Predicting the occurrences of earthquakes is difficult, but because they often bring huge catastrophes, it is necessary to launch relief logistics campaigns soon after they occur. This paper proposes a stochastic optimization model for post-disaster relief logistics to guide the strategic planning with respect to the locations of temporary facilities, the mobilization levels of relief supplies, and the deployment of transportation assets with uncertainty on demands. In addition, delivery plans for relief supplies and evacuation plans for critical population have been developed for each scenario. Two objectives are featured in the proposed model: maximizing the expected minimal fill rate of affected areas, where the mismatching distribution among correlated relief demands is penalized, and minimizing the expected total cost. An approximate lexicographic approach is here used to transform the bi-objective stochastic programming model into a sequence of single objective stochastic programming models, and scenario-decomposition-based heuristic algorithms are furthermore developed to solve these transformed models. The feasibility of the proposed bi-objective stochastic model has been demonstrated empirically, and the effectiveness of the developed solution algorithms has also been evaluated and compared to that of commercial mixed-integer optimization software. 1. Introduction Earthquakes are a special type of natural disaster that can result in huge catastrophes. The Great Sichuan Earthquake, which occurred on May 12, 2008, in Wenchuan County, Sichuan Province, China, was one such case. It had a recorded magnitude of 8.0 and imposed tremendous suffering on the local residents, causing over 69,000 deaths, 18,000 missing persons cases, 374,000 injuries, and huge loss of property. Although thousands of networked seismograph stations have been installed around the world and although these stations use powerful computers, it is still difficult to predict when and where an earthquake will strike. Therefore, an effective and necessary approach coping with an earthquake disaster is to plan and implement a disaster relief campaign that would reduce the damage right after the earthquake disaster takes place. Emergency logistics plays a vital role in disaster relief campaigns. However, the planning and implementation of relief logistics are challenging, especially when help is needed in mountainous areas, where the transportation infrastructure, such as roads, bridges and railway lines, can suffer severe damage after large earthquakes. The following three factors
M.-S. Chang, Y.-L. Tseng, and J.-W. Chen, “A scenario planning approach for the flood emergency logistics preparation problem under uncertainty,” Transportation Research E, vol. 43, no. 6, pp. 737–754, 2007.
H. O. Mete and Z. B. Zabinsky, “Stochastic optimization of medical supply location and distribution in disaster management,” International Journal of Production Economics, vol. 126, no. 1, pp. 76–84, 2010.
H. D. Sherali and B. M. P. Fraticelli, “A modification of Benders' decomposition algorithm for discrete subproblems: an approach for stochastic programs with integer recourse,” Journal of Global Optimization, vol. 22, no. 1–4, pp. 319–342, 2002.
X. Prats, V. Puig, J. Quevedo, and F. Nejjari, “Multi-objective optimisation for aircraft departure trajectories minimising noise annoyance,” Transportation Research C, vol. 18, no. 6, pp. 975–989, 2010.