Article Preview
Top1. Introduction
Nuclear energy (NE, Weinberg, 1994) amounts to the energy that is required in order for an atom to retain its stability, i.e. for the protons and neutrons that comprise the nucleus of the atom to remain bound to each other. NE is released when
- (1)
the nucleus of an atom is split into smaller nuclei (nuclear fission, NFi),
- (2)
the nuclei of two or more atoms are integrated into a larger nucleus (nuclear fusion, NFu),
where, in both cases, the released neutrons of the nuclei of the atom(s) involved in the process are vital not only for producing NE, but also for sustaining the chain reaction. In a nutshell:
- (1)
the difference in mass (and, thus, energy) between the original and resulting nuclei causes the release of significant amounts of NE (especially when compared to the size of the interacting elements), which - following collection and conversion - can be used for turning turbines, and consequently driving generators to produce electricity 1;
- (2)
the neutrons released from these nuclei sustain the NFi/NFu phenomena.
The practical exploitation of NE has become of particular interest since the last century2, with - to date - NFi constituting the main means of energy production. The last 20 years have further brought about a shift in nuclear (power) plant (N(P)P) construction and operation, with the focus moving away from building new and towards maintaining existing N(P)Ps. Consequently, special emphasis has been placed upon the need for (i) comprehensive plant life management (PLiM) and (ii) cost-effective as well as reliable instrumentation & control (I&C), both of which are crucial not only for avoiding a forced shut-down due to unavailability, but also for maintaining optimal functionality of the ageing N(P)Ps.
Control, diagnostics and fault detection, monitoring, N(P)P operations, proliferation and resistance applications, sensor and component reliability, spectroscopy and – finally - fusion supporting operations have been established as key-issues of safe, maximally efficient real-time adjustable N(P)P operation (Ma & Jiang, 2011). Complementary to the traditional signal and image/sound processing techniques that have been applied to these key-issues, the computational intelligence (CI) (Pedrycz, 1997) paradigm - which was put forward in the early 1990s as a set of computationally effective, adaptive, resistant-to-noise as well as to missing and/or partly erroneous information - has provided alternative on-line/real-time, robust, non-invasive methodologies for monitoring, controlling and predicting N(P)P operation. For the last 25 years, the resulting CI-based implementations and applications have afforded prompt, reliable as well as robust response to these key-issues, thus demonstrating the potential of CI either as a superior stand-alone tool, or - whenever advantageous - in combination with traditional signal and/or image processing/analysis techniques.
The present review communicates the application of CI methodologies to the aforementioned N(P)P key-issues, as reported in the relevant primary literature of the major publication and dissemination media (listed in alphabetical order):
- (1)
Annals of Nuclear Energy (ANE), published by Elsevier,
- (2)
Fusion Science and Technology (FST), Nuclear Science & Engineering (NSE) and Nuclear Technology (NT), all three published by the American Nuclear Society (ANS), and
- (3)
Progress in Nuclear Energy (PNE), also published by Elsevier