Skip to Main Content
This paper describes an on-going research work on the train dispatching problem. This dispatching or rescheduling problem (which is NP-complete in general cases) lacks a benchmark or a library to test different algorithms. Therefore, we present a modelisation of the problem, on which our future tool is based. The tool is decomposed into a simulation engine (event driven) which may use different algorithms to solve the problem. Different algorithms may be used and tried.