Identification of noise sources on an automatic railway system using statistical techniques

Describes the process by which sources of intermittent electromagnetic interference (EMI) noise were identified and how the significance of measurement results could be verified with the aid of statistical techniques. The most significant parameters in the failure data were time of the event, the location, the vehicle, and the nature of the event. Checking the actual probability density distribution of the failures by location or by vehicle with a Gaussian distribution confirmed that most events were not vehicle related but had their origins trackside