One of the workstreams of the NEXUS project focuses on the adaptability of metros, with a distinct emphasis on understanding, predicting and responding to fluctuations in passenger demand. To do this, scenario-based simulation methods are used to validate and exploit the tools, concepts and applications developed within the NEXUS project. A comprehensive set of simulation modelling scenarios have been carried out to ensure that all the metro system models, concepts and applications are accurate, consistent, and complete. All the metro system models, concepts and applications have been properly tested and validated, to ensure an accurate representation of real-world metro system operations which can be used to support operators’ decision-making.
The simulation models, in particular, adopted a multi-perspective (using a suite of validated metro simulation models developed over three different simulation platforms), multi-purpose (according to 23 operational scenarios that organised into 5 distinct themes), multi-level (looking at macro-, meso-, and microscopic level impacts) approach to evaluate metro system performances and the possible impacts of future Grade of Automation 4 (GoA4) – fully automated operations. Each metro simulation model was calibrated individually against real-world observations for accuracy, and cross-validated against their counterparts to ensure consistency. Scenarios were developed in close coordination with the two metro operators (i.e. AMT in Genoa and MetroS in Sofia), ensuring accurate reflection of practical metro operational concerns and relevance.
Impacts were revealed at network-wide, metro line, metro station, as well as for crowd management levels. The primary focus of the evaluation was to identify the strengths and weaknesses of the current metro systems in Genoa and Sofia, and the potential impacts of future fully automated operations in dealing with the current passenger demand and some notable disruption scenarios. It was found that GoA4 (in the form of reduced train headways) is most effective as a disruption-mitigation tool when the disruption affects a single node rather than the entire line. When considering switching to GoA4 operations, a suitable trade-off between reducing passenger wait times and improving train capacity utilisation should be made when determining the optimal train headways.
Building upon these insights, the suite of simulation models could be further exploited in association with advanced optimisation techniques to further improve metro service operations (e.g. refining train headways and/or dwell times to minimise passenger waiting time and queue lengths).
Dr Hing Yan Tong
On behalf of the Aston University Team
NEXUS WP7 Leader