Improved Genetic Algorithm for Workplace Facility Layout

于瑞峰,王永县,彭海

DOI: https://doi.org/10.3321/j.issn:1000-0054.2003.10.015

2003-01-01

Abstract:Present genetic algorithms cannot effectively model the workplace facility layout problem (FLP). An improved genetic algorithm was developed based on a new encoding method and mutation operator. The algorithm provides better representation of the facility layout, enhances the efficiency of the calculation, and dramatically reduces the number of illegal offsprings generated during the evolutionary process. In addition, the algorithm dynamically decides the number of rows for the workplace facilities, which removes the restriction of onedimensional optimization in old algorithms. The algorithm optimizes workplace layouts in two dimensions. The effectiveness of the proposed algorithm was evaluated with a typical layout problem. The results show that the algorithm has good global searching ability and rapid convergence. 