Modeling and optimization of facility layout problem using nature inspired algorithm

Abstract

Productivity and efficiency of an organization greatly depends on how people plan, organize and utilize the facilities in that organization. From an upfront investment and recurring project expense, facilities planning are a critical issue in today’s competitive manufacturing and service sectors. In addition to the upfront investment involved in facilities planning, there are operational issues that make facilities planning a critical issue. The most obvious impact is on material handling expenses. The impact of the facility layout goes beyond material handling costs. An effective facility layout implies that departments with high flow are close together. In facility layout problems, objective functions are modeled with different objectives in mind examples of which include minimization of cost or flow of materials, maximizations of closeness rewards etc. In this thesis, a mathematical model with a continuous representation of distance based adjacency matrix is developed. The resulting exact model will consider every all-rectangular-department solution. Solution from the new model is compared with solutions found from models based on binary based adjacency matrix. In this thesis, exact algorithm is used for finding feasible solution set from the total solution space. Further research can be done using other heuristic algorithms. In the function [adjacency=1/k1*e(k2*x)] proposed in this thesis for generating continuous value adjacency matrix has two co-efficient, namely k1(Denominator co-efficient) and k2 (exponential co-efficient). Unit value for both of the co-efficient was assumed even of the fact that, there are strong rationales behind these two having industry specific values. There is huge scope of econometrical research to come up with series of values of k1 and k2 for different industries.