Otimização de arranjos submarinos por algoritmo genético utilizando diferentes técnicas de tratamento de restrição

Abstract

This work develops an integrated optimization model of the subsea layout, which combines the performance of the Genetic Algorithm and a modified A* Algorithm, seeking to minimize the installation costs (CAPEX) of the subsea layout. The following are optimized: 1) the positioning of the platform; 2) positioning of subsea manifolds; 3) number of required manifolds; 4) system pipeline routing; and 5) connections between the subsea equipment. Minimum distance between equipment, interference with subsea obstacles and pipeline routing limitations are considered as restrictions. Additionally, the performance of the methodology is evaluated through the application and comparison of different constraint handling techniques, including static penalty, adaptive penalty, adaptive ε-constraint, tournament selection method and multiple constraint ranking. The case study addresses the application of the methodology to evaluate the importance of choosing the constraint handling technique for the evolution of the algorithm and, therefore, in the quality of the result. It is verified that the static penalty technique produces better results in terms of fitness than the others applied.