Трансформация методологии планирования и прогнозирования перевозок грузов на железнодорожном транспорте

Abstract

Objective: To develop a system of approaches, methods, and models for planning and forecasting the railway freight traffic using big data acquisition and analysis. This is conditioned by the today’s era of fierce competition, global digitalization, and a big amount of chaotic and multi-directional information flows, where the effective progressive development of transport systems depends very heavily on the methodology used for planning and forecasting freight traffic. The reliability of such a methodology is determined by the quality of the studied data sample, the integrity of analytical methods and procedures for their processing, and effectiveness. Methods: In order to classify the methods of planning and forecasting freight traffic, we have analyzed and systematized the main indicators of the transportation process, the degree of formalization of the methods used, and the general principles of their operation and obtaining forecast information. Results: Logical and comparative analyzes, synthesis of forecasting models made it possible to propose a conceptual approach to the practical application of the econometric methods and forecasting models used in foreign studies and analyzed cases. Practical importance: A technical approach to the use and scaling of the polynomial logit (logistic regression), the ARMA model (ARIMA) and the HIST algorithm of the Computing Center of the Russian Academy of Sciences has been proposed for forecasting the freight traffic and digitalization of planned and predictive business processes in the railway transport of the Russian Federation