磁力泵滑动轴承磨损监测系统的设计_曾培.pdfVIP

24 5 Vo.l 24 No. 5 2006 10 D rainage and Irrigation M achinery Oct. 2006 曾 培, 肖开华, 袁铁军, 孔繁余 ( , 212013) : 针对磁力泵滑动轴承磨损在线监测, 设计制作了 感器数据采集试验台, 大量采集各种工况 下的数据,综合使用 XC L, MATLAB, N UROSH LL等数据处理软件,用神经网络法和回归分析 法进行了数据分析.在此基础上, 提出了一种多 感器数据采集系统和多神经网络模型迭代求解的 方法,用该方法实现的磁力泵滑动轴承磨损检测, 可消除磁力泵轴向窜动变量的影响. 试验结果表 明测量值与实际值很接近,建立的检测系统能检测滑动轴承的间隙, 并达到 一定的测量精度. : 磁力泵; 滑动轴承; 神经网络; 磨损监测 : TH35; TH 133. 3 : A : 1005- 6254( 2006) 05- 0036- 04 Design ofwearmonitoring system for sliding bearings ofmagnetic drive pump ZEN G P ei, X IAO K ai-hua, YUAN T ie-jun, K ON G Fan-yu ( T echnical and Research Center of F luidM achinery ng ineering, JiangsuU niversity, Zhenjiang, Jiangsu 212013, China) Abstract: An expermi ental devicew ith data collecting sensors to on-linemonitor sliding bearings ofmag- netic drive pump isdeveloped andm assive data under differentwork condition are gathered. The data are processed w ith data processing softwares such as XC L, MATLAB, N UROSH LL, neuralnetwork and regression analysis. A method of mult-i sensors data acquisition system and mult-ineural network model iteration isproposed. This testingm ethod can elmi inate the effect caused by axialmove of sliding bearings ofm agnetic drive pump. The expermi ental results indicate that the observed value and the actual value agreewel.l Key words: magnetic drive pump; sliding bearing; neural network; wearmonitoring [ 1] , , . , , , . ., 20, , . , , : ,