灰靶理论在液压泵故障模式识别中的应用

发布时间：2018-04-20 04:23

本文选题：液压泵 + 故障诊断　；参考：《燕山大学》2011年硕士论文

【摘要】：液压系统以其功率大、响应快等优点在工程领域中得到了广泛应用,它在许多设备中起着核心控制或传动作用。液压泵是液压系统的“心脏”,对液压泵的状态监测和故障诊断成为保证液压系统正常运行的关键。现实中由于液压泵故障检测手段的不完善性、信号获取装置的不稳定性,或者缺少有效的观测工具,造成检测到得信息不完全。本文将灰靶理论分析方法用于液压系统故障诊断,利用存在的已知信息去推知含有故障模式的不可知信息的特性、状态和发展趋势,并对液压系统未来的发展做出预测和决策。本文利用LabVIEW编制了液压泵数据采集系统,并驱动NI数据采集卡,实现了液压泵信号采集。借助于MATLAB小波工具编制程序,对振动信号进行小波包分解重构,有效去除信号中的高频噪声。研究了主分量分析方法,将互相影响的液压泵复合故障信息,经过一系列变换后,在保留原始信号足够多信息量的同时,使各种故障相互独立,为进一步确定故障类型奠定了基础。对比最大熵谱估计和经典功率谱估计,将最大熵谱估计用于液压泵复合振动信号分析,在样本数据少的情况下,取得相对准确的时频信息。分析过程中没有固定的窗函数,因此可以避免传统谱分析中加窗函数的能量泄漏问题。借助MATLAB软件编制程序,提取液压泵松靴故障信号特征,得到故障特征频率。研究了灰色理论中的灰靶理论分析方法,确定信号幅值域特征为特征向量,对不同严重程度的液压泵松靴滑靴磨损复合故障进行分析,建立标准模式和故障模式,通过靶心度计算确定评估等级。
[Abstract]:Hydraulic system is widely used in engineering field because of its high power and fast response. It plays a core control or drive role in many equipments. Hydraulic pump is the "heart" of hydraulic system. The condition monitoring and fault diagnosis of hydraulic pump become the key to ensure the normal operation of hydraulic system. In reality, because of the imperfection of the fault detection method of hydraulic pump, the instability of signal acquisition device, or the lack of effective observation tools, the detected information is not complete. In this paper, the grey target theory analysis method is applied to the fault diagnosis of hydraulic system. The characteristics, status and development trend of unknowable information containing fault mode are deduced by using the known information, and the future development of hydraulic system is predicted and decided. In this paper, the hydraulic pump data acquisition system is programmed by using LabVIEW, and the NI data acquisition card is driven to realize the hydraulic pump signal acquisition. With the help of MATLAB wavelet tool, the wavelet packet decomposition and reconstruction of vibration signal are carried out, and the high frequency noise in the signal is effectively removed. In this paper, the principal component analysis (PCA) method is studied. After a series of transformations, all kinds of faults are independent of each other while retaining enough information of the original signal. It lays a foundation for further determining the fault type. Compared with the maximum entropy spectrum estimation and the classical power spectrum estimation, the maximum entropy spectrum estimation is applied to the analysis of the hydraulic pump compound vibration signal, and relatively accurate time-frequency information is obtained under the condition of less sample data. There is no fixed window function in the analysis process, so the energy leakage problem of windowed function in traditional spectral analysis can be avoided. With the help of MATLAB software, the fault signal features of hydraulic pump loose boots are extracted and the fault characteristic frequency is obtained. The grey target theory analysis method in grey theory is studied. The characteristic of signal amplitude range is determined as the characteristic vector. The wear and tear composite faults of hydraulic pump loose boots with different degrees of severity are analyzed, and the standard mode and fault mode are established. The evaluation grade is determined by the calculation of the target center.
【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】：TH137.5

【参考文献】