This paper investigates solving the transportation problem with fuzzy demands and fuzzy supplies using a two-stage genetic algorithm (GA). At the ﬁrst stage, we simulate a fuzzy number by distributing a fuzzy value into certain partition points. We then use GA to evolve the values in each partition point and the ﬁnal values represent the membership grade of that fuzzy number. As a result, we obtain the estimated values of all fuzzy demands and fuzzy supplies and the original fuzzy problem becomes a defuzziﬁed instance. The best solution to the defuzzified instance is then solved by the following stage via evolution process. The experimental results show that the proposed two-stage GA approach outperforms the other fuzzy approach to solving the transportation problem with fuzzy demands and fuzzy supplies.
International Journal of Innovative Computing, Information and Control 5(12)pt.B, pp.4775-4785