Computational method for feedback perimeter control of multiregion urban traffic networks with state-dependent delays

Perimeter control is the manipulation of traffic flow by adjusting traffic signals on the borders between regions. It can be used to address the problem of traffic congestion in urban areas. Although there are many papers devoted to perimeter control, the dependence of time delay on traffic state is rarely considered. Moreover, most of them depend on accumulation reference points of the network and the stability of given equilibrium points to derive a perimeter controller. In reality, such equilibrium points and the optimal reference points are unavailable in advance. With these in minds, this paper formulates an optimal feedback perimeter control problem (OFPCP) subject to state-dependent delays without given information about the optimal reference point and equilibrium point. Different from the control method based on linearization or partial linearization, a hybrid algorithm combining filled function method and gradient -based method is introduced to solve this challenging optimal feedback control problem. Some experiments are performed to demonstrate the effectiveness of our method.