Bi-level and optimal control in urban transportation network.

The control object is the urban traffic of network of eight crossroad sections as a part of density populated area in Sofia. The goal is minimization of vehicle queue lengths in front of the junctions, which leads to faster driving, reducing the air and noise pollution and better ecological picture. The paper analyzes two types of modeling and optimizations, applied for the traffic control: classical optimization (one-level) and bi-level one. The bi-level optimization has the priority to co-ordinate the two-level hierarchical problem and to take into account the interconnections between the crossroad sections. It contains two goal functions, which are considered for the control process. The bi-level and classical one-level linear-quadratic optimizations are numerically applied for the traffic light control of urban network. The comparison gives preferences to the bi-level optimization, confirmed by numerical simulations. [ABSTRACT FROM AUTHOR]