This work defines transit schedule design (TSD) as an optimization problem to construct the transitschedule with the decision variables of the location of timing points and the amount of slack time associated with ea...
详细信息
This work defines transit schedule design (TSD) as an optimization problem to construct the transitschedule with the decision variables of the location of timing points and the amount of slack time associated with each timing point. Two heuristic procedures, Ant Colony and Genetic Algorithms, are developed for constructing optimal schedules for a fixed bus route. The paper presents a comparison of the fundamental features of the two algorithms. They are then calibrated based on data generated from micro-simulation of a bus route in Melbourne, Australia, to give rise to (near) optimal scheduledesigns. The algorithms are compared in terms of their accuracy and efficiency in providing the minimum cost solution. Although both procedures prove the ability to find the optimal solution, the Ant Colony procedure demonstrates a higher efficiency by evaluating less scheduledesigns to arrive at a 'good' solution. Potential benefits of the developed algorithms in bus route planning are also discussed. Crown Copyright (C) 2011 Published by Elsevier Ltd. All rights reserved.
This work proposed a model for transit schedule design by integrating the spatial-equity considerations into the slack-time setting problem. The proposed model can help to design a transit service for more equitable r...
详细信息
This work proposed a model for transit schedule design by integrating the spatial-equity considerations into the slack-time setting problem. The proposed model can help to design a transit service for more equitable reliability of all stops distributed along the route. The transit operators' risk taking attitudes are also taken into consideration by introducing the travel-time budget (TTB) and mean-excess travel time (METT) as two risk measures. The proposed model can be formulated as a stochastic optimization problem and solved by a subsequently designed Monte Carlo (MC) simulation based genetic algorithm (GA) procedure. Scenario-based numerical examples from a real bus route in Beijing city of China are presented to demonstrate the characteristics of the model. The sensitivity of the parameters setting are also taken into consideration as well, which reflects the impact on the schedule setting of various factors including actual travel-time variability, risk-taking attitudes of transit operators, the bus drivers' adjustment, and bus-frequency setting.
This paper proposes a general model for the bus route scheduledesign (BRSD) problem considering the uncertainty in the bus travel times to minimize the expected total schedule deviation from a reliability standpoint....
详细信息
This paper proposes a general model for the bus route scheduledesign (BRSD) problem considering the uncertainty in the bus travel times to minimize the expected total schedule deviation from a reliability standpoint. The preferences of bus operators are considered in the objective function. Based on the general model, four stochastic linear programming models are subsequently developed by successively incorporating different operation strategies;they include (1) no control strategy (NCS), (2) bus drivers' schedule recovery (DSR), (3) holding control strategy (HCS), and (4) both DSR and HCS (BRHS). Then, a Monte Carlo simulation based solution method is designed to solve these four models, respectively. Finally, numerical tests based on a real bus route are used for a comparative analysis of different operation strategies used in the BRSD. The results show that HCS and BRHS outperform other operation strategies in terms of the schedule deviation of the buses. Interestingly, HCS performs slightly better than BRHS when bus operators assume that later arrival at scheduled point is more harmful than earlier arrival. Compared to NCS, HCS can save objective function approximately 67.9% in all cases. In addition, operation strategies that could better fit a particular scenario are suggested.
暂无评论