A sampling-based winner determination model and algorithm for logistics service procurement auctions under double...

Area of Science:

• Logistics and Supply Chain Management
• Operations Research
• Risk Management

Background:

• Fourth-party logistics (4PL) platforms face significant risks from demand and disruption uncertainty.
• Traditional winner determination models struggle to account for these dual uncertainties.
• Effective risk mitigation strategies are crucial for 4PL operational efficiency.

Purpose of the Study:

• To develop a robust winner determination model for 4PL platforms under demand and disruption uncertainty.
• To propose a hybrid risk mitigation strategy integrating temporary outsourcing and fortification.
• To minimize total operational costs while hedging against identified risks.

Main Methods:

• Constructed a two-stage stochastic winner determination model.
• Transformed the model into a mixed-integer linear programming problem using an improved sample average approximation (SAA) algorithm.
• Developed a sampling-based heuristic algorithm combining dual decomposition, Lagrangian relaxation, and scenario reduction.

Main Results:

• The proposed heuristic algorithm demonstrated superior performance compared to CPLEX in solving complex scenarios.
• Numerical examples and a real-world case validated the model's and algorithm's effectiveness.
• Sensitivity analysis confirmed the significant impact of demand fluctuations and disruption probability on strategy selection.

Conclusions:

• The hybrid mitigation strategy effectively hedges against double uncertainty in 4PL winner determination.
• The developed SAA and heuristic algorithms provide efficient solutions for complex stochastic optimization problems.
• Findings offer valuable insights for 4PL platforms in optimizing risk management and operational costs.