Robust
Yao et al. \cite{YaoEtal2009} deal with Dynamic traffic assignment in a disaster aftermath, in which nodes ensembles regions and arcs correspond the roads and the evacuation plan consists of a flow over time (evacuation of effected people) on the transportation network from source nodes (affected areas) to sink nodes (shelters). Under demand uncertainty, the transportation network is decomposed into cells and the problem is formulated as a deterministic cell transmission model (CTM) (for detailed info about CTM refer to \cite{Ziliaskopoulos2000}). For obtaining more realistic model in disaster context, the authors introduce a time and space dependent coefficient of threat level, hence defining the objective function as minimization of the evacuees’ total threat exposure. Since infeasibility cost in enormous in disaster context ( lack of proper evacuation may result in life loss), they reformulated the model as a robust optimization model, and by providing several theorems equivalent deterministic counterpart is obtained. Evaluation of the deterministic and robust model shows significant outperformance of robust model in term of solution feasibility and quality.