Designing a Demand Responsive Timetable for MRT Services

1 Mass rapid transit systems (MRT) have become the most important public transport mode in many 2 major cities globally, given its advantages in larger capacity, faster velocity and higher reliability 3 compared with other modes. Hence, various projects directed to improve service quality of MRT 4 systems by optimizing operation strategies have been conducted worldwide. Among numerous 5 methods to adjust operation parameters for MRT services, timetable optimization has emerged as 6 the simplest and most effective cure. On one hand, a demand responsive timetable has the potential 7 to improve service quality by reducing both waiting time and crowdedness on trains. On the other 8 hand, for transit operators, a demand responsive timetable will improve operation stability and 9 hence reliability. In this study, we propose an optimization model to solve timetable scheduling 10 considering fine grained dynamic demand data. The objective is to minimize the total waiting time 11 for all the passengers. Historical demand data extracted from smart card system is used as inputs. 12 Then, a predefined demand responsive timetable will be generated with the optimization model. 13 Since operational constraints such as the number of available trains and minimal headway are 14 considered as boundary conditions, we ensure that the proposed schedule actually has the potential 15 to be implemented in actual operation. 16 A case study on a representative MRT service is conducted and presented in this paper. In 17 the absence of effective/operated timetable, simulations on the optimal timetable obtained from 18 this proposed model and simplified timetable as extracted from Google Maps are performed based 19 on detailed travel demand extracted from smart card transactions. The result demonstrates the 20 applicability of the proposed model in designing a demand sensitive timetable and its potential 21 to minimize waiting times. However, simplified assumptions concerning potential operational 22 constraints when switching directions or manpower scheduling, restrain the immediate application 23 of the obtained results. 24