The diagnosis of failures in train traffic installations can be done in several ways: direct observations and measurements, software assisted diagnosis using specific software packages, process
variable monitoring for electronically centralized installations. This work presents basic concepts for Model Based Diagnosis (MBD) that uses fuzzy logic to analyse the integrity of Centralized Traffic Control Installations (CTC) in train traffic. We define the diagnosis relations to be used and show how to apply them to train traffic security installations. Implementing these concepts into an expert system assists maintenance operators in quick
failure diagnosis of the train traffic security installations.