基于健康样本的风电机组滚动轴承状态评估

State evaluation of wind turbine roller bearing based on normal samples

针对风电机组运行工况复杂,故障样本的缺乏,难以进行有效诊断的情况,提出了一种基于健康样本的风电机组滚动轴承状态评估方法。该方法充分利用风电机组健康状态监测数据资源,采用径向基函数插值理论,建立了风速、转速和轴承状态参数间的映射关系,通过振动偏差有效地对风电机组滚动轴承运行状态进行实时评估。数据分析表明,该方法克服了单一静态阈值报警的不足,能有效地识别出风电机组工况变化时轴承存在的异常,为风电机组轴承健康评估提供了新的思路。

Due to the complex operating condition of wind turbine and lack of its fault samples, it is diffi- cult to effectively diagnose its faults, thus a state evaluation method of wind turbine roller bearing based on normal samples is proposed in this paper. The method makes full use of monitoring data of wind tur- bine in its normal conditions and establishes a mapping between bearing condition parameters and wind speed, rotating speed by using radial basis function interpolation theory. The vibration deviation is used to effectively evaluate real-time running condition of wind turbine roller bearing. The results of data analysis show that the method can overcome the shortcomings of a single static threshold alarm and effectively identi- fy existing abnormalities of wind turbine bearing under the time-variant conditions of wind turbine. This pa- per provides a new way for state evaluation of bearings.