Integrating Microscopic and Macroscopic Models for Railway Network Evaluation
M Kettner, B Sewcyk, C Eickmann, IVE, University of Hannover, DE
Microscopic and macroscopic models for transport and operational infrastructure planning are widely used by railway companies throughout the world. The presented paper describes the process of integrating these models in order to simplify the data storage, data administration and especially the process of data acquisition. The approach taken by the IVE (Institute of Transport, Railway Construction and Operation at the University of Hanover, Germany) to integrate these two models is presented. This approach will be exemplified by the integration of two models developed at the IVE.
h4. The microscopic model
Extensive infrastructure modifications in railway systems generate high costs and predetermine the basis of the railway operation (e. g. timetables) in such systems for many years. Operational changes due to or independent of infrastructure modifications can adversely affect the operational flexibility and reliability if unconsidered or unexpected problems occur. Microscopic operational simulation enables the railway operator to determine the consequences of infrastructure or operational changes in railway systems, before investing funds or putting an unfeasible timetable into operation.
There is a long experience in simulation and evaluation of railway operations at the IVE. Development of the microscopic simulation system RailSys started in the 1980s mainly for the German Federal Railways. During the last few years rail operators throughout the world discovered the advantages of operational simulation. It has become an inherent part of the planning process. RailSys offers the features necessary to accurately calculate and evaluate infrastructure and operational measures.
h4. The macroscopic models
The macroscopic simulation model NEMO, developed at the IVE since 1999, is a strategic planning tool for the evaluation of infrastructure and operational measures in guided transport systems. Based on a macroscopic network, NEMO models the complex interaction between network infrastructure, railway operation and transport demand.
NEMO provides a way to efficiently evaluate different production and infrastructure scenarios in passenger and freight traffic. Transport supply can be optimised based on demand and capacitive as well as operational bottlenecks can be detected. Thus, efficient use of investment funds is supported by the model. Furthermore, an economic evaluation of planning scenarios can be used to identify saving possibilities.
Combining microscopic and macroscopic models Precise data forms the basis of all results and calculations of a macroscopic simulation instrument like NEMO. Thorough data preparation and care are indispensable, but often involve much effort. Minimising this work and the costs involved requires utilisation of existing resources. The idea presented in the complete paper is to integrate the two models, which would give several advantages for railway companies. The infrastructure data has to be maintained only once in the microscopic model and can be effectively reused in the macroscopic one. In order to be able to reuse the data of the microscopic model, the generation of the macroscopic network has to be automated.
The main element needed by the macroscopic model is a network graph with associated running time and minimum train headway information. This network graph is at an higher abstraction level than the network graph of the microscopic model. The train headway specifies the time interval at which two successive trains can be operated in the same area of the infrastructure without hindering each other. Train headways are used in the calculation of infrastructure occupation times and therefore are required for the identification of capacity bottlenecks.
Mapping of model trains on an infrastructure represented with high accuracy is realised in the macroscopic model by using functionalities of RailSys, which provides microscopically precise mapping of the infrastructure. In connection with NEMO, this microscopic model calculates the running times and train headways for all model trains defined at the infrastructure of the macroscopic network. The communication between the two models is done via a program-program interface. All data is exchanged automatically. This is realised by the microscopic model acting as a server-application in the background, while the macroscopic model requests all network infrastructure and operational data using socket communication. This process guarantees that all required data is always available at high accuracy.
The described procedure offers two main advantages. On the one hand, all changes to the rail network must only be updated by the infrastructure operator in the microscopically precise representation. This prevents contradictory data in different network representations. Moreover, the work required for data care is remarkably reduced. On the other hand, if the macroscopic network graph has been generated on the basis of a microscopically fine data basis, all running times and train headways can also be determined on this highly accurate basis and be transferred to the network graph.
The complete paper will demonstrate the approach chosen to combine the two models in great detail and give some examples of its applications. In addition it will discuss further possible improvements to the software that will provide a simplified automated generation and exchange of all necessary information.
Association for European Transport