그룹 테크놀러지에서의 기계 및 부품군을 형성하기 위한 발견적 해법

A heuristic algorithm for forming machine cells and part families in group technology

A similarity coefficient based algorithm is proposed to solve the machine cells and part families formation problem in group technology. Similarity coefficients are newly designed from the machine-part incidence matrix. Machine cells are formed using a recurrent neural network in which the similarity coefficients are used as connection weights between processing units. Then parts are assigned to complete the cell composition. The proposed algorithm is applied to 30 different kinds of problems appeared in the literature. The results are compared to those by the GRAFICS algorithm in terms of the grouping efficiency and efficacy.