Adaptive NEURO-Fuzzy Inference System (ANFIS) Based Worker Assignment Model For Large Sized Virtual Cells
In this paper, the author aims at developing an optimized worker assignment model for large sized virtual
manufacturing cells by using adaptive neuro-fuzzy inference system (ANFIS). Improved productivity and superior quality in
operations with maximum utilization of existing resources are always the primary objectives of any manufacturing
organization. Many manufacturing philosophies have been developed to achieve the above objectives and Virtual Cellular
Manufacturing System (VCMS), a logical extension of Cellular Manufacturing System (CMS), is one of such philosophy
developed quite recently. Worker assignments are invariably a challenging task under any manufacturing environments.
Though there are many techniques and algorithms developed, tested for worker assignment tasks, application of ANFIS into
this task is an innovation and novel which has enough potential to be used as a tool. ANFIS, functionally equivalent to fuzzy
inference system, is a convenient way to map an input dataset to an output dataset. In this paper, datasets corresponding to three
cell configuration problems under VCMS environment from literature are implemented in the ANFIS model as input and
output datasets. Results of worker assignments obtained from the present ANFIS model and the previous published models are
then compared, analysed and discussed. The study and results obtained affirm that ANFIS also shows prominence and promise
in solving problems related to workforce assignment into virtual cells of higher configurations.
Index Terms- worker assignment, adaptive neuro-fuzzy inference system, worker fitness attributes, fuzzy logic.