Search

Vehicle behavior of a PRT (Personal Rapid Transit system) multiberth station has already been analyzed as a continuous time system under the assumption of Poisson vehicle arrival and exponential passenger loading and unloading time. However, it is more realistic to model it as a discrete time system considering a widely used point-follower control. Thereby a general passenger loading and unloading time can be easily treated. An off-line station with two berth in PRT, in which vehicles move on a guideway under point-follower control, is modeled and analyzed as a discrete-time Markov chain in this paper. The state quantity is defined based on the physical state of two berths at the arrival instance of a cell. First it is shown that the inter-entrance time can be calculated numerically for a general passenger loading and unloading time. Then, several measures of the station behavior such as mean inter-entrance time, entrance frequently and entrance factor are analytically obtained for the case of fixed passenger loading and unloading time, the opposite of the exponential distribution. The differences between two berths and one berth are discussed by showing numerical examples. [ABSTRACT FROM AUTHOR]

A control algorithm and a probability-theoretical traffic analysis of a grade-separated intersection model of the personal rapid transit (PRT) systems under the quasi-synchronous scheme is presented. At the intersection, vehicles diverge from the less preferential subline and merge into the preferential mainline. To realize efficient cell-assignment, this algorithm makes a slot-slipping decision before the diverge point and performs the actual slot-slipping after it is on the off line. Even a diverge-minded vehicle is forced to go straight (be adopted) when the number of maneuvering cells exceeds its limit and passes the intersection without the maneuver. The behavior of this algorithm under the random arrival assumption is formulated as a Markov chain, the analysis of which leads to the explicit expressions for the delay distribution, the average delay, and the abort rate. [ABSTRACT FROM AUTHOR]