A time-space (TS) traffic diagram is one of the most important tools for traffic visualization and analysis. Recently, it has been empirically shown that using parallelogram cells to construct a TS diagram outperforms using rectangular cells due to its incorporation of traffic wave speed. However, it is not realistic to immediately change the fundamental method of TS diagram construction that has been well embedded in various systems. To quickly make the existing TS diagram incorporate traffic wave speed and exhibit more realistic traffic patterns, the paper proposes an area-weighted transformation method that directly transforms rectangular-cell-based TS (rTS) diagrams into parallelogram-cell-based TS (pTS) diagrams, avoiding tracing back the raw data of speed to make the transformation. Two five-hour trajectory datasets from Japanese highway segments are used to demonstrate the effectiveness of the proposed methods. The travel time-based comparison involves assessing the disparities between actual travel times and those computed using rTS diagrams, as well as travel times derived directly from pTS diagrams based on rTS diagrams. The results show that travel times calculated from pTS diagrams converted from rTS diagrams are closer to the actual values, especially in congested conditions, demonstrating superior performance in parallelogram representation. The proposed transformation method has promising prospects for practical applications, making the widely-existing TS diagrams show more realistic traffic patterns.
To assess road traffic safety risk in civil aviation airports and develop effective accident prevention measures, this study proposed a risk assessment method based on accident tree and Bayesian network for airport aircraft activity areas. It identified influencing factors in the aircraft activity area from the perspectives of person-vehicle-road-environment-management and analyzed their relationships. The Bayesian network was utilized to determine initial probabilities for each influencing factor. Findings indicated a relatively high overall safety level in the airport's road traffic system. Accident trees were employed to qualitatively and quantitatively analyze common human-vehicle accident patterns. The initial probabilities obtained from the Bayesian network served as basic event probabilities in the accident tree to determine the occurrence probability of the top event. Taking a 4F airport in China as an example, accident cause analysis identified five important risk sources in human-vehicle accidents, including blind spots for special vehicles, illegal driving by drivers, pedestrians violating regulations, passengers entering restricted areas, and blind spots at intersections. Corresponding safety management measures were formulated. The study concluded that the integration of Bayesian networks and accident trees effectively determines accident probabilities and offers specific solutions, thus playing a crucial role in enhancing road traffic safety management within aviation airports.
