Authors:   Manfred Gronalt, Edith Schindlbacher

Abstract

This work presents an agent-based simulation model for the planning and evaluation of intermodal freight transportation networks with a system-wide scope. This approach covers actors, transportation infrastructure, terminal operation strategies and train services. The relevant actors in continental intermodal freight transport, their contributions and tasks are identified and their specific planning rules are as well as the dedicated coordination structures and the code of conducts between the agents are modelled in order to analyse the overall behaviour of the system. Each agent acts according to its own goals and rules and has to consider its domain specific restrictions. They have the ability to make autonomous decisions. These may be set actively or according to system?s restrictions or requests. Typical agents considered are terminal agents, link agents, train operator, route planner and container owner respectively. The developed class model structure is also shown. The developed agent-based simulation model enables the dynamic performance evaluation of the whole intermodal freight transport system, as well as on the level of individual elements. We define an arbitrary freight network with 15 terminals to supply 23 demand regions. Source-sink relations are generated from databases of joint research centres of the EU commission. Also we consider empty container depots in the network. Model trains and corresponding schedules for each terminal and load units are defined in order to generate a basic network load. The modelling approach is applied to standard processes like execution of regular transportation orders, balancing of empty container stocks and introduction or close down of transportation services. A deviation management is applied for booked load units which are too late, late arrivals of trains, insufficient terminal capacity (loading tracks, container storage) and availability of load units. We apply a scenario approach to observe system?s behaviour if we are changing transportation demand, modify terminal network or adopt train schedules. We can conclude that the agent-based simulation is an appropriate approach for evaluation the intermodal freight network.