Cellular Manufacturing through Composite Part Formation: A Genetic Algorithm Approach

Maintenance of cell efficiency in Cellular Manufacturing System is important. Once the cell is formed, new parts come naturally for processing. These parts may not be completely processed in a single manufacturing group and need different machines for processing. This paper presents a procedure for allocating new parts to machine groups using the concept of composite parts. Initially for each of the part families, the hypothetical composite parts are prepared from historical data and coded in terms of alphabetical string. The method starts with identifying the specific part family from the drawing of the new part. Primitive libraries for surface elements are prepared from historic data for identification of processes of the new components. Combinations of these surface elements make the features of the parts. The strings of all new coming components are compared with hypothetical composite parts that represents the part family. The optimization criterion for proposed method is taken as minimization of summation of Levenshtein distance between composite strings and the set of new part strings that reach for processing in the dedicated manufacturing groups. The model developed by the authors can be utilized to guide planners to operate the manufacturing cells efficiently.