Revistes Catalanes amb Accés Obert (RACO)

A fuzzy logic approach to assembly line

Daniel J. Fonseca, Matthew Elam, Charles L. Karr, C.L. Guest


This paper deals with the use of fuzzy set theory as a viable alternative
method for modelling and solving the stochastic assembly line balancing
problem. Variability and uncertainty in the assembly line balancing problem
has traditionally been modelled through the use of statistical distributions.
This may not be feasible in cases where no historical data exists. Fuzzy set
theory allows for the consideration of the ambiguity involved in assigning
processing and cycle times and the uncertainty contained within such time
variables. Two widely used line balancing methods, the COMSOAL and
Ranked Positional Weighting Technique, were modified to solve the balancing
problem with a fuzzy representation of the time variables. The paper shows
that the new fuzzy methods are capable of producing solutions similar to, and
in some cases better than, those reached by the traditional methods.

Text complet: HTML