Simulating Annual Long-Distance Travel Demand

Abstract

Research in the field of transport planning focuses today on urban mobility and, thus, on daily activities. Hence, most of the surveys, models and simulations aim to add valuable contributions to this topic. However, long-distance travel is growing, because travelling has become more accessible and cheaper recently. Nowadays, almost half of all vehicle miles travelled are generated by long-distance travel. Therefore, it is crucial to transport planners and policy makers. Nevertheless, so far only few surveys or models cover this field. One of the reasons is the challenging task to capture data describing long-distance travel. This thesis shows that big data can potentially help to overcome this burden using GSM data. Due to different challenges there has been also no effort to develop an agent-based simulation for long-distance travel. The major obstacle is the time horizon that such a simulation has to cover. Long-distance travel can not be analyzed focusing on a single day, because long-distance journeys usually consume more time and are also planned well in advance. This thesis introduces an agent-based simulation covering long-distance and long-term travel demand and, thereby, closes the research gap described above. Furthermore, the thesis demonstrates that the simulation presented can adequately simulate large scale scenarios in reasonable time. The main value of such a simulation is support of policy makers for big infrastructural investments such as new bridges, tunnels or airports.