Development of travel time functions for disrupted urban arterials with microscopic traffic simulation

Abstract

Urban traffic networks consisting of partially blocked roads often need to remain open to traffic before, during and after disasters because of their vital roles to hazard preparation, emergency response and recovery of urban communities. To conduct effective traffic planning of disrupted transportation networks highly depends on accurate prediction of travel time on partially blocked roads, which is very different from that on intact roads. Due to the lack of appropriate models, travel time prediction approaches developed for intact roads have often been directly applied to partially blocked roads, leading to inaccurate travel time estimates. Unrealistic travel time estimates of partially blocked roads as well as the whole transportation network further affect traffic planning, emergency response and other decision-makings which are heavily reliant on travel time prediction. A new approach to develop travel time functions for partially blocked roads in urban areas is proposed to close this gap based on microscopic traffic simulation. First, an improved model for simulating traffic on partially blocked roads is developed by extending the existing cellular automaton model. Second, the improved traffic model is validated at microscopic and macroscopic levels with measured traffic data from an urban road. Third, traffic simulations under various scenarios with different demand flow rates, truck ratios and blockage ratios are conducted through microscopic simulation experiments. Finally, a set of continuous traffic time functions are further developed for disrupted traffic flow with parameters estimated from the generated traffic data. The developed travel time functions for a typical urban arterial road are then compared with the standard Bureau of Public Roads function. The comparison suggests that the standard Bureau of Public Roads function would considerably underestimate the travel time on partially blocked roads and the proposed travel time functions can offer a more realistic prediction. The proposed methodology of developing the travel time functions of partially blocked roads will be helpful for accurate estimation of traffic demand of post-hazard transportation networks.