Traffic simulation models allow the evaluation of road networks and are used as a fundamental tool for traffic management, including parking management. Traffic simulation is a mature field with several decades of research. However, most existing studies focus on traffic simulation under normal traffic conditions. Traffic simulation of mixed networks in conditions close to exceeding capacity or in emergency cases is still a challenge. This research focuses on the development of solutions for the management of large-scale parking facilities and depots (for either passenger vehicles or commercial fleets) under constraints including near-capacity demand, temporally concentrated arrivals/departures and need for emergency evacuation. An integrated methodological framework will be developed that will operationalize a cycle created from the following main methodological and technological challenges/ components, mainly microscopic parking facility simulation, vehicle localization support in indoor environment (where GPS devices are unavailable), information generation- dissemination and control and strategy generation for the optimal parking management schemes. For understanding of the needs that exist in the context of this research two experiments were conducted in the area of the National Technical University (NTUA) campus, in a similar environment to that in a large parking facility. The experimental setup is analyzed in detail and from the first analysis of the available data some conclusions are drawn.
|Tags||big data, machine learning, parking|