风力发电机齿轮箱故障诊断方法研究

发布时间：2018-10-04 19:17

【摘要】：作为风力发电机的重要传动部件，齿轮箱是风力发电机组中故障率较高的部件之一，因此研究风力发电机齿轮箱故障诊断方法尤为重要。本文首先介绍了风力发电的国内外发展状况及风力机齿轮箱故障诊断方法的研究现状，阐述了时域指标诊断方法的基本理论，简述了去趋势波动分析及多重分形去趋势波动分析的基本原理。针对去趋势波动分析方法中区间长度选取的随机性和主观性，根据互信息的基本理论，提出采用符号分析的方法进行互信息计算，对去趋势波动分析的长程相关性指数进行优化，避免了区间长度对长程相关性指数的影响，，确保了去趋势波动分析改进算法的稳定性。然后根据信号故障时会出现多重分形的特征，通过计算分形图偏离程度，得到图形偏离程度（DGD）作为故障识别的时域指标。进行了滚动轴承的信号仿真，对仿真信号进行去趋势波动分析改进算法运算，得到仿真信号的分形图和DGD特征因子，根据所得分形图和DGD特征因子来分辨正常与故障仿真信号。通过对仿真信号的分辨来证明了DGD特征因子的有效性和准确性。在仿真信号中加入不同程度的噪声，验证噪声对DGD特征因子的影响，证明了噪声的加入不影响DGD特征因子对故障识别。最后将去趋势波动分析改进算法应用到实际的信号中，对齿轮箱振动信号和风力机齿轮箱振动信号进行分析，分别求取其分形图和DGD特征因子，并将DGD特征因子与其他时域指标进行对比。通过对比发现，峭度在对实际故障识别的过程中出现了不稳定性，而DGD特征因子对风力机齿轮箱故障识别上有很好的稳定性，证明了本文提出的DGD特征因子对风力机齿轮箱故障识别上的优越性。
[Abstract]:As an important transmission part of wind turbine, gearbox is one of the components with high failure rate in wind turbine generator, so it is very important to study the fault diagnosis method of wind turbine gearbox. This paper first introduces the development of wind power generation at home and abroad and the research status of wind turbine gearbox fault diagnosis method, and expounds the basic theory of time domain index diagnosis method. The basic principles of detrend fluctuation analysis and multifractal detrend fluctuation analysis are briefly introduced. Aiming at the randomness and subjectivity of interval length selection in the detrend fluctuation analysis method, according to the basic theory of mutual information, a symbolic analysis method is proposed to calculate mutual information. The long range correlation index of detrend volatility analysis is optimized to avoid the influence of interval length on long range correlation index and the stability of the improved algorithm is ensured. Then according to the feature of multifractal when signal fault occurs, the deviation degree of fractal graph (DGD) is obtained as the time domain index of fault identification by calculating the degree of deviation of fractal graph. The signal simulation of rolling bearing is carried out, and the improved algorithm of de-trend fluctuation analysis is applied to the simulation signal. The fractal figure and DGD characteristic factor of the simulation signal are obtained, and the normal and fault simulation signals are distinguished according to the score diagram and the DGD characteristic factor. The validity and accuracy of DGD feature factor are proved by the resolution of simulation signal. The effect of noise on DGD feature factor is verified by adding noise to the simulation signal. It is proved that the addition of noise does not affect the DGD feature factor for fault identification. Finally, the improved algorithm of detrend fluctuation analysis is applied to the actual signal, the vibration signal of the gear box and the vibration signal of the gear box of the wind turbine are analyzed, and the fractal diagram and the DGD characteristic factor are obtained, respectively. The DGD characteristic factor is compared with other time domain indexes. Through comparison, it is found that kurtosis is unstable in the process of actual fault identification, and DGD characteristic factor has good stability for wind turbine gearbox fault identification. It is proved that the DGD characteristic factor proposed in this paper is superior to the wind turbine gearbox fault identification.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM315

【参考文献】