Integrating LOPCOW-DOBI method and possibilistic programming for two-stage decision making in resilient food supply chain network

Abstract

This study focuses on resilient supplier selection and order allocation — two crucial aspects of modern supply chain management (SCM). Globalization and strategic sourcing expose supply chains to disruptions, making resilient sourcing strategies essential for adapting to fluctuations in supply and demand. This paper proposes a novel integrated hybrid model that combines multi-attribute decision-making (MADM) with a possibilistic multi-objective programming model (PMOPM) to enhance decision-making in supply chain resilience (SCR). In the first stage, we present an MADM model that integrates the LOgarithmic Percentage Change-driven Objective Weighting (LOPCOW) and DOmbi Bonferroni (DOBI) methods. The LOPCOW-DOBI method enables decision-makers (e.g., purchasing team) to evaluate and rank multiple suppliers based on normal business criteria ($NBC$) and resilient pillars ($RPs$). In the second stage, a PMOPM is employed to determine optimal order allocations when supply, demand, and cost parameters are fuzzy in nature. The quantitative and qualitative evaluation of decision-makers’ opinions is integrated into mathematical optimization by combining the MADM output with PMOPM. Using the $ϵ$-constraint method, the model was optimized to obtain Pareto solutions, with the final solution identified via the global criteria method. A real-world food industry case study validated the MADM-PMOPM model. Our results show that suppliers with higher resilience performance receive larger orders. Sensitivity analysis confirms that the two-stage model consistently delivers stable and adaptable solutions. Comparative analysis further demonstrates that the proposed approach is effective and reliable for rational decision-making.