Abstract：Neural network has increasingly been used in a fault diagnosis system for wind turbines. The choice of input layer characteristic parameters plays an important role in solving the unstable problem in convergence during network training. First of all, the selection of input layer characteristic parameters of neural network was studied in a fault diagnosis system for wind turbines. It was identified that a wind turbine has three typical components, such as gear box, rotor and blade, in which a fault can occur frequently. Then the fault type and its mechanism were analyzed. It was shown that the frequency characteristics of the gear box can be used to characterize its fault types, a fault in the rotor can be related to an axis orbit;however the fault diagnosis of blade needs an acoustic emission system.Based on these ana-lytical results, several kinds of methods for determining the input layer characteristic parameters were proposed at last. For the gear box, the input layer characteristic parameters can be determined by means of time-frequency characteristics of a fault; the parameters for rotor faults can be reflected by its axis orbit, the parameters for the blade faults can be decided by the characteristic data, which are generated on the blade surfaces and can be detected with an acoustic emission system. This approach can provide a reference for neural network establishment of a fault diagnosis system for wind turbine units.
曹婷, 郑源. 风力机故障诊断神经网络特征参数确定方法[J]. 排灌机械工程学报, 2014, 32(3): 247-251.
Cao Ting, Zheng Yuan. Method for determining neural network characteristic parameters in fault diagnosis system for wind turbines. Journal of Drainage and Irrigation Machinery Engin, 2014, 32(3): 247-251.
[1]World wind energy association. Highlights of the world wind energy report 2008[EB/OL]. 2013-11-01.http://www.wwindea.org/home/index.php.[2]Jiang Liping, Chi Yongning, Qin Haiyan,et al. Wind energy in China [J]. IEEE Power & Energy Magazine, 2011,9(6):36-46.[3]Wilkinson M, Spianto F, Knowles M, et al. Towards the zero maintenance wind turbine[C]//Proceedings of the 41st International Universities Power Engineering Conference, 2006: 74-78.[4]张绪锦,谭剑波,韩江洪.基于 BP 神经网络的故障诊断方法[J].系统工程理论与实践,2002,22(6): 61-66. Zhang Xujin,Tan Jianbo, Han Jianghong. A method of fault diagnosis based on BP neural network[J]. Systems Engineering-Theory & Practice,2002,22(6):61-66.(in Chinese)[5]Watson S J, Xiang B J,Yang W X,et al. Condition monitoring of the power output of wind turbine generators using wavelets[J]. IEEE Transaction on Energy Conversation, 2010,25(3):715-721.[6]Ribrant J, Bertling L M. Survey of failures in wind po-wer systems with focus on Swedish wind power plants during 1997—2005[J]. IEEE Transactions on Energy Conversion, 2007, 22(1):167-173.[7]李虎.大型风电机组振动状态监测系统开发[D].北京:华北电力大学机电学院,2009.[8]王伟杰,赵学增,黄文涛.基于BP网络的故障诊断正向推理方法[J].车用发动机,2001(4):33-35. Wang Weijie, Zhao Xuezeng,Huang Wentao. Forward reasoning method for fault diagnosis based on BP neural networks[J].Vehicle Engine, 2001(4):33-35.(in Chinese)[9]李少华,赵明浩,陈杰,等.小波变换在风力机齿轮箱故障特征提取中的应用[J].振动与冲击,2010,29(S): 201-203. Li Shaohua, Zhao Minghao, Chen Jie, et al. Application of wavelet transform on fault feature extraction for wind turbine gearbox[J].Journal of Vibration and Shock, 2010,29(S): 201-203.(in Chinese)[10]王红君,刘冬生,岳有军.基于小波分析和神经网络的电机故障诊断方法研究[J].电气传动,2010,40(3): 69-73. Wang Hongjun, Liu Dongsheng, Yue Youjun. Study of the fault diagnosis method based on wavelet time and frequency analysis and the neural network in the motor[J]. Electric Drive, 2010,40(3): 69-73.(in Chinese)[11]葛哲学, 孙志强. 神经网络理论与MATLAB 2007实现[M].北京:电子工业出版社,2007:108-109.[12]王光研,许宝杰.弹性BP神经网络的旋转机械故障诊断[J].北京机械工业学院学报,2007,22(2):9-12. Wang Guangyan, Xu Baojie.Fault diagnosis for rotating machinery based on RPROP neural network[J]. Journal of Beijing Institute of Machinery, 2007,22(2):9-12.(in Chinese)[13]高晓清, 姚竹亭.基于BP神经网络的齿轮箱故障诊断[J].机械工程与自动化, 2008(3):117-118. Gao Xiaoqing,Yao Zhuting.Fault diagnosis of gear box based on BP neural network [J]. Mechanical Enginee-ring & Automation,2008(3):117-118.(in Chinese)[14]李登峰,杨晓慧.小波基础理论和应用实例[M].北京:高等教育出版社, 2010:130-143.[15]羊拯民,朱忠奎.小波包分解与神经网络相结合的变速箱齿轮故障识别[J].农业机械学报,2001,32(5):78-81. Yang Zhengmin,Zhu Zhongkui. Fault recognition of automotive transmission based on wavelet packet decomposition and neural network[J]. Transactions of the Chinese Society for Agricultural Machinery, 2001,32(5):78-81.(in Chinese)[16]李伯全,王翔.小波包分析在变速箱故障自动诊断中的应用[J]. 江苏大学学报:自然科学版,2003,24(2):11-14. Li Boquan,Wang Xiang. Application of wavelet package analysis to fault diagnosis of tractors gearbox[J]. Journal of Jiangsu University: Natural Science Edition, 2003,24(2):11-14.(in Chinese)[17]赵利美.基于小波和神经网络的旋转机械故障诊断研究[D].西安:西安科技大学理学院,2012.[18]李德江.旋转机械轴心轨迹的理论计算、提取和自动识别研究[D].济南:山东大学机械工程学院,2007.[19]Hu M K. Visual pattern recognition by moment invaria-nts[J]. IRE Transactions on Information Theory,1962,8(2):179-187.[20]葛新峰.基于混合神经网络的水电机组运行特性动态报警及故障诊断[D].南京:中国水利水电科学研究院,2011.[21]沈功田,万绍光,刘时风,等. 多层包扎压力容器的声发射检验和安全评定[J].无损检测,1997,19(3):67-70. Shen Gongtian, Wan Shaoguang, Liu Shifeng,et al. Acoustic emission test and safety evaluation of multilayer-wrapped pressure vessels[J].NDT, 1997,19(3):67-70.(in Chinese)
