Statistical Models for Health Monitoring of Rare Events in Railway Tracks
摘要
Squats are punctual material failures at railway tracks which can lead to critical effects when not detected or removed in time. This paper proposes a statistical modelling approach to describe the relation between risk of squat occurrence and potential root causes, such as track characteristics, traffic patterns and load data, considering multi-collinearity and rare events. Data from the Swiss railway were used to develop models for track health monitoring.