Abstract：To meet the needs of signal processing on pump unit fault diagnosis, the principle of invariant moment theory was introduced. In addition, the neural network modeling as well as the sample acquisition in detail was discussed. As the shape of axis orbit responded the pump unit operation is related to a variety of fault, the real-time detection swing signals of axis on invariant moment were processed according to the invariant features of translation, scaling and rotation of invariant moment. And then the shape of axis orbit was determined by using BP neural network on pattern recognize. The combination of numerical simulation and on-site test were used to compensate the shortage of neural network training samples. All samples of both processed on invariant moment and the corresponding actual shape of the samples are of the neural network training ones. After network training completed, the output was compared with the actual shape of axis loci to validate this method. Taken the fault detection and diagnosis of Dayudu Pump Station in Shanxi for example, 10 sets of data of the sample were selectd to be compared, and the results show that the neural network recognition of the results are accurate. The method can provide the basis for orbit shape automatic identification and realizing fault diagnosis system intellectualization of pump unit.
陈 坚, 叶渊杰, 陈 抒, 陈光大, 于永海, 王建明. 基于不变矩和神经网络的泵机组轴心轨迹自动识别[J]. 排灌机械工程学报, 2011, 29(1): 67-71.
Chen Jian, Ye Yuanjie, Chen Shu, Chen Guangda, Yu Yonghai, Wang Jianming. Automatic identification of pump unit axis orbit based on invariant moments features and neural networks. Journal of Drainage and Irrigation Machinery Engin, 2011, 29(1): 67-71.
付波,周建中,彭兵,等.基于仿射不变矩的轴心轨迹自动识别方法[J].华中科技大学学报:自然科学版,2007,35(3):119-122.Fu Bo,Zhou Jianzhong,Peng Bing,et al.Auto-recog-nition of shaft orbits using affine moment invariants[J].Journal of Huazhong University of Science and Technolo-gy:Nature Scienee Edition,2007,35(3):119-122.(in Chinese)彭兵,周建中,付波,等.径向基神经网络在水轮发电机组故障诊断中的应用[J].水电自动化与大坝监测,2006,30(1):35-38.Peng Bing,Zhou Jianzhong,Fu Bo,et al.Application of RBF neural networks to fault diagnosis of hydraulic turbine generator units[J].Hydropower Automation and Dam Monitoring,2006,30(1):35-38.(in Chinese)杨廷勇,施冲.基于神经网络的轴心轨迹的自动识别[J].水力发电,2002(12):34-37.Yang Tingyong,Shi Chong.Automatic identification of core path based on the neural net[J].Water Power,2002(12):34-37.(in Chinese)Lin Yong,Zhou Xiaojun,Zhang Wenbin,et al.Auto-matic identification of axis orbit based on normalized pattern spectrum[C]//Proceedings of 1st International Congress of Image and Signal Processing,CISP 2008.Sanya,Hainan,China:IEEE,2008,4:115-119.许飞云,钟秉林,黄仁.轴心轨迹自动识别及其在旋机诊断中的应用[J].振动、测试与诊断,2009,29(2):141-145.Xu Feiyun,Zhong Binglin,Huang Ren.Automatic rec-ognition of shaft orbit and its application to fault diagno-sis of rotating machinery[J].Journal of Vibration,Measurement & Diagnosis,2009,29(2):141-145.(in Chinese)Hu M K.Visual pattern recognition by moment invari-ants[J].IRE Transactions on Information Theory,1962,8(2):179-187.Song Zhenghe,Zhao Bo,Zhu Zhongxiang,et al.Re-search on traffic number recognition based on beural net-work and invariant moments[c]∥Proceedings of the Sixth International Conference or Machine Learning and Cybernetics.Hong Kong:IEEE,2007:389-393.刘厚林,王勇,谈明高,等.基于神经网络的离心泵汽蚀性能预测[J].排灌机械,2008,26(3):15-18.Liu Houlin,Wang Yong,Tan Minggao,et al.Prediction on cavitation performance of centrifugal pumps based on artificial neural network[J].Drainage and Irrigation Machinery,2008,26(3):15-18.(in Chinese)符向前,刘光临,蒋劲,等.基于不变性矩的径向基函数神经网络转子轴心轨迹自动识别[J].武汉大学学报:工学版,2003,36(2):133-136.Fu Xiangqian,Liu Guanglin,Jiang Jing,et al.Auto-matic identification of axis orbit using RBF network and moment invariants[J].Engineering Journal of Wuhan University,2003,36(2):133-136.(in Chinese)Jain A K,Duin R P W,Mao Jianchang.Statistical pat-tern recognition:A review[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(1):4-37.冯胜强,赵梨,任登春.基于频谱与轴心轨迹图的汽轮机RBF神经网络故障诊断研究[J].机械与电子,2009(9):21-23.Feng Shengqiang,Zhao Li,Ren Dengchun.Study on fault diagnosis of RBF neural network for steam turbine based on spectrum graph and chart of axle center track[J].Machinery & Electronics,2009(9):21-23.(in Chinese)