风机关键部件测点优化定位与早期故障预测算法研究

项目名称： 风机关键部件测点优化定位与早期故障预测算法研究

项目编号： No.51277074

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 电工技术

项目作者： 赵洪山

作者单位： 华北电力大学（保定）

项目金额： 83万元

中文摘要： 本项目主要针对目前风电机组故障率高、健康状态监测信息不足这一突出问题，研究风电机组关键部件状态监测点优化定位和这些部件早期故障预测方法等问题。具体研究内容：（1）鉴于设备故障可诊断性与传感器配置的数量、位置等关系密不可分，提出从风机物理结构进行分析，建立以传感器最少数量为目标，以各部件动态特性和故障可测性为约束的测点优化定位模型，并研究解该优化问题的计算方法；（2）建立风机关键部件关于故障的物理模型，研究基于物理模型的叶片/变桨距系统和传动系统的故障预测方法及相关算法的实现；（3）建立风机监测数据与故障的关系模型，研究基于数据关系模型的叶片/变桨距系统和传动系统的故障预测方法及相关算法的实现；（4）研究构建风机实时健康状况评估的智能传感器网络的相关技术问题。课题的开展将为确保风机安全、可靠运行提供理论和技术支撑，同时也为设计功能更加完善的风机提供科学依据。

中文关键词： 风电机组；测点优化定位；故障预测；；

英文摘要： Focusing on the presents of high failure rate and absence of healthy status information of current wind turbines, the proposal mainly works on the optimal placement of status monitoring transducers for critical components of wind turbines and the failure' early warning mechanism for wind turbines' safe and reliable operation. The research focuses on the following aspects: (1) It is known whether the failure of wind turbine is able to be diagnosed is closely related to the number and the position of the transducers. After a detail analysis on the physical structure of wind turbine, an optimal placement model of sensors is proposed with an objective of minimization the number of the sensors and a set of constraints of the dynamic characteristic of each component and the condition of failures' predictable. Detail study on the solving method for the model is performed. (2) The physical model on failure of key components of wind turbine is established. Research is performed on the failure prediction method and associated solving algorithm of blade/pitch system and drive system. (3) A relation model of wind turbine status monitoring data and failure is established. Research is performed, based on the relation model, on the Research is performed on the failure prediction method and associated solving algorithm of blade

英文关键词： wind turbine；optimal sensor placement；fault prediction；；