Research reactor loading pattern optimization using estimation of distribution algorithms

Conference · Sat Jul 01 04:00:00 EDT 2006

OSTI ID:22039573

Jiang, S ^[1]; Ziver, K ^[1]; Carter, J N; Pain, C C; Eaton, M D; Goddard, A J. H. ^[1]; Franklin, S J; Phillips, H J ^[2]

---

1. Dept. of Earth Science and Engineering, Applied Modeling and Computation Group AMCG, Imperial College, London, SW7 2AZ (United Kingdom)
2. Imperial College, Reactor Centre, Silwood Park, Buckhurst Road, Ascot, Berkshire, SL5 7TE (United Kingdom)

A new evolutionary search based approach for solving the nuclear reactor loading pattern optimization problems is presented based on the Estimation of Distribution Algorithms. The optimization technique developed is then applied to the maximization of the effective multiplication factor (K{sub eff}) of the Imperial College CONSORT research reactor (the last remaining civilian research reactor in the United Kingdom). A new elitism-guided searching strategy has been developed and applied to improve the local convergence together with some problem-dependent information based on the 'stand-alone K{sub eff} with fuel coupling calculations. A comparison study between the EDAs and a Genetic Algorithm with Heuristic Tie Breaking Crossover operator has shown that the new algorithm is efficient and robust. (authors)

• Research Organization: American Nuclear Society, 555 North Kensington Avenue, La Grange Park, IL 60526 (United States)
• OSTI ID: 22039573
• Country of Publication: United States
• Language: English

Related Subjects

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS
ALGORITHMS
CALCULATION METHODS
CONSORT-2 REACTOR
CONVERGENCE
MULTIPLICATION FACTORS
OPTIMIZATION
REACTOR FUELING