic. The interpretation of such results requires however much care because of the finite order of the harmonic expansion which causes aliasing of any unresolvable features and produce unavoidable artefacts. For this reason, multiple tomographic inversions, based on different function expansions, are required in order to obtain a 0018-9499/96

SPIDER CODAS Evolution Toward the ITER Compliant Neutral Beam Injector CODAS

05

Distributed Continuous Event-Based Data Acquisition Using the IEEE 1588 Synchronization and FlexRIO FPGA

The ITER neutral beam test facility, currently under the construction at Padua, Italy, consists of two experiments: source for production of ions of deuterium extracted (SPIDER) to test the ITER size ion source and megavolt iter injector and concept advancement (MITICA) to test the full ITER neutral beam injector. While the former represents an intermediate step, also finalized to the ITER Diagnostic Neutral Beam, the latter will represent the final design and implementation of the full prototype of the ITER heating neutral beam (HNB) injectors. Consequently, for the control and data acquisition system (CODAS) of SPIDER, there are no requirements for compliance with ITER, whereas the MITICA HNB Plant System will comply with the ITER directives for plant instrumentation and control because final system will be eventually integrated in ITER. SPIDER CODAS is currently under commissioning and the integration of the different frameworks that operate in an integrated way is proven to be successful. The experience and solutions gained in SPIDER CODAS development will be reused as far as possible in MITICA and to provide at the same time the required ITER plant system compatibility, a set of ITER-like layers (networks) will be defined, using exactly the same protocol defined for ITER interfaces.

Communication and synchronization aspects of a mixed hardware control and data acquisition system

High-speed event driven acquisition is normally performed by analog-to-digital converter (ADC) boards with a given number of pretrigger sample and posttrigger sample that are recorded upon the occurrence of a hardware trigger. A direct physical connection is, therefore, required between the source of event (trigger) and the ADC, because any other software-based communication method would introduce a delay in triggering that would turn out to be not acceptable in many cases. This paper proposes a solution for the relaxation of the event communication time that can be, in this case, carried out by software messaging (e.g., via an LAN), provided that the system components are synchronized in time using the IEEE 1588 synchronization mechanism. The information about the exact event occurrence time is contained in the software packet that is sent to communicate the event and is used by the ADC FPGA to identify the exact sample in the ADC sample queue. The length of the ADC sample queue will depend on the maximum delay in software event message communication time. A prototype implementation using a National FlexRIO FPGA board connected with an ADC device is presented as the proof of concept.

General-purpose framework for real time control in nuclear fusion experiments

Aspects of the control and data acquisition system of the RFX nuclear fusion experiment are dealt with. The system is built around a local area network that connects programmable controllers, minicomputers with CAMAC front-end, and personal computes as operator consoles. These three types of nodes use compatible software that contains a set of low-level routines corresponding to levels one to four of the ISO OSI (Open Systems Interconnection) recommendations. A detailed description is provided of how the overall system synchronization is achieved. All subsystems (computer and programmable-controller based) follow the same state transition diagram (scheduler) and are coordinated by a supervisor system based on a programmable controller. The solution proposed for the precision timing and waveform generation and its integration with the overall system synchronization are described. >