基于模糊神经网络的智能故障诊断专家系统

PDF(2438 KB)

振动与冲击 ›› 2017, Vol. 36 ›› Issue (4) : 164-171.

论文

基于模糊神经网络的智能故障诊断专家系统

司景萍，马继昌，牛家骅，王二毛

作者信息 +

An intelligent fault diagnosis expert system based on fuzzy neural network

SI Jiping , MA Jichang , NIU Jiahua, WANG Ermao

Author information +

文章历史 +

摘要

发动机是车辆的核心部件，及时有效地发现并排除故障,对降低维修费用,减少经济损失,增加发动机工作时的可靠性，避免事故发生具有重大的意义。以某型号发动机为研究对象，运用测试技术、信号处理、小波分析、神经网络和模糊控制理论，提出了基于模糊神经网络的智能故障诊断系统。本文建立了发动机故障信号采集试验台，在试验台上人工模拟3种转速下6种工况，通过加速度传感器采集正常工况和异常工况的振动信号，之后利用小波包技术进行消噪处理，并提取出故障信号的特征值，作为网络训练和测试的样本数据。用样本数据训练和检测自适应模糊神经网络，完成对信号的离线模式识别，之后以测试样本数据实现在线故障诊断，通过仿真分析，取得了很好的诊断效果。与传统的BP神经网络故障诊断方法进行对比，无论在诊断精度上还是学习速度上，模糊神经网络在故障诊断中更具有优势。同时，在专家系统的理论基础上，将模糊神经网络与专家系统进行信息融合，实现数据接口通信，利用网络的自学习能力建立智能故障诊断数据库和诊断规则库，通过程序语言快速高效的设计出智能诊断系统。最后，通过发动机故障诊断实例仿真分析，验证了基于模糊神经网络的智能故障诊断专家系统的可行性。

Abstract

Engine is a very important part of vehicles. Fault diagnosis and removal of faults for engine timely and effective thus has important significance,which can not only reduce maintenance costs,reduce economic losses,increase the reliability of the engine at work,but also avoid accidents. A model engine was used as an example in this study. Testing techniques,signal processing,wavelet analysis,neural networks and fuzzy control theory were applied. An intelligent fault diagnosis method based on fuzzy neural network was proposed. The paper established a fault signal acquisition engine test stand,and simulated six kinds of artificial conditions under three kinds of speed. An acceleration sensor was used to collect the vibration signals of the normal condition and abnormal conditions. And then wavelet theory was used to denoise the collected vibration signal. The extracted fault characteristic value of the signal was used as network training sample data and testing sample data. The sample data was used to train and test adaptive fuzzy neural network and complete the signal pattern recognition offline. Online fault diagnosis was then realized. Compared with the traditional BP Neural Network diagnostic methods,the fuzzy neural network has more advantages in fault diagnosis,no matter in learning speed or accuracy. At the same time,on the basis of the theory of the expert system,the fuzzy neural network information fusion was combined with the expert system. The data communication interface was implemented. The network self-learning ability was used to establish a database of intelligent fault diagnosis and the rules library of diagnosis. A fast and efficient design intelligent diagnosis system was completed through the programming language. Finally,the engine fault diagnosis example simulation analysis proved that the intelligent fault diagnosis expert system based on fuzzy neural network is feasible.

导出引用

司景萍，马继昌，牛家骅，王二毛. 基于模糊神经网络的智能故障诊断专家系统[J]. 振动与冲击, 2017, 36(4): 164-171

SI Jingping,MA Jichang,NIU Jiahua,WANG Ermao. An intelligent fault diagnosis expert system based on fuzzy neural network[J]. Journal of Vibration and Shock, 2017, 36(4): 164-171

参考文献

[1] 陈朝阳，张代胜，任佩红. 汽车故障诊断专家系统的现状与发展趋势[J].机械工程学报,2003,39(11): 01-06.
Chen Zao-yang, Zhang Dai-sheng, Ren Pei-hong. Present situation and the development trend of auto fault diagnosis expert system[J].Journal of Mechanical Engineering,2003,39
(11): 01-06.
[2] 杨朔，姚寿广等. BP神经网络在专家系统开发中的应用与研究[J].科学技术与工程, 2007,7(21)：5580-5583.
Yang Suo, Yao Shou-guang. BP neural network in the application of expert system development and research[J].
Journal of Science Technology and Engineering, 2007,7(21)：5580-5583.
[3] 吴永恒，肖涵. 基于自适应非趋势波动分析的齿轮振动信号特征提取[J]. 振动与冲击, 2015, 34(9): 154-157.
Wu Yong-heng, Xiao Han . Adaptive Detrended Fluctuation Analysis as a Feature Extraction Method to Study Gear’s Vibration Signal. Journal of Vibration and Shock, 2015, 34(9): 154-157.
[4] 蒋静芝，孟相如等. 基于Matlab/VB的网络故障诊断软件设计[J].计算机测量与控制, 2010,18(6): 1440-1443.
Jang Jing-zhi, Meng Xiang-ru,et al. Network fault diagnosis software design based on Matlab/VB[J]. Journal of Computer Measurement and Control, 2010,18(6): 1440-1443.
[5] 李国勇，杨庆佛.基于模糊神经网络的车用发动机智能故障诊断系统[J].系统仿真学报，2007,(3):1034-1037.
Li Guo-Yong,Yang Qing-Fu.Intelligent fault diagnosis system of Automotive engine based on fuzzy neural network[J].Journal of system simulation,2007,(3):1034-1037.
[6] 张云强，张培林，吴定海，等.基于最优广义S变换和脉冲耦合神经网络的轴承故障诊断[J].振动与冲击, 2015,34(9):26-31.
Zhang Yun-qiang,Zhang Pei-lin,Wu Ding-hai,et al.Bearing fault diagnosis based on optimal generalized S transform and pulse coupled neural network[J].Journal of Vibration and Shock, 2015,34(9):26-31.
[7] B. Samanta , K. R. Al-Balushi, S. A. Al-Araimi. Artificial neural networks and genetic algorithm for bearing fault detection[J]. Soft Comput, 2006,10: 264–271.
[8] 彭斌,刘振全. 基于改进小波神经网络的旋转机械故障诊断[J].振动、测试与诊断, 2007,27(1): 32-35.
Peng Bin, Liu Zhen-quan. Based on the improved wavelet neural network of rotating machinery fault diagnosis[J].Journal of Vibration、Test and Diagnosis, 2007,27(1): 32-35.
[9] 王新志，曹爽，丁海勇. VB调用神经网络工具箱在测绘软件设计中的应用[J].测绘通报, 2011(11)：59-62.
Wang Xin-zhi, Cao Shuang, Ding Hai-yong. VB call neural network toolbox in the application of surveying and mapping software design[J].Bulletin of Surveying and Mapping, 2011(11)：59-62.
[10] 王远航,邓超等.基于混合型专家系统的重型机床故障诊断[J].计算机集成制造系统,2010,16(10): 2139-2147.
Wang Yuan-hang, Deng Chao. Heavy duty machine tool fault diagnosis based on hybrid expert system[J].Journal of Computer Integrated Manufacturing System,2010,16(10): 2139-2147.