An intelligent system for constructing metro train driver working schedules

Abstract

Aim. The paper aims to tests and analyse the results of application of an intelligent system for constructing metro train driver working schedules that is intended for automatically improving the efficiency of utilisation of the working time of metro train drivers. In this case, depending on the task at hand and the chosen criterion, efficiency is understood as a reduction of the number of drivers involved in the implementation of the specified standard train schedule, improved uniformity of alternation of driver work and rest periods, as well as reduction of the duration of down time within work shifts.Methods. The study presented in the paper uses the graph theory, recursive and heuristic algorithms.Results. The authors have developed recursive algorithms for preparing work schedules for main metro train drivers and substitute drivers that operate during main drivers’ lunch breaks. The developed algorithms are used in the intelligent system for metro train driver work schedule planning. The algorithm for generating main driver work schedule includes driver allocation to a number of work lines preliminarily defined based on the rolling stock operation schedule for the purpose of ensuring traffic according to the planned metro train schedule. The algorithm for generating substitute driver work schedule involves substitute driver work time planning based on the possibility of arranging lunch breaks for main drivers.Conclusion. The paper presents the results of a trial of the developed intelligent system for train driver work planning for the Moscow Metro using the example of the Zamoskvoretskoye Depot of the Zamoskvoretskaya Line, the Vykhino Depot of the Tagansko-Krasnopresnenskaya Line, as well as the results of the system’s adaptation to the Moscow Central Circle. It also presents a comparative analysis of driver work schedules, i.e., the actual one and one obtained using the developed intelligent system. The application of the developed system may enable as much as a 28% improvement of the efficiency of metro train drivers’ work hour use.