Multi-Agent Systems-based Coordination to drive Autonomous Self-Organized Systems

Funder

Organizational Unit

Publications

Complex Intersections with a Dedicated Road Lane per Crossing Direction

Publication . Reddy, Radha; Almeida, Luis; Santos, Pedro M.; Tovar, Eduardo

Complex intersections are often busier with a separate road lane per crossing direction, i.e., left, straight, and right. These intersections eliminate the diverging and merging conflicts; thus, vehicles only fall under crossing conflicts within intersections. However, the traditional way of serving vehicles from one road at a time increases traffic congestion and hinders performance. To address this issue, we extended the synchronous framework for complex intersections with a separate road lane per crossing direction, which was initially presented for single-lane and two-lane intersections in which roads are shared among vehicles with different crossing directions. We compare the performance of our synchronous framework against the traditional Round-Robin (RR) intersection management approach.

2023-07-20Other

Open access

Synchronous Management of Mixed Traffic at Signalized Intersections towards Sustainable Road Transportation

Publication . Reddy, Radha; Almeida, Luis; Gutiérrez Gaitán, Miguel; M. Santos, Pedro; Tovar, Eduardo

In urban road transportation, intersections are traffic bottlenecks with increased waiting delays and associated adverse effects. A recently proposed intelligent intersection management (IIM) approach, the Synchronous Intersection Management Protocol (SIMP), synchronizes the vehicles access to simple single-lane isolated intersections, outperforming competing approaches in various performance metrics. In this paper, we apply SIMP to multi-lane intersections, increasing significantly the applicability of the protocol while dealing with the additional complexity emerging from the multiple crossing conflicts. Using the SUMO simulator, we compare the performance of SIMP with two conventional (Round-Robin - RR and Trivial Traffic Light Control - TTLC) and two IIM approaches (Intelligent Traffic Light Control - ITLC and Q-learning based Traffic Light Control - QTLC) under continuous and interrupted upstream traffic flows scenarios in urban settings. The results using a maximum speed of 30km/h confirm the superiority of SIMP, improving traffic throughput (~14.4%) and reducing travel delays (~64.4%) and associated fuel consumption (~25.5%) when compared to the best of the other approaches.

2023-05-22Journal article

Open access