Comparative study of Constructive and Improvement Algorithms for Cell Formation with Alternative Routings

Abstract

This paper describes a comparison of genetic algorithm and state space search algorithm for simultaneous formation of manufacturing cells and selection of process plans for parts with alternative plans. Both algorithms perform two simultaneous decisions: selection of process plan for parts between several alternatives, and assignment of parts and required machines to manufacturing cells. For genetic algorithm, a chromosome represents an assignment of machines to cells, and fitness function defined as number of inter-cell transfer is used to select the part route alternative with minimal inter-cell movements. The space search algorithm is based on classification and ordering of machines and application of space search to recursively assign the most critical machines and parts to cells. Both algorithms have been studied with several decision making criteria: intercellular traffic, size of the machine cells, ratio of state space expansion, and so on. The results presented in the paper include inter-cell transfers and the cell size. Both algorithms have been applied to the same set of cell formation problems with alternate routings from literature and representative results are shown. Performance of both algorithms is evaluated with respect to the quality of the obtained solution (as compared with different algorithms on the same problem from literature). The conclusions from this experiment are presented at the end of the paper.