Proceedings of ICAPP'10, San Diego, CA, USA, June 13-17, 2010
In order to find the most economical loading pattern (LP) considering multi-cycle fuel
loading, multi-objective fuel LP optimization problems are examined by employing an adaptively
constrained discontinuous penalty function (ACDPF) method. This is an improved method to
simplify the complicated acceptance logic of the original DPF method in that the stochastic effects
caused by the different random number sequence can be reduced. The effectiveness of the multiobjective simulated annealing (SA) algorithm employing ACDPF is examined for the reload core
LP of Cycle 4 of Yonggwang Nuclear Unit 4. Several optimization runs are performed with
different numbers of objectives consisting of cycle length and average burnup of fuels to be
discharged or reloaded. The candidate LPs obtained from the multi-objective optimization runs
turn out to be better than the reference LP in the aspects of cycle length and utilization of given
fuels. It is note that the proposed ACDPF based MOSA algorithm can be a practical method to
obtain an economical LP considering multi-cycle fuel loading
