基于非线性状态估计的风电机组振动建模研究  

A model for wind turbine vibration based on nonlinear state estimate technique

作　　者：苏连成[1] 孙伟[1] 董金国 SU Liancheng;SUN Wei;DONG Jinguo(School of Electrical Engineering,Yanshan University,Qinhuangdao,Hebei 066004,China;Shijiazhuang New Century Patent and Trademark Firm,Ltd.,Shijiazhuang,Hebei 050000,China)

机构地区：[1]燕山大学电气工程学院,河北秦皇岛066004 [2]石家庄新世纪专利商标事务所有限公司,河北石家庄050000

出　　处：《燕山大学学报》2018年第4期331-341,共11页Journal of Yanshan University

基　　金：河北省自然科学基金资助项目(F2015203412);无损检测技术教育部重点实验室开放基金资助项目(ZD201329007)

摘　　要：本文引入振动加速度信号,与SCADA数据结合进行风电机组关键部件的状态监测及故障诊断。实验数据为某风电机组连续6个月SCADA数据,振动建模方法采用非线性状态估计技术,在风电机组故障高发部位——齿轮箱、发电机、主轴承等处加装振动加速度传感器,通过对各部位振动特性的分析,分别建立相应部位的振动模型进行故障监测与诊断,并验证其可行性。研究表明,利用非线性估计的方法建立风电机组关键部位的振动预测模型简单有效,预测精度高,为后续风电机组振动分析提供了新的思路。A condition monitoring and fault diagnosis method based on the vibration acceleration signal combined with SCADA data of the key components for wind turbines is introduced in this paper.Experimental data are from a wind turbine in six months with continuous SCADA data.Then three vibration models are established using the nonlinear state estimation technique according to the SCADA data collected during the wind turbine is working nor^nally.The three vibration models are used to describe the complex relationship between the gearbox vibration,generator vibration,main bearing vibration and the parameters that govern its behavior.These models can also be used as a reference to help detect incipient failure.The study shows NSET is an effective and simple way to model the relationship between the gearbox vibration,generator vibration,main bearing vibration with different related parameters in a high precision.At the same time it provides a new train of thought for wind vibration analysis.

关 键 词：风电机组状态监测 非线性状态估计 振动 SCADA 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]