风电机组及其齿轮箱寿命的统计建模

发布时间：2018-02-28 04:25

本文关键词： 风电齿轮箱平均故障间隔时间最优置信限法 Bayes方法可靠性　出处：《兰州理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：为了从可靠性数据的角度对风电机组及其齿轮箱进行可靠性分析,本文以30台1.5MW水平轴风电机组52个月的可靠性数据为研究对象,建立风电机组平均故障间隔时间分布模型；基于风电齿轮箱的无故障数据,分别运用Bayes方法和最优置信限法进行分析,预测风电齿轮箱的可靠度及可靠寿命；并收集风电齿轮箱零部件相关的经验信息,运用Bayes方法或最优置信限方法分析其可靠性。主要研究内容如下： (1)风电机组在实际工作中是可修复系统,对于可修复系统,寿命就是指平均故障间隔时间。所以,提取30台风电机组52个月的故障间隔时间,借助于MATLAB软件分别对几类寿命分布模型进行参数估计、曲线拟合及假设检验,最终建立风电机组平均故障间隔时间分布模型。 (2)由于这30台风电机组的齿轮箱在数据采集截止时间内没有发生故障,可靠性数据表现为无故障数据。根据无故障数据的相关处理方法,选择应用Bayes方法和最优置信限法,分别分析其在无故障数据情形下的可靠性,将两种方法的估计结果进行比较验证,并对最优置信限法中置信水平的取值规律进行了探讨,最终得到风电齿轮箱可靠度和可靠寿命预测。 (3)在经过Bayes方法和最优置信限法对风电齿轮箱无故障数据分析研究后,通过有目的地收集相关经验数据,根据所获得的信息选取最优置信限法分别在形状参数未知与已知的两种情况下,对风电齿轮箱的几类轴承进行可靠性预测。基于可靠性数据对现役风电机组及其齿轮箱进行可靠性分析预测,更符合风电机组实际运行的情况,这将为风电机组及其齿轮箱设计、制造及维修提供参考标准。同时,为国产风电齿轮箱在无故障数据的情形下的可靠性分析提供新的分析思路。
[Abstract]:In order to analyze the reliability of wind turbine and its gearbox from the point of view of reliability data, this paper takes the reliability data of 30 1.5 MW horizontal shaft wind turbine units for 52 months as the research object, and establishes the average fault interval time distribution model of wind turbine units. Based on the fault free data of wind power gearbox, the reliability and reliability life of wind power gearbox are predicted by using Bayes method and optimal confidence limit method, and the relevant empirical information of wind power gearbox parts are collected. The Bayes method or the optimal confidence limit method are used to analyze its reliability. The main research contents are as follows:. Wind turbine is a repairable system in practice. For a repairable system, the life span of a repairable system is the mean time between failures. Therefore, a fault interval of 52 months is extracted from 30 wind turbines. With the help of MATLAB software, the parameter estimation, curve fitting and hypothesis test of several kinds of life distribution models are carried out, and finally, the average fault interval time distribution model of wind turbine is established. 2) since the gearbox of the 30 wind turbines has no fault within the data acquisition deadline, the reliability data is shown as no fault data. According to the relevant processing methods of the non-fault data, the Bayes method and the optimal confidence limit method are selected. The reliability of the method in the case of no fault data is analyzed, the estimation results of the two methods are compared and verified, and the rule of taking the confidence level in the optimal confidence limit method is discussed. Finally, the reliability and reliable life prediction of wind power gearbox are obtained. After analyzing and studying the non-fault data of wind power gearbox by Bayes method and optimal confidence limit method, the relevant empirical data are collected purposefully. According to the obtained information, the optimal confidence limit method is selected to predict the reliability of several kinds of bearings in wind power gearbox under the condition of unknown shape parameter and known shape parameter respectively. Based on the reliability data, the reliability analysis and prediction of the active wind turbine and its gearbox are more in line with the actual operation of the wind turbine, which will provide a reference standard for the design, manufacture and maintenance of the wind turbine and its gearbox. This paper provides a new analysis method for the reliability analysis of domestic wind power gearbox without fault data.
【学位授予单位】：兰州理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM315

【参考文献】