Scenario dominance for acceleration of robust service system design method

This paper deals with robust emergency system designing. Usually, the problem of optimal resource deployment under standard conditions is formulated as the weighted p-median problem. The weights of the distances are often proportional to population in the individual users' locations. The demand for system robustness brought constitution of detrimental scenarios on the scene. The scenarios comprise impact of possible fatal failures on time distances between facilities and the users' locations. Each scenario corresponds to a particular objective function, and thus searching for the robust service system design consists in minimizing the maximum of particular objective functions. The resulting problem is hard to be solved due to the model size and usage of min-max constraints, which link up the individual scenario objective functions with their upper bound representing the objective function of the robust problem. We focus here on handling big size of the original model by scenario dominance analysis to reduce the set of scenarios and to accelerate associated solving methods.