Mathematical Model for Safety Evaluation of Distributed Interlocking System

Each safety function (SF) which is realized by an interlocking system (IS) have to be subjected to proving safety. Proving safety is also about the evaluation of a random failures safety integrity level of the safety function. In the case of modern (electronic) interlocking systems, the evaluation of the influence of multiple failures is based on using graphic-mathematical models. If safety function is realized on multiple levels of the interlocking system, then the creation of the model is demanding on an analyst's work and by that a probability of an analyst's mistake is also increasing. This paper describes the procedure of the creation of the model which is based on the decomposition of the interlocking system into individual submodels with exactly defined logical links. Knowledge of basic parameters of submodels and knowledge of logical links of submodels allows to merge submodels into one model and to calculate a hazard rate for the evaluated safety function. Fault Tree Analysis (FTA) is used as a basic method. Input parameters for a quantitative part of FTA (parameters of basic events) are defined by Markov Chain.