Research On Structural Method Based Fault Detection For Wind Turbines

The high failure rate of the wind turbine is the main reason of the low average utilization rate. The failure of the wind turbine will cause the loss of wind power generation and the maintenance cost, thus increasing the operation cost of the wind farm. So one way to improve the operating efficiency of the wind farm is to reduce the failure rate of the wind turbine.The engine room of the wind turbine is installed in a few meters from the ground. The interior space is very narrow. If the fan is in trouble, the maintenance of the wind turbine is very difficult. Even more important, if the wind turb ine can not be found in time and take reasonable measures to deal with, it will be possible to lead to the occurrence of the whole wind turbine catastrophic accident. So it has important theoretical significance to study the fault detection of the wind turbine to understand the operation state of the wind turbine, to find and to accurately locate the wind turbine generator fault, to develop wind farm maintenance plan, to improve the operation reliability of the wind turbine s.In this paper, the fault detection method of wind turbine system based on structural analysis method is presented. The method of structural analysis is used to detect and isolate the fault detection and isolation by the residual equations made of Measurement information. The method can detect the state of the wind turbines from the system level. By analyzing the common faults of the gearbox of the wind turbine, the fault information is imported into the normal dynamics model of the gearbox, and the dynamic model of gearbox containing the fault information is obtained. Under the premise of ensuring the accuracy of capture information, the structural model of gearbox containing fault information is studied and the problem of fault detection and isolation was solved. On the platform of MATLAB, the above algorithms are realized by programming. The common faults of the gearbox under the condition of the common faults and optimal sensor configuration are detected and identified. The simulation results show the correctness and effectiveness of the proposed method.