Scheduling Emergency Materials Using Road-Network Repair Time Sequence

Abstract

Natural disasters, such as earthquakes, can cause significant damage to road networks, leading to reduced rescue efficiency and hindrance to rescue operations. This study focuses on the time-sequence repair strategy for the damaged road network to optimize material scheduling and rescue efforts. The disaster time is divided into multiple cycles, considering the quantitative impact of road damage on passage time. A dynamic decision-making model is formulated to optimize the repair of the damaged road network and the material scheduling simultaneously. To tackle the complexity of the model, a combination of genetic algorithm and Benders decomposition algorithm is employed for solving. The effectiveness of the proposed algorithm is evaluated using a case study based on the Ya’an earthquake. The solution is compared with the performance of mixed-integer programming solvers DICOPT (discrete and continuous optimizer) and SBB (simple branch and bound) of GAMS software. The results indicate that the Benders decomposition algorithm achieves an average gap rate of 2.4% compared with DICOPT and SBB solvers when solving the problem with the objective of the weighted sum of the transport time and the volume of urgent materials. Furthermore, when dealing with the problem with the objective of the weighted sum of the transport time, the volume of urgent materials, and out-of-stock rate, DICOPT and SBB fail to find acceptable solutions within reasonable time frames. This demonstrates the efficacy of the Benders decomposition algorithm in addressing emergency-material-scheduling problems efficiently.