علي ثنائي ايلخچي

With the emergence of the information era and the concurrent advancements in Information Technology, along with the recognition of the human potential in utilizing data, we are currently observing substantial expansion in the generation, documentation, and application of scientific data across diverse fields and industries. The rail transport industry accounting for approximately 30% of cargo transportation in the world, encompasses a wide range of sectors and possesses extensive volumes of data, which can be harnessed to improve performance, broaden operational capabilities, and enhance managerial decision-making capabilities. In comparison to other modes of transportation, rail transport offers notable competitive advantages such as high safety of cargo transportation, the capacity to handle massive volumes of goods, etc. Consequently, the development and investment in the rail industry has a high rank among the foremost strategic priorities for governments and Numerous studies have been conducted to eva‎luate the various dimensions and impacts associated with the modal shift from road to rail transport for cargo transportation. After that, this project has aimed to initially define various scenarios in the first phase of the research (for incorporating assumptions and different situations), accompanied by examining the statistical and recorded data related to intercity road and rail transportation and waybills in Iran for previous years to discover and identify suitable loads (rail-friendly loads) for rail transportation (by taking into account the important characteristics of the loads and utilizing data mining methods such as K-mean Clustering method). After extracting the share of this type of load, in the second phase of the research, various statistical, data mining, and machine learning methods (such as ARIMA, ANN, etc) have been implemented and the best methods have been selected according to the results of Appropriate measures of the accuracy of the forecasts for each forecasting scenario. Also, in the third phase, by using the selected prediction models from the previous phase, the volume of the demand for rail and road cargo transportation has been forecasted for the next two, five, and ten years, and the share of rail-friendly loads from the total land loads has been determined. The implementation results show that the total share of rail-friendly loads for two-year, five-year, and ten-year time horizons is equal to 20.7%, 20.1%, and 20.5%, respectively, for the rail freight net tonnage index and the net ton-kilometer index of transported freight equal to 32.1%. 32.2% and 31.6% are obtained.

پيش‌بيني تقاضا , حمل‌ونقل ريلي , حمل‌ونقل جاده‌اي , بار ريل‌پسند , تغيير شيوه حمل , داده‌كاوي

Demand forecasting , Rail transport , Road transport , Rail-friendly cargo , Modal shift , Data-mining

Ali Sanaei Ilkhechi

Mehdi Ghazanfari