The possible applications of fuzzy logic for the treatment of conflicts in the dispositional tasks of railway traffic control centers

The handling of traffic conflicts of trains is rather complicated if full dynamic models are used. Often it is senseless to take into consideration all details of the motion equations and environmental parameters especially as the main parameters, like maximum speed of the trains involved are not exactly known. Also, a large number of noise type variables are always available. The paper proposes the use of the Mamdani type fuzzy rule based reasoning algorithm for handling railway traffic conflicts. The essential point is the granulation of all important parameters into a relatively small number of categories labeled by terms such as small, large, etc. This way the possible decisions for a concrete traffic conflict situation get various fuzzy weights after accumulation of all the fired rules. The final (non-fuzzy) decision is taken by choosing the option with the largest weight. A simplified example is presented for a delay and overtaking type conflict.