The aperiodic facility layout problem with time-varying demands and an optimal master-slave solution approach

Yiyong Xiao, Yue Zhang, Sadan Kulturel-Konak, Abdullah Konak, Yuchun Xu, Shenghan Zhou

Research output: Contribution to journalArticlepeer-review

2 Scopus citations


In many seasonal industries, customer demands are constantly changing over time, and accordingly the facility layout should be re-optimized in a timely manner to adapt to changing material handling patterns among manufacturing departments. This paper investigates the aperiodic facility layout problem (AFLP) that involves arranging facilities layout and re-layout aperiodically in a dynamic manufacturing environment during a given planning horizon. The AFLP is decomposed into a master problem and a combination set of static facility layout problems (FLPs, the slave problems) without loss of optimality, and all problems are formulated as mixed-integer linear programming (MILP) models that can be solved by MIP solvers for small-sized problems. An exact backward dynamic programming (BDP) algorithm with a computational complexity of O(n 2) is developed for the master problem, and an improved linear programming based problem evolution algorithm (PEA-LP) is developed for the traditional static FLP. Computational experiments are conducted on two new problems and twelve well-known benchmark problems from the literature, and the experimental results show that the proposed solution approach is promising for solving the AFLP with practical sizes of problem instances. In addition, the improved PEA-LP found new best solutions for five benchmark problems.

Original languageEnglish (US)
Pages (from-to)5216-5235
Number of pages20
JournalInternational Journal of Production Research
Issue number17
StatePublished - 2021

All Science Journal Classification (ASJC) codes

  • Strategy and Management
  • Management Science and Operations Research
  • Industrial and Manufacturing Engineering


Dive into the research topics of 'The aperiodic facility layout problem with time-varying demands and an optimal master-slave solution approach'. Together they form a unique fingerprint.

Cite this