Comparison of annual railway freight turnover and automobile transport of the Russian Federation: modeling, verification, point and interval forecast

Abstract

Purpose: a study was conducted comparing the freight turnover of rail and road transport in the Russian Federation over the last twenty years and a forecast assessment was made. The comparative assessment was carried out due to the excess of freight turnover by rail over road transport. The study includes the following main stages: modeling based on the observed annual values of cargo turnover; model verification by comparing model and experimental values in years not involved in the construction of the model; model point and interval forecasting; analysis of the forecast trend of changes. The mathematical model of the excess of freight turnover by rail over road transport from the year of cargo delivery is based on a sample of statistical data from Rosstat from 2004 to 2021. The model was verified based on the results of cargo turnover in 2022 and 2023. Methods: to build a mathematical model, test it and carry out forecasting, a number of theoretical provisions of econometrics were used, including dynamic series, regression analysis, the least squares method, point and interval forecasting. The applied calculation apparatus was used using the Excel PPP. Results: the quality of the constructed model was evaluated by the following indicators: approximation errors, correlation and determination coefficients, and the Fisher criterion. A comparison of the observed and model values of excess cargo turnover showed that in the period from 2004 to 2021, according to the above indicators, they have minor deviations. The verification performed confirmed the satisfactory simulation. Practical significance: the conducted research has shown that it is possible to predict the results of transport operations according to the described algorithm and it is justified to make management decisions at the beginning of their implementation. Recommendations: to continue improving the modeling and forecasting apparatus, taking into account the assessment of risk indicators.