Design of embedded bearing fault diagnosis system based on Zynq-7000

The vibration signal of bearing faults is often non-stationary and nonlinear. The signal-processing methods based on Hilbert-Huang transform can effectively extract the fault features. The fault diagnosis of machinery is a typical small sample problem. Support vector machine is a machine learning method that is suitable for small sample classification problem. It is effective to the fault diagnosis of bearings. Zynq-7000 is a type of novel heterogeneous processor, which has both embedded resources and logical resources. It can be designed flexibly and has a powerful performance, well suited to the demand for bearing fault diagnosis. After completion of the hardware platform built on Zynq-7000, we develop the host computer which provides users a good display. Moreover, we use the open data of the rolling bearing at case western reserve university and the data collected from INV1612 type multifunction flexible rotor experiment platform, developed by Beijing COINV, to make the algorithm and hardware platform function test and verification. Test results demonstrate that the hardware platform based on Zynq-7000 can effectively make fault diagnosis of bearings.