Compressed Sensing Technology Applied to Fault Diagnosis of Train Rolling Bearing

Compressed sensing technology is a new approach in the field of signal processing. It can achieve data acquisition and compression at the same time, and express the original signal in a simple way. Network monitoring technology used in detecting the fault of the train's rolling bearing always produce large amounts of datum. Too much of the parse inspection data cause a lot of unnecessary cost of energy and network bandwidth. In this article we propose compressed sensing method to deal with the inspection date of train's rolling bearing which will be transferred in the sensor network to reduce the data quantity. Experimental result show that, after compression and reconstruction, the reconstructed signal still contains most of the information and energy of the original signal, we can still detect the fault of train's rolling bearing from the reconstructed signal accurately.