Encoding Structures and Operators Used in Facility Layout Problems with Genetic Algorithms

Abstract

The allocation of facilities in a plant layout is a complex problem. For solving it, many authors have used Genetic Algorithms (GAs) with the objective of reaching an efficient plant layout design. To represent the plant layout design as a data structure, GAs require a defined encoding scheme. Such a structure defines the types of solutions that can be obtained, and influences the GA´s ability to find good solutions. There are a few surveys on facility layout problems, but they have not addressed evolutionary issues in depth. This work presents a review that focuses on encoding schemes and related operators used in GAs, and suggests a method of classifying the different encoding structures described in the bibliography. We also studied their main characteristics and objectives; and successfully identified the crossover and mutation operators that could be utilized depending on the type of encoding scheme.