Abstract
In this paper mathematical model and a suboptimal control scheme for automated public transportation systems with a closed-loop single-lane guideway, is proposed. Instead of time, distance is chosen as an independent variable taking into account the simplicity of implementation in real systems. The deviation from the scheduled time interval between the adjacent vehicles and the deviation from the scheduled velocity of each vehicle are chosen as the state variables of our discrete-distance dynamic model for the traffic control, and the deviation from the scheduled acceleration of each vehicle is chosen as the control variable.
The suboptimal control scheme, which is designed by a linear regulator with upper and lower bounds for control inputs, is proposed, and is compared with the optimal control scheme obtained by a mathematical programming technique. It is shown that the suboptimal control, which can be easily implemented as an on-line real-time traffic control, is very close to the optimal control. The performance of the control proposed in this paper is compared with those of two other methods in which the time is chosen as an independent variable. It is found that the performance is not degraded appreciably by choosing the distance as the independent variable. Therefore, the effectiveness of the model and the control scheme for automated public transportation systems in this paper is ascertained from the viewpoint of their realizability in a real system.