Development and adaptation of meta-heuristic optimization methods in nuclear fuel management of soluble boron-free system-integrated modular advanced reactor to effectively increase the operation cycle length.

The optimal arrangement of fuel assemblies inside a nuclear reactor core is one of the most challenging topics in nuclear engineering. Three different powerful optimization methods (particle swarm optimization, genetic algorithm method, and dragonfly algorithm) are developed and applied on the System-integrated Modular Advanced ReacTor (SMART). A new program named "DYNAMIC-FMO" is written in MATLAB language programming to solve a multi-objective NP-Hard problem of determining the fuel assemblies' optimal position inside the SMART core. To attain the required nuclear data and simulate the SMART core, DRAGON/PARCS codes are coupled and employed in our "DYNAMIC-FMO" program. Optimization of the SMART core loading pattern has led to a significant increase in cycle length by 185 days and a decrease in power peaking factor by 12.5%. According to the results, the dragonfly algorithm has better performance than the particle swarm optimization and the genetic algorithm for core loading pattern optimization of SMART. [ABSTRACT FROM AUTHOR]