Performance and design of mobility allowance shuttle transit services: Bounds on the maximum longitudinal velocity

摘要

The traditional fixed bus route system is considered inconvenient by the passengers either because of the pick-up or drop-off points or the schedule of the system that does not match the rider's needs. Flexible systems such as demand responsive transit (DRT) are more costly and largely limited to specialized operations. The mobility allowance shuttle transit (MAST) services are a new concept in transportation that merges the flexibility of demand responsive transit (DRT) with the low-cost operability of fixed route bus systems. The MAST service has a fixed base route that covers a specific geographic zone, with a set of mandatory checkpoints with fixed schedule departure times conveniently located at major connection points. Given an appropriate slack time, these buses are allowed to deviate from the fixed path to pick up or drop off passengers at their desired locations. The customers are divided into three catgories, namely, hybrid, regular, and random. The MAST system is already implemented in a small scale at the Metropolitan Transit Authority (MTA) of Los Angeles County in California. In MAST services, the main purpose is to move the customers along a primary direction which may be around a loop or back and forth between two terminal checkpoints. A higher demand served per vehicle would allow the service to be more cost efficient. However, if the demand is more, the slower the vehicle would move because of the deviations and service time causing the service to be less attractive. Therefore, the present research work tries to model the relationship between the velocity along the primary direction of a MAST service and the demand to assess the performance of these types of systems and help in the design process. The introduction part of the paper explains the MAST system, the problem undertaken for the study, and states reasons as to why the problem is considered NP hard. In Section 2 of the paper, the literature review for the present work is presented. In Section 3 of the paper, the system definition, the model considered, its length and width, and assumptions regarding the demand are presented. The vehicles are assumed to follow a forward progression through the rectangle in either a left-right or right-left direction. In Section 4 of the paper, the lower bound of V (longitudinal velocity of the vehicle) is considered. The researchers have assumed a no backtracking policy allowing the vehicle to move in the forward direction. In Section 5 of the paper, the optimality of the no back tracking policy is presented. This is done to determine whether there exist sufficient condition on the locations of the demand points that would guarantee optimality of a no backtracking routing policy. In Section 6 of the paper, the first upper bound of V (longitudinal velocity) is formulated. For this the researchers have identified a subset using an algorithm. In Section 7 of the paper, the second upper bound of V is shown by removing certain constraints in the original problem. In Section 9 of the paper, the researchers analyze two different cases that are consistent with the present MAST system. Section 10 of the paper presents the results and some future research directions. (47 refs.)