基于灰色神经网络的液压泵故障诊断研究

发布时间：2018-01-02 04:37

本文关键词：基于灰色神经网络的液压泵故障诊断研究　出处：《燕山大学》2011年硕士论文　论文类型：学位论文

【摘要】：随着液压技术的不断发展,液压系统在各种领域使用的频率不断增加,其功能也不断的增多,结构越来越复杂,相应的也增加了液压系统出现故障的可能性。液压泵属于连续工作元件,其结构比较复杂,容易发生故障,在近年的研究中,液压泵的故障诊断越来越受到人们的重视,随着新技术的出现,液压泵的故障诊断技术朝着自动化,智能化的方向发展,应用此类方法对液压泵进行有效地故障诊断是目前的研究热点。本文提出一种将神经网络理论和灰色理论结合的方法对泵的故障进行诊断。神经网络和灰色理论都是在故障诊断领域应用比较成熟的理论方法,将两种理论结合进行的研究也越来越多。本文将神经网络理论和灰色理论中的灰色关联度理论进行结合,应用Matlab软件的神经网络工具箱设计了一个自定义的灰色神经网络模型,对该网络的传输函数等属性进行自定义,使得该网络模型能够实现液压泵的未知故障模式和已知故障模式之间的灰色关联度并行计算,并且可以由网络计算结果判断出未知故障模式的故障类型,直接将诊断结果进行输出。本文基于材料试验机的液压油源系统,以轴向柱塞泵为对象,进行状态监测并对泵在多种工作状态下端盖部位的振动加速度信号进行采集,应用小波理论和Hilbert变换方法对信号进行处理,提取幅值域和时频域的特征向量,以特征向量作为灰色神经网络模型的输入,对此类泵的几种典型故障进行诊断,并通过Matlab软件进行仿真,证明了灰色神经网络故障诊断新方法的有效性。
[Abstract]:With the development of hydraulic technology, the frequency of hydraulic system in various fields is increasing, its function is increasing, and the structure is becoming more and more complex. It also increases the possibility of hydraulic system failure. Hydraulic pump is a continuous working component, its structure is more complex, prone to failure, in recent years in the study. People pay more and more attention to the fault diagnosis of hydraulic pump. With the emergence of new technology, the fault diagnosis technology of hydraulic pump is developing towards the direction of automation and intelligence. The application of this method to the effective fault diagnosis of hydraulic pumps is a hot topic at present. In this paper, a method combining neural network theory with grey theory is proposed to diagnose the faults of hydraulic pumps. Neural network and grey theory are mature theory methods in the field of fault diagnosis. There are more and more researches on the combination of the two theories. In this paper, we combine the neural network theory with the grey relational degree theory. A self-defined grey neural network model is designed by using the neural network toolbox of Matlab software, and the transmission function and other attributes of the network are customized. The network model can realize the parallel calculation of grey correlation degree between unknown fault mode and known fault mode of hydraulic pump, and can judge the fault type of unknown fault mode from the result of network calculation. The diagnosis results are directly outputted. In this paper, based on the hydraulic oil source system of the material testing machine, the axial piston pump is taken as the object to monitor the state and collect the vibration acceleration signal of the end cover of the pump in various working conditions. Wavelet theory and Hilbert transform are used to process the signal and extract the eigenvector of amplitude range and time-frequency domain. The eigenvector is used as the input of the grey neural network model. Several typical faults of this kind of pump are diagnosed and simulated by Matlab software, which proves the validity of the new method of fault diagnosis based on grey neural network.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH165.3

【参考文献】