Model for predicting the frequency of broken rails

Abstract

Broken rails can cause train delays, trains cancelations and, unfortunately, they are common causes of accidents. This affects planning of a resources, budget and organization of railway track maintenance. Planning of railway track maintenance cannot be done without an estimation of number of rails that will be replaced due to the broken rail incidents. There are many factors that influence broken rails and the most common are: rail age, annual gross tonnage, degree of curve and temperature in the time of breakage. The fuzzy logic model uses acquired data as input variables to predict the frequency of broken rails for the certain rail types on some Sections.