A rapid transit system is called robust if it maintains its
functionality under perturbations. To be robust, strategies for producing
less vulnerable plans and procedures of recovery actions in case of
disruptions (including timetable adjustment, rolling stock re-scheduling
and crew re-scheduling) are used. This paper deals with the first task and
develops an effective plan for allocating fleet frequencies at stops along
a line based on three objectives: minimizing passenger overload, maximizing
passenger mobility and minimizing passenger loss. Schedules for
decongesting and recovering the line are determined by means of optimization
models. Heuristic approaches are discussed and computational
results for a real case study are provided