Blixten 2 - Decision Support for Traffic Managers: Approximate and Exact Optimizing Methods
The need and potential for introducing a wider scope of computational functionality to support proactive train traffic management has become increasingly apparent in recent years. There are a few systems on the market that promise computational functionality to support optimizing, proactive control and planning in real time during various types of disruptions. However, a compilation of the current state of knowledge shows that these systems are designed for specific contexts and that experiences of how they work in practice are rarely published.
Aim of the project
The project aims to develop, apply and evaluate concepts and methods for efficient computational support in operational train traffic management. The focus will be on a further development of the parallelized search algorithm developed in the KAJT-related project TRANSFORM. The algorithm development and evaluation will be based on relevant scenarios defined in consultation with the Swedish Transport Administration and focus on three main aspects:
- Quality measures of proposed replanning measures
- Computational efficiency
- Stability
By quality measures we mean defined indicators that jointly illustrate and succinctly describe the prerequisites and expected effects of a solution. By computational efficiency we refer to the performance of the algorithm over time and by stability we refer to how the performance of the algorithm varies depending on the type and extent of disturbance. The application and evaluation will be done through computer simulations.
Implementation of the project
The project has the following objectives:
- To monitor the research front in the specified area and make available a compilation of the current state of knowledge.
- In collaboration with the Swedish Transport Administration and the FR8Rail2 project, to conduct case studies that highlight more practical aspects of the future application of optimizing calculation support in the operational operation of (Swedish) train traffic.
- To further develop the calculation methods developed in the TRANSFORM project and systematically evaluate them experimentally, as part of the above-mentioned case studies.
Financier: Trafikverket
Status: Ended
Area: Computer Science
Project start: 2020-02-01
Project end: 2022-04-30
Contact person: Johanna Törnqvist Krasemann