Early Fault Diagnosis Research On Offshore Wind Turbines Transmission System

Wind energy is a kind of important renewable energy. In our country,onshore wind power development is very quickly, however, the development ofoffshore wind power is in its infancy. Transmission system which is theimportant part of offshore wind turbines, has high failure rate affected by thespecial operation environment. And its failure is easy to cause great influenceto the wind turbines. Therefore, the research of early fault diagnosis foroffshore wind turbine transmission system has very important significance toimprove the efficiency, reliability and economy of wind turbines. Based onfailure characteristics of the transmission system, this paper adopts the methodof vibration analysis to study the fault vibration model of each component. Forthe fault vibration signal is relatively complex and has the characteristics ofnonlinear non-stationary, using Empirical Mode Decomposition (EMD) as thebasic analysis method of signal decomposition analysis. This paper mainly dothe following several aspects of the research.(1)In view of the EMD has mode mixing phenomenon in signaldecomposition, this paper put forward an improved differential Empirical ModeDecomposition (DEMD) on the basis of the differential methods. Thesimulation research show that the improved DEMD can improve the modemixing phenomenon in EMD, and also improve the low-level intrinsic modefunction (IMF) distortion phenomenon in traditional DEMD.(2)This paper also combined with improved DEMD and independentcomponent analysis (ICA) for fault diagnosis method to extract faultcharacteristic signals. On the basis of improved DEMD and ICA, combiningHilbert envelope spectrum analysis and the simulation verified its feasibility inthe early wind turbine transmission. (3)Because of the LabVIEW has graphical language, professional signalprocessing module and open platform. The transmission fault diagnosis systembased on LabVIEW is designed in this paper. The improved DEMD algorithmis used in this system and can analysis the original fault vibration signal withthe typical time-domain analysis parameters. Also its interface is designed andit makes the transmission system fault diagnosis system based on LabVIEWcan reflect the characteristics of transmission system fault more intuitively andrecord fault data to determine the fault trends.