Signal Setting with Dynamic Process Assignment

Traffic signal setting can be cast within the general problem of area traffic control. This chapter presents results of a general modeling approach to global optimization for signal setting, which address the problem by including in the optimization model both equilibrium constraints as well as stability ones. Day-to-day dynamics is approached through simple but still effective dynamic process models based on exponential smoothing filters. Local stability of fixed-point states (consistent with equilibrium patterns) is investigated through the spectral analysis of the Jacobian matrix of the recursive equation defining the day-to-day dynamics. An in-depth analysis of the mathematical features of the proposed method is reported together with results of examples ranging from very simple ones, allowing for graphical representation, to more complex ones, useful to address implementation at real scale.