Development of a robust bi-objective model for Closed Loop Supply Chain Network

Closed Loop Supply Chain (CLSC) is very influential in improving the company’s reputation and business competitiveness. One of the most critical strategic decisions in CLSC planning is network design. Decision makers and supply chain planners are struggling with conflicting objectives with uncertain and ambiguous data. Furthermore, uncertainty results in infeasibility and most of the methods applied for dealing with uncertainty are unable to incorporate this infeasibility and only focus on the improvement of the obejective value. Therefore, decision makers need to apply methods which enable them to strike a balance between model feasibility and solution optimality. In order to address this issue, a bi-objective model applying robust optimization in CLSC network design is proposed in this research work. This model minimizes total cost and total environmental impact of the supply chain while defining the location of facilities and the quantities of products transported among different facilities, quantity of products to be produced, total cost and total environmental impact of different configurations. Measuring environmental impact of the supply chain is implemented using a method based on LCA (Life Cycle Assessment), i.e., ReCipe 2008. The data used regarding environmental impact scores were derived from ECO-it software. The augmented ε-constraint method is used to solve the bi-objective model. To be more precise, this study gives an insight to managers in striking a balance between economic and environmental aspects of the supply chain. The results of this research were able to define the optimum network design with minimum cost and environmental impact under uncertain demand of customers. The proposed robust model has been validated by obtaining some basic data from a case study conducted by various authors in a pulp and paper industry in Europe. However, further relevant information required for the model was obtained from various sources. The efficiency of the robust model was then verified by a comparison with the equivalent deterministic model using two performance measures: mean value to validate the solution quality and standard deviation. Computational results of the model show that robust solution reduces mean and standard deviation of total cost and total environmental impact comparing to deterministic model.

Preview Text FK 2014 142IR.pdf Download (1MB) | Preview

Actions (login required)

View Item