Online Surge Detection Method Based on Axial Displacement Sensor of MSCC

Magnetically suspended centrifugal compressor (MSCC) can get higher speed with no friction and improve transformation efficiency. However, with speed increases, air flows more easily to separate from the blade and cause the gas pressure unsteadily, which leads to surge occur, causing the system destroyed. Aiming at the problem that existing surge detection methods usually increase complexity or cannot detect surge in time, this paper proposes an online surge detection method and corresponding surge judgment criteria of MSCC system using the signal of axial displacement sensor based on adaptive interference canceller (AIC) and average phase forward linear predictor (APFLP). At first, the system model of the MSCC is established and the transfer function between surge pressure signal and rotor displacement signal is derived. On these basis, an AIC is designed to eliminate the synchronous vibration signal, which has serious interference to surge detection and an APFLP is designed to estimate the frequency of the surge signal. Finally, the simulation and experimental research of the new method are carried out. The simulation and experimental results show that the method can detect surge quickly and effectively.