基于数据融合的大型风电机组故障诊断方法

项目名称： 基于数据融合的大型风电机组故障诊断方法

项目编号： No.61273172

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 杨苹

作者单位： 华南理工大学

项目金额： 80万元

中文摘要： 风力发电机组是风力发电系统的关键设备,长期运行在恶略的工作环境下，由于缺乏有效的故障诊断方法，多种常见故障无法提早预警，导致故障停机时间较长，造成巨大的经济损失。本项目通过分析风力发电机组的常见故障传播机理，从时间和空间多维度出发，从多个数据源中提取风力发电机组的故障特征；研究和建立表征风力发电机组常见故障的故障特征最小集；建立描述风力发电机组整体特性的故障分析模型；分析研究风力发电机组故障过程的动态特性，以及风力发电机组在各种运行模式下机械与电气等部件综合因素对风电机组故障特征量的影响；采用数据挖掘技术进行基于D-S证据理论的数据融合，以最少维数的故障特征量完全表征某一类常见故障，建立风电机组状态识别、故障预测、故障排除的完整解决方案。目标是建立风力发电机组状态监测与故障诊断研究的整体分析理论体系，为风力发电机组的状态监测与故障诊断的研究、设计、运行提供新的理论支持和分析手段。

中文关键词： 故障诊断；风力发电机组；数据融合；故障特征提取；

英文摘要： As the key component of wind power generation system, wind turbines operate in those harsh environments during its long-term lifetime. Due to lack of efficient diagnosis approaches, many common faults could not be detected and altered in advanced, thus results in long downtime and huge economic lost. By analyzing those common fault propagation mechanisms of wind turbine, the research of project extracts fault features of wind turbine from several data source based on multi-dimension of time and space. A minimum set of fault features describing common wind turbine faults, as well as a fault analysis module describing the overall characteristics of wind turbine, is then researched and established. It is resolved the dynamic characteristics of faults of wind turbine, and analyzed how much the fault features will be affected by mechanical and electrical combined factors under different operation modes of wind turbine. Data mining technology is adopted to conduct data fusion based on D-S evidence theory, and to fully describe one typical fault by the fault feature with minimum dimension, thus a total solution for wind turbine status identification, fault forecasting and fault elimination is then established. This project aims to establish a theoretical mechanism for resolving and analyzing wind turbine status monitor

英文关键词： Fault Diagnosis；Wind Turbine Generator System；Data Fusion；Fault features extraction；