evolutionary otimization of CANDU Reactor

evolutionary otimization of CANDU Reactor

(Parte 2 de 2)

The multi-cycle multi-objective optimization method was developed for the refueling simulation of a CANDU reactor by the proposed evolutionary algorithm, which combines the simple genetic operators, the elitism strategy and the heuristic rules, including the experience-based channel selection strategy together with a rough estimation of the core reactivity decay rate due to an irradiation and the reactivity increment due to a refueling operation. The implementation of the genetic algorithm was verified by a 6-month refueling simulation for a natural uranium CANDU-6 reactor, which was compared with the results obtained by the reactor-operation-experience-based and the generalized perturbation methods. The comparison showed that though the genetic algorithm successfully found the optimum solution, the benefit was relatively small from the viewpoints of the reactor performance parameters and the computing time when compared to the existing methods. In the future, the computing time of the genetic algorithm can be reduced if a parallel computing is introduced.

[1] H. Choi, “A fast-running fuel management program for a CANDU reactor,” Ann. Nucl. Energy, vol. 27, p. 1, 2000. [2] H. Choi and D. H. Kim, “Refueling simulation strategy of a CANDU reactorbasedon optimumzone controllerwaterlevels,” Nucl. Sci.Eng., vol. 151, p. 8, 2005. [3] M. D. DeChaine and M. A. Feltus, “Fuel management optimization using genetic algorithms,” Nucl. Technol., vol. 1, p. 109, 1995. [4] ——, “Fuel management optimization using genetic algorithms and expert knowledge,” Nucl. Sci. Eng., vol. 124, p. 188, 1996. [5] A. K. Ziver, J. N. Carter, C. C. Pain, C. R. E. de Oliveira, A. Goddard, and R. S. Overton, “Multi-cycle optimization of advanced gas-cooled reactor loading patterns using genetic algorithms,” Nucl. Technol., vol. 141, p. 122, 2002. [6] Y. Kobayashi and E. Aiyoshi, “Optimization of a boiling water reactor loading pattern using an improved genetic algorithm,” Nucl. Technol., vol. 144, p. 143, 2003. [7] D. E. Goldberg, Genetic Algorithms in Search Optimization and Machine Learning. Reading, MA: Addison-Wesley, 1989. [8] A.Jaszkiewicz, “Genetic localsearch formulti-objectivecombinatorial optimization,” Eur. J. Oper. Res., vol. 137, p. 50, 2002. [9] S. Esquivel, S. Ferrero, R. Gallard, C. Salto, H. Alfonso, and M. Schutz,

“Enhanced evolutionary algorithms for single and multiobjective optimization in the job shop scheduling problem,” Knowl.-Based Syst., vol. 13, p. 15, 2002. [10] G. T. Parks, “Multi-objective pressurized water reactor reload core design by non-dominated genetic algorithm search,” Nucl. Sci. Eng., vol. 124, p. 178, 1996. [1] A. Yamamoto, “A quantitative comparison of loading pattern optimization methods for in-core fuel management of PWR,” J. Nucl. Sci. Technol., vol. 339, p. 34, 1997. [12] B. Rouben, “RSFP-IST, the industry standard tool computer program for CANDU reactor core design and analysis,” in Proc. 13th Pacific Basin Nuclear Conf., Shenzhen, China, 2002, p. 75, PBNC-2002.

Authorized licensed use limited to: UNIV ESTADUAL PAULISTA JULIO DE MESQUITA FILHO. Downloaded on July 13, 2009 at 17:53 from IEEE Xplore. Restrictions apply.

(Parte 2 de 2)

Comentários