Solution of facility layout problems with pickup/drop-off locations using random search techniques

Due to non-polynomial hardness, the facility layout problem (FLP) becomes more critical when pickup/drop-off (P/D) locations are considered in the design of an open field layout under a manufacturing environment. This paper proposes an indigenous model of the facility layout problem based on random search techniques and its solution methodology using a genetic algorithm (GA), simulated annealing (SA) and a hybrid algorithm (HA). The paper illustrates the performance of different random search operating parameters in solving the facility layout problem considering P/D locations along the periphery of rectangular machine blocks. The preliminary experiments were carried out on three facility layout test problems having six, eight and ten machines in order to fix the different operating parameters such as crossover operator, crossover rate, initial temperature, temperature reduction factor, number of generations, population size, etc. The results of extensive preliminary experimentation were utilized to solve facility layout problems having 12 and 18 machines and, finally, were compared with the existing procedures in the literature. The experimental tables and related analysis performed via the solution methods by applying GA, SA and HA revealed that random-search-based modeling of FLP considering P/D and its solution as suggested in this paper is worth pursuing.