Devising a method for the automated calculation of train formation plan by employing genetic algorithms

Abstract

EN: We devised a method for calculating the plan of formation of single-group freight trains, which is based on the use of genetic algorithms as the optimization method. Development of the method is predetermined by the need to improve accuracy in calculations and quality of making management decisions in the area of organization of railcar traffic under modern conditions. A mathematical model is constructed that uses accumulation parameters as stochastic variables. This will make it possible to find the most rational variant of the plan for train formation. The simulation we performed demonstrated effectiveness of the method devised. The method developed demonstrated improved accuracy of about 3 per cent relative to the classical analytical methods. The method makes it possible to consider the limitations on the throughput and processing capacity of technical stations and throughput capacity of the sections. This method uses parameters of accumulation as stochastic variable that makes it possible to find a more efficient variant of the plan for train formation. These capabilities allow us to consider a possibility of applying this method as a basis for building an integrated automated system for managing railcar traffic, which will bring together the tasks of strategic and operational planning at the new qualitative level. Creating such a system, in turn, might provide opportunities for strengthening the systems effect, reducing downtime of railcars, increasing profitability and competitiveness of freight rail transport.

UA: Розроблено метод розрахунку плану формування одногрупних вантажних поїздів, який базується на використанні генетичних алгоритмів. Метод демонструє не лише високу точність розрахунків, але і забезпечує можливість урахування обмежень по пропускній і переробній спроможності об’єктів залізничної інфраструктури. Застосування даного методу в сукупності з використанням можливостей сучасних обчислювальних систем дозволить вирішувати задачу розрахунку плану формування поїздів для всього полігону залізниць України.