Institute of Fundamental Technological Research

In the present paper, the authors try to find distinguishable characteristics based on acceleration data obtained from a railway car body. Those quantities allow constructing diagnostic maps, which in turn can be used to classify vehicles clearly as damaged or undamaged. In this approach, the assessment is based on measurements taken from both an undamaged and damaged rail vehicle. The data are compared by methods of artificial intelligence. The aim of this paper is to study aspects of a pure signal-based approach to the end of a quick recognition of damage by assessment of selected parameters of sensor data in real time.

analysis of measured signals, experimental investigations, artificial intelligence, railway wagon, quick recognition of damage, damage diagnostic, neural networks