A Fast and Effective Greedy Heuristic for On-line Train Calendars Generation

This paper describes a new approach for train calendars textual generation, that is a heuristic algorithm designed to automatically generate a text to final customers, concisely and clearly, from a service calendar represented by a boolean vector as input. This niche problem belongs to the transportation field, specifically, to railway services. The new heuristic algorithm, developed in the C programming language, guarantees a constant computation time, between 0 and 16 ms. Tested on several real railway timetables, this new approach was extensively compared with existing mathematical programming models presented in Amorosi et al. (2019), with a significant reduction of computation times that makes it applicable in practical contexts.