Nuclear fuel loading pattern optimisation using a neural network

An algorithm to optimise the fuel loading pattern (LP) in nuclear reactors was developed using an artificial neural network (ANN) to generate arrangements for the fuel in the core. The core parameters were calculated with the WIMS-D4 and CITATION-LDI2 codes, and the minimization of the maximum power peaking factor (FPmax) was used to choose the best arrangements. To verify the algorithm a PWR reactor with approximately 1/3 reprocessed fuel loaded was considered. The neutronic performance of the obtained arrangements and the efficiency of the implemented method were analysed. Several configurations were found for the core presenting better characteristics than the reference configuration adopted, so indicating the viability of the developed methodology. The algorithm was applied to a core considering part of the loading with reprocessed fuels, however this technique can be used for standard loadings.