An Interactive Genetic Algorithm with c-Means clustering for the Unequal Area Facility Layout Problem

Unequal Area Facility Layout Problem (UA-FLP) has been addressed by several methods. However, UA-FLP has only been solved regarding quantitative criteria. Our approach includes subjective features to UA-FLP, which are difficult to take into account with a classical heuristic optimization. For that, an Interactive Genetic Algorithm (IGA) is proposed that allows an interaction between the algorithm and the Decision Maker (DM). Involving the DM knowledge into the approach guides the search process, adjusting it to the DM´s preferences at every iteration of the algorithm. The whole population is evaluated through the DM subjective evaluations of the representative solutions, which are different enough and are chosen by means of c-Means clustering method. The empirical test results show that the proposed IGA is capable of capturing DM preferences and that it can progress towards a good solution in a reasonable number of iterations.