An Intelligent Fuzzy Multi-Agent System for Reduction of Bullwhip Effect in Supply Chains

Abstract

This paper presents a Multi-Agent System (MAS) for reduction of the bullwhip effect in fuzzy supply chains. First, it is shown that, even using an optimal ordering policy, without data sharing the bullwhip effect still exists in the supply chain. Then a multi-agent system is proposed to manage the bullwhip effect. The multi-agent system has four different types of agents. The multi-agent system applies Tabu Search algorithm for fuzzy rules generation and a new data filtering method for extraction of training and testing data from the supply chain data warehouse. The results show that the proposed MAS is capable of managing the bullwhip effect efficiently.