| Popis: |
Dynamic rerouting framework can improve urban traffic management by mitigating urban traffic congestion. Emerging technologies such as fog-computing offers low-latency capabilities and facilitates the information exchange between the vehicles and infrastructure systems, and this fosters dynamic rerouting efficiency. In this study, a 2 stage-method combining GAQ (Graph Attention Network- Deep Q Learning) and EBkSP (Entropy Based $k$ Shortest Path) is proposed using a fog-cloud architecture to reroute the vehicles in a dynamic urban environment to achieve improved travel efficiency. First, GAQ analyzes the traffic conditions on each road and for each fog area and assigns a road index based on the information attention from both local and neighboring areas. Second, the route for each vehicle is assigned using EBkSP based on the vehicle priority and route popularity. The results demonstrate attainment of higher speed and lower total travel time for each vehicle in the network, thereby indicating the efficacy of the proposed framework in dynamic rerouting. |
