Research On Fault Diagnosis System Of Gearbox In Large Wind Turbines

Fault diagnosis of wind turbine and early warning is an important assurance for the safe operation of the wind farm.With the expansion of the single unit capacity and large-scale wind farm putted into operation, realizing convenient fault diagnosis and warning has became one of the highlights in agro-scientific research in the wind power generation technology. Wind turbines mainly include gear box, generator, blades, hydraulic system and components such as yaw system.Gear box is the key to the units speed, energy transfer components, which is also the highest incidence of failure parts whose operation affects the stability of the whole machine performance.In bad environment, the change rate of gear box may even reach100%.After analyzing the main failure types and various fault mechanism comprehensively of gearbox in wind turbine, proposing the corresponding diagnostic tests.Taking vibration signals of gearbox as the source of information to build a wind turbine gearbox fault diagnosis system. The hardware and software structure and function of the system are determined.Hardware part of the fault diagnosis system includes the vibration sensor, charge amplifier, data acquisition card and industrial computer. The common PC institutions and ordinary PC in the engine room bottom, the rest is placed in the fan at the top of the cabin, adopts Ethernet cable communication between industrial PC with PC. Vibration sensor is tightly attached to the gear box of the box, the vibration signal into a voltage signal after the charge amplifier amplification, through data acquisition card is saved to the industrial control, completion of data analysis and processing, processing of data, together with the original data is passed on to the bottom of the PC upply wind farm staff observation record, also can through the wireless network card installed in the PC master control room transfer data through wireless WIFI LAN to ind farms, wind farm workers in master control room can perform data observed and recorded. Using Matlab and VC mixed programming to complete the system software, in Matlab environment to establish several."M" file, compile time domain statistical value calculation. FFT and wavelet analysis, wavelet packet analysis and time-frequency processing algorithm is then implemented by using VC call Matlab engine data processing, data extraction of fault features.System has been conducted a field test in red bay wind farm of shanwei city, the test shows that the system is running in good condition and the results are accurate and comprehensive.