Network Design in Strategic Alliance under Uncertainty with a Trade-off between Risk and Performance

The purpose of this study is to present a new mathematical model to design a supply network by considering the strategic alliance and the relationships between the supply chain members under uncertainty. This study attempts to create a suitable decision-making environment for managers to optimize the network and make appropriate strategic decisions accordingly. Since the mathematical model of network design has computational complexity, providing a suitable solution method for the proposed model is another goal of this research.
As this paper is an applied study, a new mixed-integer linear programming model (MILP) has been presented. The robust optimization method has been used to deal with uncertainty risks such as the risk of changing sales and the return of products. In the mathematical model, strategic alliance levels and the level of risk for each of the partners are considered. The model has two objectives; minimizing the cost, and minimizing the risk of establishing a strategic alliance. The optimal location of the facilities, the selection of production methods, the capacity of facilities, the selection of colleagues, the level of strategic alliance, and their control level have been determined in the presented mathematical model. Considering the computational complexity of the mathematical model, The Benders decomposition method has been applied, and a solution for the proposed mathematical model has been developed and localized by using acceleration mechanisms.
The results show the effectiveness of strategic alliances in reducing the costs of the production and distribution network of goods. The quantification of the concepts related to the strategic alliance in the supply chain, and the establishment of a trade-off between the risk and benefits of the strategic alliance are other research findings. Considering the different levels of strategic alliance and risk for each partner, the results of the current research show that a strategic alliance reduces the cost, and this cost reduction depends on the risk level of the partners. In addition, the computational results show the efficiency of the accelerated Benders decomposition algorithm for solving mathematical models in large-scale problems. In some problems that the Gams software is not able to provide the right answer in the appropriate time, the algorithm based on benders methods provided acceptable answers in a shorter time frame.
Applying the industry data shows the effectiveness of the model in creating a decision-making environment for managers and decision-makers. Also, the results show the appropriate performance of the solution method. Therefore, the finding of this research indicates a new research viewpoint in the field of network design under strategic alliance for the production and distribution of products.