Abstract:This article is based on the mechanism of acoustic emission generation and conducts theoretical and experimental research on the fault diagnosis method of sliding bearings based on acoustic emission parameter analysis. Firstly, three lubrication states of sliding bearings were simulated using a steam turbine generator unit simulation rotor experimental platform. Through the experimental platform and the designed experimental plan, acoustic emission signals of different lubrication states were collected using acoustic emission acquisition equipment. Secondly, for the collected acoustic emission signals of different lubrication states, the energy mean and power spectral entropy mean were calculated. A divergence index based on the energy mean and power spectral entropy mean of acoustic emission was proposed for the diagnosis of sliding bearing lubrication states. This method was used to diagnose the simulated signals and compared with the single energy parameter analysis method, It was found that the energy parameter analysis method cannot effectively reflect the three lubrication states of sliding bearings, while the index diagnosis method proposed in the article using a combination of multiple parameters has better signal adaptability and higher discrimination.