Application of CSA to loading pattern optimization problems for reload core in nuclear power plant

Abstract

Reload loading pattern design is the key content to economy, safety and flexibility of the operation after reloading in nuclear power plant. Loading pattern optimization is a typical combinatorial optimization problem with multi-objective and multi-parameter. Furthermore, the reload loading pattern optimization design is a very short time requirement, which makes it difficult for traditional optimization algorithms. In this study, a combined simulated annealing algorithm (CSA algorithm) are proposed. CSA effectively utilizes the characteristic regularity in the optimization space. Additionally, the combined database created by CSA can significantly increase the diversity of the algorithm, and greatly improve the optimization ability. In comparison to the traditional simulated annealing method, the optimization results of CSA are noticeably better. By contrasting the CSA with the improved genetic algorithm (GA), it can be seen that CSA can exploit the characteristic regularity of the reload loading pattern optimization more effectively, and the optimization efficiency and optimization results of the method are noticeably superior to GA.