Evolutionary algorithm approach in unequal area layout planning problem

Placement of components on a layout based on a priori determined criteria and limitations is a well-known problem in facility layout planning, circuit design, or machine placement areas. Most layout problems are NP-Hard, and they cannot be solved optimally. The main objective in layout problems is to determine the relative positions of departments that minimize the flow cost among them. This study focuses on Unequal Area Facility Layout planning (UA-FLP), one of the most challenging areas in facility layout research. The problem is solved as a static flow problem, assuming that flows between departments are not changing over time. As a first step of the study, mathematical models used previously to solve linear and nonlinear problems are thoroughly examined. Secondly, heuristic and meta-heuristic algorithms have been investigated since mathematical models are not capable of solving problems with a large number of departments. As a result, a new evolutionary algorithm has been developed. Parameters of the algorithm were selected using the Taguchi experimental method. The algorithm was tested on problem sets that are also frequently used in the literature, and its performance was analyzed. The proposed approach provides alternative layout configurations under a single framework. Contrary to the previous works that focus on a single configuration, the proposed evolutionary algorithm approach generates three alternative solutions, each with a separate layout configuration. Three alternatives considered include horizontal and vertical FBSs (Flexible Bay Structure) and STS (Slicing Tree Structure).