ABSTRACT
Many traffic-responsive ramp metering methods have been proposed, and there exist some successful ideas we can borrow. For instance, an active disturbance rejection control approach is proposed for the traffic density control via ramp metering. This new approach consists of a tracking differentiator, an extended state observer, together with a nonlinear feedback control law [1]. A new method based on the composite of PID controller and cerebella model articulation controller is proposed for ramp metering, and the density control problem is formulated as an output tracking and disturbance rejection problem [2]. A hierarchy control strategy and a genetic algorithm optimization for the coordinated ramp control are proposed. There are two control layers in the coordinated ramp control system: the coordination control layer and the direct control layer. The direct control layer is to keep the actual values of traffic densities in the vicinity of the desired traffic densities via PI controllers. Genetic algorithm optimization is used to search the optimal PI parameters [3]. A PI control method based on particle swarm optimization is proposed to regulate the number of vehicles entering a freeway entrance point, and the established traffic flow model is a simple first-order model [4]. The design of dual heuristic
programming for the optimal coordination of freeway ramp metering is presented. Specifically, the dual heuristic programming method is implemented to solve congestions. A neural network controller is achieved by the dual heuristic programming method with traffic models [5]. The iterative learning control approach is applied to address the traffic density control problem with ramp metering. The iterative learning approach can effectively deal with this class of control problem and significantly improve the traffic response [6].
