Abstract:
In a competitive and maintainability context, each company looks for optimizing its supply chain in order to satisfy its customers by providing the best quality products in the best delays and with the lost costs. In this sense, we are interested in a single- commodity stochastic supply chain design problem. Our supply chain is composed of suppliers and retailers. The objective is to find the best location for distribution centres (DCs) and to serve retailers from suppliers through DCs in a random supply lead time. We presented a non-linear optimization model that integrates the selection of suppliers, the location of DCs, and the retailers’ allocation decisions with an oriented cost function to minimize. Note that the determination of exact solutions to this problem is a NP-hard problem. Accordingly, we proposed an optimization approach using three different metaheuristics: genetic algorithm, simulated annealing, and taboo search to solve this problem and find the best supply chain structure (location of DCs, allocation of suppliers to DCs and DCs to retailers). Computational results are presented and compared to evaluate the efficiency of the proposed approaches.
Machine summary:
We presented a non-linear optimization model that integrates the selection of suppliers, the location of DCs, and the retailers’ allocation decisions with an oriented cost function to minimize.
Accordingly, we proposed an optimization approach using three different metaheuristics: genetic algorithm, simulated annealing, and taboo search to solve this problem and find the best supply chain structure (location of DCs, allocation of suppliers to DCs and DCs to retailers).
The problem resolution allows us to determine the decision variables Xj, Yij ,and (View the image of this page)The objective function (1) minimizes total cost which represented the sum locating facilities with fixed cost, the costs of shipment, and transportation from DCs to retailers and from suppliers to DCs. By assuming that each DC uses an EOQ policy, the total inventory costs at the DCs are represented by the third term.
Table 1 illustrates results the supply chain structure (located DCs and selected suppliers) and the global generated cost obtained for six different case simulations for each instance of the studied problem.
We have proposed three metaheuristics-based optimization approaches to solve a stochastic distribution network problem design by integrating the strategic decisions of supplier selection, location of DCs, and retailer allocation.