A META-HEURISTIC ALGORITHM BASED ON MODIFIED GLOBAL FIREFLY OPTIMIZATION: IN SUPPLY CHAIN NETWORKS WITH DEMAND UNCERTAINTY

Nowadays, many challenges affect global supply chain networks including disruptions, delays, and failures during shipment of products. These challenges also incur penalty costs due to customers’ unmet demands and failures in supply. In this dissertation, the model was developed as a multi-objective supply chain network under two risk factors including failure in supply and unmet demand based on three different scenarios. The objective of scenario I was to minimize the total expected transportation costs between stages for each supply chain and penalty costs associated with shortage of products. Supply chain with no failure in supply will communicate with supply chain with failure to deliver its product to the final customer. For scenario II, the objective was to maximize the profits of the supply chain that face extra inventory. This supply chain with surplus products will collaborate with supply chains with shortage of products to prevent any undesirable costs associated with extra inventory. The objective of scenario III was to develop a multi-objective function, which maximizes the profit and minimizes the total costs associated with production, holding, and penalties due to supplier failure of raw materials. Once a supply chain faces failure in supply of raw materials, other supply chains with no supply failure will collaborate to prevent any associated costs. This research investigates the applicability of the Modified Firefly Algorithm for a multi-stage supply chain network consisting of suppliers, manufacturers, storages, and markets under risks of failure. Commercial software cannot obtain the optimal results for these problems considered in this research. To achieve better findings, we applied a Modified Firefly Algorithm to solve the problem. Two case studies for a pipe and a steel manufacturing integrated supply chain demonstrated the efficiency of the model and the solutions obtained by the Firefly Algorithm. We used four optimization algorithms in ModeFRONTIER and MATLAB software to test the efficiency of the proposed algorithm. The results revealed that when compared with other four optimization algorithms, Firefly Algorithm can help achieve maximum profits and minimizing the total expected costs of supply chain networks.