SIMULTANEOUS OPTIMIZATION OF THE DESIGN AND OPERATIONS PLAN OF SCHEDULED MASS TRANSIT LINES.

摘要

This dissertation refines an existing scheduled passenger service process optimization model (SPASM) intended to deal simultaneously with the design and operations plan of scheduled mass transit lines. Although crew rostering, facility design, fleet selection and operations planning of scheduled passenger carriers have been the subject of much study in recent decades, there have been few attempts to establish or formulate the relationship between these subject areas in one model. Most of the previous models focus on individual aspects of the system design problem separately or sequentially and therefore may not reach the true optimal design of the system.;With this model, one can easily examine the substitution possibilities of producing given any output. These possibilities of factor substitution in principle exist among various resources and are shown to be substituted in the case of transit. Comparing the results obtained from this model with the traditional design approaches which ignore the possibility of substitution or of alternative designs reveals that the traditional design approach might yield a design which is clearly non-optimal and hence more costly than necessary.;The primary motivation of this study was to develop a model for identifying optimal designs of a system, given the prices of various resources, the capacity and expected traffic, minimum level of service, and any appropriate constraints on the technology. It is hoped that this study will contribute to the better understanding of the design and operation plan of scheuled passenger carriers. This should be helpful in achieving more efficient public transit systems and in reducing total long run costs.;This study uses mathematical programming techniques to model an urban mass transit line to consider the design options including labor, vehicle frequency, vehicle size and number of vehicles in the fleet, storage spaces, and guideway expansion, with the objective of minimizing total long run daily costs. Also, we develop heuristic procedures which make application and solution of the model easier and less costly in analyst's time and computer time. This model is suited for application in the evaluation of existing scheduled modes of passenger transportation, and design of new transportation systems. It should also be useful in the analysis of cost features of transit systems.