This study aims to apply the concept of multi-agent algorithms to coordinate signalised intersections along a study corridor. A single intersection is modeled as a local autonomous controller equipped with local optimisation model to generate the control plans under real-time traffic conditions. For a network of intersections, an optimal signal plan at each local controller does not imply that the network performance is optimal. The study grouped the intersections to zone level based on location and traffic volume.The common cycle time generated by zone is applied to all zone members.The study also investigated the determination of offset between signals, i.e., the time displacements of green splits along a movement direction.Unlike the typical signal coordination which applying offsets to all intersections along the arterial, the proposed control system divides the coordination to intra-zone and inter-zone levels which aim to maximise flow in both local and network conditions. A comparative evaluation between the proposed system and other control strategies is presented under different congestion scenarios through simulation experiments. (Author/publisher) For the covering entry of this conference, please see ITRD abstract no. E214133.
Samenvatting