Towards a resilient metro system

Abstract

The Metro system plays an important role in public transportation due to its high speed and large volume. However, disruptions in the metro system may significantly impact the welfare of passengers. The resilience of metro systems, i.e., the ability to recover from disturbances, is of great concern to urbanized societies. In the rare literature about the resilience of a metro system, none has defined and measured resilience as a minimax problem. Therefore, it is significant to propose a framework to study the resilience of metro systems. This thesis aims to propose a resilience measurement framework for metro systems to evaluate their ability to recover from disruptions and explore the resources and time required to realize a higher resilience. Before analyzing the metro resilience, a completely bi-directional metro system model is established considering its functional and structural characteristics. The author considers situations that a large passenger flow occurs at directed platforms. To assess the resilience in such cases, we have to solve four key tasks: (1) propagation of the imbalance, (2) identification of critical platforms, (3) A recovery strategy to minimize the effects of imbalance on critical platforms, (4) A resilience measurement framework. First, the imbalance propagation is analyzed and predicted, where the variance of propagation caused by the occurrence time, the type, and the level of primary disruption is examined. Then, time-variant critical platforms are identified at the directed platform level, where critical platforms are evaluated considering the usage level in normal daily conditions and the effects on system performance when dynamic events occur. A combined skipping and detainment recovery model is then proposed to restore the system performance. Finally, a resilience assessment framework based on a minimax model is presented using the above-combined recovery considering multiple oversaturation situations. The above methods are applied to Shenzhen Metro system for validation. Some managerial insights are summarized: (1) The disruption type, the disruption degree, and the occurrence time are determining factors for the propagation results. The type determines the essential characteristics of the propagation trend, the occurrence time determines the affected space and time scope, and the degree mainly determines the range of the extra waiting cost. (2) The results of critical platforms over a day and the day of a week showed differences. Most of the critical components are located around transportation hubs or high-density residential areas. (3) Compared with the skipping strategy and the detainment strategy, the combined strategy can output the shortest recovery time and the smallest amount of extra waiting time to recover the metro system. (4) The resilience results are closely related to the disruption degree and available resources when the critical platform suffers one type of disruption. Under one disruption, the proposed approach can output the required extra trains and recovery time to recover to the normal condition. Metro companies can prepare extra trains for some target disruptions to improve resilience.