Optimizing community post-earthquake emergency medical capability via mobile hospital configuration

Emergency medical capability is critical for communities to reduce casualties after earthquakes. In this research, a bi-level design optimization model for the configuration of mobile hospitals is developed to minimize both the post-earthquake transfer time of casualties from their residential areas to the mobile hospitals, and the within-hospital service time including mean queue waiting time and average processing time. The upper level of the model characterizes community officials’ decisions regarding possible mobile hospital configuration that consists of three decision variables: the location of mobile hospitals on candidate sites, their functional classes, and the number of attending physicians dispatched to them. The lower level model leverages the mixed user equilibrium traffic assignment to simulate the path selection behaviour of drivers transferring casualties with different injury heterogeneity under different information perceptions. Meanwhile, considering the uncertainty in the number of casualties and the post-earthquake damage state of bridges, a robust configuration model for the stochastic case is also established based on the robust optimization theory. To tackle the optimization model, an interactive solution approach combining the genetic algorithm and the modified mixed equilibrium assignment algorithm is presented. A demonstrative study is conducted on a middle-class city located in an earthquake-prone area.