Resumen
This paper presents the multimodal transport model done for Feasibility Study Update of Railway Corridor ?ilina ? Ko?ice ? Cierna nad Tisou ? state border, in the Slovak Republic. Model includes entire area of Slovakia in different levels of detail and also border and remote European external zones. The data of freight preference in Slovakia we obtained from adaptive state preference survey on main freighters in Slovakia. Each of participants was asked to deal with maximally three logistic tasks. The entire freight task was defined by transported commodity, length of transport path and transport mode used in the present. For each of the tasks respondent passed adaptive state preference experiment where mode, cost, travel time and reliability were attributes of alternatives. 51 freighters participated on the survey and they were asked about 71 freight tasks. We find out that railroad or road mode was not preferred significantly but freighters are resistant to change routine freight traffic mode. In general, specific constant of present usage mode was positive and almost thirty times larger than value of time. This is complication for standard modal shift traffic model because alternative specific constant could not be assigned to specific traffic mode equation. Moreover, actual data of freight transport in Slovakia have different level of accuracy. Very accurate and detailed information (commodity and values in origin-destination pairs) were available for railway transport. On the other hand, there was lack of information about road freight transport. This difficulty leads us to build double modal split model for road and railway freight transport. Trip generation process is comprised of two components which represents two (railway and road) currently used modes. First part of trip generation process contains matrix of volumes of freight currently transported by rail. The second part of the process estimate origin-destination matrix by trip production and attraction model and represented road freight transport. This model is based on knowledge about location of big companies and population density. The two products of trip generation processes enter modal split model which is doubled again. One of the modal split models is for freighters who currently use railroad and second of them is model for current road freight users. Each of the two models prefers? own traffic mode. Four origin-destination matrices are product of the entire modal split process. One pair of the OD matrices is for railway freight transport and the other pair is for road freight transport. Corresponding matrices are put together for the next step of the modelling ? assignment of mode specific freight demand matrices to model transport network, by the generalized costs method. In the part of transport model supply, freight trains and their attributes (type, capacity, speed, and time intervals) were put in the model. The model can simulate both direct cargo transport from origins to destinations and cargo movement by trucks, manipulation trains and relational trains in order to optimize transport of carried goods.