Combined Optimization of One-Way Streets Configuration and Signal Setting in Urban Transportation Networks

When optimizing traffic management strategies of urban transportation networks, one-way streets configuration and signal setting problems are often dealt with separately, i.e., when considering one of them, the other is assumed to be fixed and given, or ignored. However, the interactions and interdependences between the above two aspects suggest that simply combination of individual optimal schemes is not necessarily the combined global optimal scheme. This paper does address the combined optimization problem of one-way streets configuration and signal setting taking account of route choice behaviors. One-way streets configuration and signal setting are described by binary integer and real-valued variables respectively. A mixed integer bilevel programming model is formulated to tackle the Stackelberg game of the decision process in which the upper level sub-problem aims to minimize the total travel time encountered by all drivers considering signal delays; while the lower one, a stochastic user equilibrium model, aims to describe route choice behaviors of users reacting to the given one-way streets configuration and signal setting scheme from the upper level. Then, a hybrid GA-MSA solution algorithm is proposed to solve this BLP model. Finally, a numerical experiment is included to demonstrate the application of the proposed model and algorithm.