The Research On Facility Layout Problem Based On Multi-Objective Genetic Algorithm And Differential Evolution

In recent years, with the rapid development of social economy, it becomes more and more important for manufacturers to improve the production efficiency, reduce production cost, and improve enterprise’s competitive advantage, win the market advantages. The facility layout problem is to allocate N machines within a specified space, which can be optimized to achieve the goal of improving production efficiency. The extend double row layout problem is to allocate N machines on two rows separated by a straight aisle with fixed width. This problem involves two objectives of material handling cost and layout area and requires determination of not only machine sequence on both rows (relative location), but also exact machine location (absolute location). This problem is hard to solve since it includes both combinatorial and continuous aspects and involves two objectives that may conflict with each other.In this paper, a methodology is proposed to create a set of Pareto solution for this problem. First, a multi-objective genetic algorithm (MGA) is devised to obtain a set of non-dominated machine sequences. Secondly, a multi-objective differential evolution (MDE) is suggested to produce Pareto solutions for each non-dominated machine sequence. Finally, finial Pareto solutions are created from those Pareto solutions produced by MDE. This methodology is tested by a number of problem instances and compared with an exact approach (CPLEX), Experimental results show that our approach outperforms the exact one for problems of more than toy-size both in terms of solution quality and computational time.