Analysis Of Wind Turbine Reliability Based On Failure Data

In recent years,the accumulative installed capacity and new installed capacity of wind turbines in China were ranked first in the world,and the market scale was huge.However,the failure rate of wind turbine was still at a high level because of the harsh working conditions,the vary manufactural levels of the early machine,the immature maintenance technology and so on.Unfortunately,it could severely affect the economic benefit of wind farms,and restricted the normal development of wind power industry.Therefore,we constructed a model to analyze the main component of the wind turbines according to the Weibull distribution model by using the actual failure data of a wind farm in Zhangjiakou,Hebei province.On the basis of the distribution parameters of wind turbine components and the quantitative index of relative reliability,the failure modes of wind turbine components were judged and their maintenance decisions were optimized.The main research work of this paper was as follows:(1)The main system of the wind turbine was introduced,and the main fault modes were simply explained according to the actual data.We focused on the actual operation failure data of 99 doubly fed wind turbines in this wind farm due to the last 6 years,then analyzed the statistics of the failure times and the downtime of the system.Meantime,we established the database for the following correlational research.(2)In this paper,two parameter Weibull distribution model is used to analyze the reliability of main components of wind turbine,and the parameter solution method and relative reliability index of Weibull distribution are introduced.The unknown parameters of Weibull distribution are solved by the least square method and maximum likelihood estimation method,and the reliability index is calculated.In fact,due to the short running time of wind turbines and the shortage of fault data samples,it is found that the accuracy of the two methods is poor when the sample size is small.(3)Support Vector Regression(SVR)is a machine learning method based on the VC dimension theory and the structural risk minimization principle of statistical learning theory,especially for the condition that the sample size of fault data is small.The application of support vector regression method for solving the two parameter Weibull distribution shape and location parameters,compared with the least square method and maximum likelihood estimation can be found for support vector machine in the small sample and large sample conditions,and in the small sample condition has certain advantages.(4)Based on the theory of reliability analysis,we analyzed the Weibull distribution parameters and related reliability index of main components,then estimated the failure mode and maintenance methods,optimization of maintenance strategy,to achieve the goal of safty and stable operation and high economic benefit of wind farm.