Hierarchical Supplier Selection Optimization with Multiple Items in Large-Scale Construction Projects

Managers of modern large-scale construction projects are under pressure to meet higher customer expectations with tighter budgets. Although they deal with numerous issues in the purchasing and manufacturing processes, selecting effective and efficient material suppliers is among the most critical one. This selection is often very challenging when lacking precise information. Moreover, the construction contractor and the suppliers often have conflicting interests and make decisions individually. As research concerning the aforementioned issues is still relatively scarce, this paper proposes a multiobjective bilevel programming model with random fuzzy coefficients for supplier selection problem with multiple items (SSP-MI) in a large-scale construction project. The upper level problem deals with the construction contractor who selects suppliers to minimize total cost, maximize service, and item quality. The lower level problem deals with the suppliers who allocate supplied items to maximize their own total profit. For solving this complex bilevel nonlinear model with uncertainties an expected value operator method is first used to deal with the uncertain variables, and then Karush-Kuhn-Tucker (KKT) conditions and a combinatorial algorithm with a sectional genetic algorithm with fuzzy logic controller (flc-SGA) and a weighted-sum method (WSM) based on satisfactory degree (SD) denoted as flc-SGA with SD-based WSM are proposed. Finally, the proposed approach is demonstrated to be effective when carried out in the Pubugou Hydropower construction project.