摘要
针对大型风电机组复杂多变的运行工况,采用单一固定阈值来评价风电机组运行状态误报警率高的问题,采用标准模糊C均值聚类进行运行工况识别;在各个子工况空间下,利用非线性状态估计方法建立齿轮箱正常工作状态下温度模型进行轴承温度估计;采用滑动窗口残差统计方法对残差进行分析,建立残差均值报警阈值。最后以某风电机组SCADA(数据采集与监视控制系统)数据进行应用研究。结果表明:基于工况辨识的模型可以准确估计齿轮箱温度并能够降低误报警率,可以实时在线监测齿轮箱运行状态。
In the light of the problem that he operating conditions of a large-sized wind power generator unit are complex and changeable and using a single fixed threshold value to evaluate the operating state of a wind power generator unit has a high false alarm rate,the standard fuzzy C mean clustering method was used to identify the operating condition. Under various operating conditions,the non-linear state estimation method was employed to establish a temperature model for gearboxes in the normal work order to estimate the temperature of the bearings and the sliding window remainder error statistical method was adopted to conduct an analysis of the remainder error to establish a remainder error alarm threshold value. Finally,the SCADA data of a wind power generator unit was used to conduct an applied study with a relatively satisfactory result being achieved.
出处
《热能动力工程》
CAS
CSCD
北大核心
2016年第7期41-46,133,共6页
Journal of Engineering for Thermal Energy and Power
基金
国家自然科学基金资助项目(61203107)
关键词
风电机组
工况辨识
状态监测
非线性状态估计
wind power generator unit
operating condition identification
state monitoring
non-linear state estimation(NSET)