Study On Life Prediction Of Wind Turbine Main Shaft Bearings

With the development of wind power technology and commercial operation of the market on the right track, Wind energy become to the actual production from the theory, at the same time in improving profits. Wind turbine in operation process will Start-Stop constantly, speed changing, the equipment under great damage, the wind farm in the area of wind well, away from the town, traffic underdeveloped, caused by wind machine maintenance costs high. At the same time due to the sudden failure caused the loss of equipment can not be ignored, if can prediction the life of wind turbine parts, have a plan to replace the failure components, will effectively reduce the wind turbine operation maintenance cost, improve the reliability of the wind turbine, improve the competitiveness of wind power generation.Firstly use the fatigue life prediction method of spindle bearing life prediction,through the S-N curve and fatigue life prediction theory, spindle bearing axial force have been obtained; through the basic principles of mechanics, the spindle and its supporting structure is simplified as a simply supported beam, the root according to the horizontal direction and the vertical direction by the force balance of spindle bearings were obtained by radial force. Using ANSYS software to realize the finite element analysis of the spindle bearing, by setting the force loading and the boundary conditions should seek maximum force value and stress maximum position. On the basis of the fatigue life calculation theory, using the fatigue life prediction method of bearing life prediction.Then in the premise of obtaining the fault characteristic value, the spindle bearing life prediction by the similarity of life prediction method. Euclidean distance function to measure the similarity function, prediction model is established, by the addition of recession coefficient of weights and make the model closer to the actual situation and improve the accuracy of the prediction results, through the MATLAB software to realize the large complicated operations and the calculation results are achieved; for less historical data, using a sample failure and the failure sample joint life prediction method, in the history of small samples can effectively increase the prediction accuracy.