大型风电机组状态监测与智能故障诊断系统研究

发布时间：2018-06-30 02:08

本文选题：风电机组 + 状态监测与故障诊断　；参考：《山西大学》2014年硕士论文

【摘要】：风力发电近年来在全球发展十分迅猛,总装机容量连年刷新纪录。在全球风电大发展的同时,产生了一系列亟待解决的问题。由于风电机组长期工作在条件恶劣的环境中,气温变化大,风沙、风速、载荷变化随机、不确定等因素导致机组各部件故障频发,加之机组分布分散,难以及时有效的发现故障,并且发生故障时,维护和检修困难,停机时间长,严重影响了风电的经济性。因此,开展风电机组的状态监控与故障诊断研究,对于提高风电运营的可靠性、安全性和经济性有着重大且长远的现实意义。本文从风力发电机组状态监测的现场实际需求出发,通过对山西某集团公司的3MW风力发电机组进行实际调研,针对风机传动系统的主要故障类型和特点,自主设计研发了风电机组传动系统故障模拟实验台。在实验台的硬件基础条件上,创新故障诊断方法,针对风电机组传动系统的故障振动信号所具有的非平稳、非线性以及复杂的调制成分的特点,提出了消除了频带错乱缺陷、分解更精细、分辨率更佳的改进型节点重构小波包算法结合包络谱的故障诊断新方法。针对风电机组传动系统的早期故障振动信号的时变微弱性以及故障与征兆的非线性映射导致的故障识别困难问题提出了改进型的节点重构小波包联合PNN(概率神经网络)的故障诊断新方法。通过采用实验台的实际数据对提出的两种故障诊断新方法进行验证,证实了方法的可行性和准确性,为今后更加深入进行风电机组的状态监测和故障诊断研究提供了有利的工具和技术支持。本文还完成了风电机组状态监测系统的传感器的选型、监测点的确定以及系统的研究设计工作和现场的测试工作。参与研发的数据采集分析仪具有多通道的振动、转速等信号的高速、同步、实时采集与故障分析诊断功能和远程数据传输功能等创新性设计。自主设计的监测系统界面充分考虑了现场的实际需求,不仅具有对风机和风机具体故障所属部位提供实时故障报警和给现场工作人员提供指导建议的功能,还具有对单台或多台风机监测部位的监测指标进行单图或多图的曲线显示的功能,方便现场人员进行对比确定故障。
[Abstract]:Wind power generation has been developing rapidly in the world in recent years, with the total installed capacity breaking the record year after year. With the great development of global wind power, a series of problems need to be solved. Because the wind turbine works in the harsh environment for a long time, the factors such as large temperature change, wind sand, wind speed, random load change, uncertainty and other factors lead to the frequent failure of the unit components, and the distribution of the unit is scattered. It is difficult to find fault in time and effectively, and when failure occurs, maintenance and maintenance are difficult, and the downtime is long, which seriously affects the economy of wind power. Therefore, the research on condition monitoring and fault diagnosis of wind turbine is of great and long-term practical significance for improving the reliability, safety and economy of wind power operation. In this paper, according to the actual demand of wind turbine condition monitoring, through the actual investigation of the 3MW wind turbine of a group company in Shanxi, the main fault types and characteristics of the fan drive system are discussed. Design and development of wind turbine transmission system fault simulation test bench. Based on the basic hardware condition of the test bench, the method of fault diagnosis is innovated. Aiming at the characteristics of non-stationary, nonlinear and complex modulation components of the fault vibration signal of the wind turbine transmission system, the defect of frequency band dislocation is eliminated. Improved node reconstruction wavelet packet algorithm combined with envelope spectrum is a new fault diagnosis method with better resolution and finer decomposition. Aiming at the time-varying weak vibration signal of the early fault of wind turbine transmission system and the difficulty of fault identification caused by nonlinear mapping of fault and symptom, an improved node reconstruction wavelet packet combined with PNN (probabilistic neural network) is presented in this paper. A new method for fault diagnosis based on network. The feasibility and accuracy of the proposed two new fault diagnosis methods are verified by using the actual data of the test bench. It provides favorable tools and technical support for further research on condition monitoring and fault diagnosis of wind turbines in the future. This paper also completes the selection of sensors, the determination of monitoring points, the research and design of the system, and the field test work of the wind turbine condition monitoring system. The developed data acquisition analyzer has the innovative design of multi-channel vibration, high speed, synchronization, real-time acquisition and fault analysis and diagnosis function and remote data transmission function. The self-designed interface of the monitoring system fully takes into account the actual needs of the site. It not only has the function of providing real-time fault alarm to the fan and its specific fault parts, but also providing guidance and advice to the field workers. It also has the function of displaying the curves of single or multiple maps on the monitoring index of single or multiple typhoon machines, which is convenient for the personnel on the spot to compare and determine the faults.
【学位授予单位】：山西大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM315

【参考文献】