An integer L-shaped algorithm for the integrated location and network restoration problem in disaster relief

Being prepared for potential disaster scenarios enables government agencies and humanitarian organizations to respond effectively once the disaster hits. In the literature, the two-stage stochastic programming models are commonly employed to develop preparedness plans before anticipated disasters. These models can be very difficult to solve as the complexity increases by several sources of uncertainty and interdependent decisions. In this study, we propose an integer L-shaped algorithm to solve the integrated location and network restoration model, which is a two-stage stochastic programming model determining the number and locations of the emergency response facilities and restoration resources under uncertainty. Our algorithm accommodates the second-stage binary decision variables which are required to indicate undamaged and restored roads of the network that can be used for relief distribution. Our computational results show that our algorithm outperforms CPLEX for the larger number of disaster scenarios as the solution time of our algorithm increases only linearly as the number of scenarios increases