Research On Information Exergy Diagnosis Method Of Lubrication Condition Of Sliding Bearing Based On Process Signal

As an important power engineering rotating machinery,sliding bearings have been widely used in electric power,aviation and marine power and heating industries since their invention.At present,the diagnostic methods applied to the lubrication state of sliding bearings,such as time domain,frequency domain or ringing coefficient,event counting,etc,have certain limitations.Therefore,in order to realize the real-time monitoring and fault diagnosis of the sliding bearing lubrication state,this paper has developed a diagnostic method that can accurately determine the lubrication state of a sliding bearing,and ensure the safe and economic operation of the sliding bearing and the unit.Based on the AE technology and the method of information entropy and information exergy signal extraction,the information entropy matrix and information exergy matrix samples under different lubrication states are obtained.The acquired sample matrix and proposed diagnostic method are used to simulate the unknown analog signals.diagnosis.Firstly,three kinds of lubrication state samples of dry friction,boundary friction and liquid friction of the sliding bearing and a set of unknown signals are simulated by the steam turbine generator set simulation rotor test rig.Then use an AE acquisition equipment to collect AE signals of different lubrication states,and The three-limit method is used to further distinguish between different lubrication states.In addition,the AE signal of the lubrication state change during the startup of the actual power plant was collected.Secondly,the AE signals and information entropy and information exergy algorithms are used to calculate the signals,and the feature signals are extracted to form the information entropy and information exergy sample matrix.The measurement features include singular spectrum information entropy and frequency domain in the time domain.The power spectrum information entropy and the wavelet spatial characteristic spectrum and wavelet energy spectrum of the time-frequency joint domain obtained the sample matrix of the lubrication state.A method for diagnosing lubrication state of sliding bearing based on information entropy distance and a method for diagnosing lubrication state of Euclidean information entropy sliding bearing are proposed.Using two methods to diagnose the unknown analog signal can effectively distinguish the three lubrication states of the sliding bearing.The accuracy of the lubrication state diagnosis can be further improved by the information entropy distance and the information exergy distance curve,which is of great significance to the running performance and safety of the sliding bearing.Finally,the research results show that the method can accurately diagnose the lubrication state of the sliding bearing,and the discrimination is high,which can provide a novel and useful method for the lubrication state diagnosis of the sliding bearing.