МAТЕМАТИЧЕСКОЕ МОДЕЛИРОВАНИЕ РЕГИОНАЛЬНЫХ ГРУЗО- И ПАССАЖИРОПОТОКОВ
Mathematical Modelling of Regional Cargo and Passenger Flows

Summary/Abstract: The creation and implementation of the strategies for economic and social development in the Russian regions for the period up to 2035 implies an adequate development of transport services affecting all economic sectors and segments of the population. In this regard, we propose a model connecting the characteristics of passenger and cargo flows with the parameters of economic and social development, as well as with the region’s demography. This model allows specifying the congestion of the transport system resulting from the implementation of plans for social and economic development and planned decisions in the sphere of economic activity. For developing the model, we selected parameters describing the economic situation, labour market, demography, living standards and social situation in the analysed subject. These parameters have the highest correlation coefficients with the analysed characteristics of the transport infrastructure. Further, we conducted a step-by-step regression analysis, adding to the already existing variables new ones that gave the greatest increase in the determinacy coefficient R2. The model shows that the main factor determining the amount of passengers transported by public buses is the annual average number of employed persons. The passenger turnover is mostly affected by the population size. The volume of goods transported by trucks is determined by parameters characterising the level of the production development (investments in fixed assets, fixed capital in the economy, and the volume of shipped goods of domestic production). The use of nonlinear models and networks did not significantly reduce the model’s errors. Additionally, we clustered the Russian regions by indicators of socio-economic development and the characteristics of transport infrastructure affecting traffic flows. Then we assessed the efficiency of transport infrastructure’s exploitation in various clusters. This allows the targeted benchmarking, namely the selection of regions mostly appropriate for comparison with the analysed one.

• Details
• Contents