My work is part of a development of a sensor system which is able to detect faulty wheels on a train that passes by a given track on the basis of the noise that is produced by the train. Minor faults can increase the level of the caused noise, disturbing the passengers as well as the inhabitants who live next to the railway. More serious faults can be causes of accidents, endangering the lives of passengers and may increase the risk of major (in some cases environmental) material damage.
I was provided with noise samples of previously known wheel faults for my thesis. I had to find a method so I could detect and identify faulty wheels in the given samples with high probability. I focused my attention on the detection of wheelflats, since this is the most common form of wheel defects on trains and can be the source of serious damages if not detected in time. To implement the measurement system I used National Instruments CompactRIO and LabView environment. My thesis summarizes my result in fault detection.