Multi-Objective Stochastic Synchronous Timetable Optimization Model Based on a Chance-Constrained Programming Method Combined with Augmented Epsilon Constraint Algorithm

• Published in: Mathematical problems in engineering Vol. 2022; pp. 1 - 18
• Main Authors: Yuan, Yu, Wang, Pengcheng, Wang, Minghui
• Format: Journal Article
• Language: English
• Published: New York Hindawi 28.08.2022; John Wiley & Sons, Inc
• Subjects: Algorithms; Buses (vehicles); Collaboration; Commuting; Constraint modelling; Engineering; Integer programming; Linear programming; Multiple objective analysis; Normal distribution; Operating costs; Optimization; Optimization models; Pareto optimization; Passengers; Planning; Public transportation; Quality of service; Random variables; Randomness; Scheduling; Sensitivity analysis; Stochastic programming; Synchronism; Timetables; Transportation services; Transportation systems; Travel time
• ISSN: 1024-123X; 1026-7077; 1563-5147; 1563-5147
• DOI: 10.1155/2022/9222636

The design of the timetable is essential to improve the service quality of the public transport system. A lot of random factors in the actual operation environment will affect the implementation of the synchronous timetable, and adjusting timetables to improve synchronization will break the order of normal service and increase the cost of operation. A multi-objective bus timetable optimization problem is characterized by considering the randomness of vehicle travel time and passenger transfer demand. A multi-objective optimization model is proposed, aiming at minimizing the total waiting time of passengers in the whole bus network and the inconsistency between the timetable after synchronous optimization and the original timetable. Through large sample analysis, it is found that the random variables in the model obey normal distribution, so the stochastic programming problem is transformed into the traditional deterministic programming problem by the chance-constrained programming method. A model solving method based on the augmented epsilon-constraint algorithm is designed. Examples show that when the random variables are considered, the proposed algorithm can obtain multiple high-quality Pareto optimal solutions in a short time, which can provide more practical benefits for decisionmakers. Ignoring the random influence will reduce the effectiveness of the schedule optimization scheme. Finally, sensitivity analysis of random variables and constraint confidence in the model is made.