Datensatz

Zusammenfassung

The evacuation during the 2011 Great East Japan Earthquake and Tsunami revealed a high preference for vehicle evacuation in coastal areas. Although evacuation by foot is generally suggested, considering the contribution to evacuating vulnerable people (e.g., the elderly, infants, and disabled), it is essential to plan for safe and reliable evacuation by car. Connected and Autonomous Vehicles (CAVs) have a great potential to improve transportation system efficiency through autonomous driving and communication technologies. Assuming that CAVs will be widely used in the future, it is crucial to examine the impact of autonomous car evacuation on traffic flow (e.g., moving speed, evacuation time, and traffic capacity). In this study, we evaluate the effectiveness of CAVs in tsunami evacuation using the flocking algorithm as the primary control model. We developed a microsimulation of CAVs self-driving along a road with obstacles using agent-based modeling. Then, we quantitatively compared the CAVs model to a human driving control model as a baseline to assess its effectiveness. The results indicated that CAV flocking control performs better than the baseline in a three-lane roadway environment with obstacles. Applying the proposed flocking algorithm method and optimizing navigation-control parameters to various traffic situations, this study contributes towards investigating safe autonomous car evacuation in the future.