基于盲源分离的风力发电机主轴承振声诊断研究

发布时间：2018-06-14 06:07

本文选题：风力发电机 + 主轴承　；参考：《沈阳工业大学》2014年博士论文

【摘要】：近年来，随着人类社会对能源需求的急速增长和日益严重的环境问题，煤、石油、天然气等传统能源所暴露的问题越来越突出。风能作为一种新型可再生能源，以其储量巨大，价格低廉，环境污染小的优势越来越受到人们的重视，使得风电装备也随之得到了迅速的发展。近年来随着我国大力发展风力发电事业，风电机组逐步增多，但随之而来的是风电机组事故频发，对风力发电机的状态监测和故障诊断显得尤为重要。在风电机组的各组成部件中，主轴承是最为重要，也是最容易出现故障的部件之一。而主轴承的工作状态是否正常，将直接影响到整个风电机组的正常运转。因而对风力机主轴承的状态监测和故障诊断显得十分有必要。目前针对风力机主轴承的诊断方法很多，其中最常用的是振动诊断法和声发射诊断法。但由于风力机运行环境经常十分恶劣，在运行过程中，反映其故障状态的特征信息经常淹没在噪声干扰信号之中，有效地提取其故障信息，对风力机主轴承的监测和诊断十分必要。国内外很多学者在这方面做了大量工作，如将专家系统、模糊系统、神经网络、小波变换、Hilbert-Huang变换、Wigner分布、支持向量机等方法应用于风力机主轴承的诊断之中，取得了很多有价值的研究成果，但同时也存在一些问题。鉴于此，本文采用盲源分离理论来探索风力机主轴承振动和声发射故障信号的提取方法，并做了如下工作：第一，介绍了风电技术的发展状况，阐述了课题的研究背景、研究的目的和意义，论述了风力机主轴承振动和声发射诊断的国内外研究现状，并指出本文的思路和采用的研究方法。第二，探讨了盲源分离的基本理论和盲源分离算法，主要阐述了FastICA算法和JADE算法的计算过程，并指出这些算法存在的不足之处。针对盲源分离算法存在的不足，探讨了采用粒子群优化算法对盲源分离过程进行的优化，并比较了各分离算法的性能。第三，建立了基于盲源分离的风力机主轴承振动诊断系统。首先探讨了振动信号的提取方法，，认为包络分析对振动信号的提取较为有效，然后分别对转子试验台、风力发电机试验台和实际风力发电机主轴承的振动信号进行了分离，以实现振动故障信号的特征提取。第四，建立了基于盲源分离的风力机主轴承声发射诊断系统。首先探讨了声发射信号的提取方法，认为小波分析对声发射信号的提取较为有效，然后对风力发电机主轴承的声发射信号进行了分离，以实现声发射故障信号的特征提取。第五，根据提取的风力机主轴承信号的特点，采用集成小波神经网络对风力机主轴承进行故障诊断。针对振动信号和声发射信号的特点分别设计子神经网络，并采用决策融合神经网络进行诊断信息融合，提高了诊断效率。并对诊断算法进行了软件实现，增强了诊断方法的实用性。第六，总结本文的主要结论并对相关的研究技术进行了展望。
[Abstract]:In recent years , with the rapid growth of energy demand and the growing environmental problems of human society , the problems of coal , oil , natural gas and other traditional energy sources have become more and more prominent . As a new type of renewable energy , the advantages of large reserves , low price and less environmental pollution have been paid more and more attention .

Many scholars have done a lot of work in this field , such as expert system , fuzzy system , neural network , wavelet transform , Hilbert - Huang transform , Wigner distribution , support vector machine and so on .

Firstly , the development status of wind power technology is introduced , the research background , the purpose and significance of the research are expounded , the research status of the vibration and acoustic emission diagnosis of the main bearing of the wind turbine is discussed , and the thinking and the research method adopted in this paper are pointed out .

Secondly , the basic theory of blind source separation and the blind source separation algorithm are discussed . The calculation process of FastICA algorithm and JADE algorithm is mainly discussed , and the shortcomings of these algorithms are pointed out . In view of the shortage of blind source separation algorithm , the optimization of blind source separation process using particle swarm optimization algorithm is discussed , and the performance of each separation algorithm is compared .

Third , the vibration diagnosis system of the main bearing of the wind turbine based on the blind source separation is established . Firstly , the extraction method of the vibration signal is discussed . It is concluded that the envelope analysis is effective to the extraction of the vibration signal , and then the vibration signals of the rotor test stand , the test stand of the wind generator and the main bearing of the actual wind power generator are separated respectively to realize the feature extraction of the vibration fault signal .

Fourthly , the acoustic emission diagnosis system based on blind source separation is established . Firstly , the extraction method of the acoustic emission signal is discussed , and the wavelet analysis is considered to be effective for the extraction of the acoustic emission signal , and then the acoustic emission signal of the main bearing of the wind turbine generator is separated to realize the feature extraction of the acoustic emission fault signal .

Fifth , according to the characteristics of the extracted wind turbine main bearing signal , the integrated wavelet neural network is adopted to diagnose the main bearing of the wind turbine . The sub - neural network is designed according to the characteristics of the vibration signal and the acoustic emission signal respectively , and the decision fusion neural network is adopted to carry out diagnosis information fusion , so that the diagnosis efficiency is improved . The diagnosis algorithm is implemented in software , and the practicability of the diagnosis method is enhanced .

Sixth , summarizes the main conclusions of this paper and looks forward to the relevant research techniques .
【学位授予单位】：沈阳工业大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TM315;TH165.3

【参考文献】