基于混沌吸引子不变量信息熵特征的转子故障诊断方法研究

发布时间：2018-05-11 01:25

本文选题：转子系统 + 量化特征提取　；参考：《兰州理工大学》2012年硕士论文

【摘要】：旋转机械是工业部门中应用最为广泛的一类机械设备，其核心部件为转子-轴承系统。利用振动信号对转子-轴承系统的运行进行实时监测、分析与诊断，是保证旋转机械稳定、高效运作的重要措施。但随着旋转机械转速的提高和新材料、新结构的出现，传统的线性理论已不能满足现代转子系统故障分析的需要，非线性理论和方法在转子故障诊断的研究中占有愈加重要的地位。然而，现有成果仍存在难以系统归纳可用于故障识别的非线性特征、难以对信号给出精确的量化评价结果的缺陷，一定程度上降低了故障识别的准确性。基于此，本文针对转子典型故障的动力学模型，运用混沌理论与信息熵方法，对其振动信号的非线性状态特征的量化提取问题开展了研究。主要工作内容及获得的研究结论如下： (1)通过对不同慢变参数下系统数学模型的建立和演算，完成了基于混沌理论的典型故障转子动力学行为分析，并归纳了四种故障在不同转速域内的动力学响应差异特征，且该差异特征在系统故障慢变过程下仍保持一致性，因此能够作为转子故障模式辨识所需的特征信息，为本研究所提出方法的合理性做出了证明。 (2)根据转子故障非线性振动信号处理的要求，设计了局部投影滤波的改进算法，将邻域半径和噪声子空间维数等参数的自适应选择环节加入原算法中，以充分保留强噪声背景下系统的混沌运动信号。将该算法用于转子试验台采集的故障信号降噪，获得了较为理想的结果。 (3)针对实际复杂不可建模系统的情况，通过计算不同故障下转子随转速变化的关联维数和最大Lyapunov指数，反映其不同的动力学行为状态，并利用故障转子模型仿真序列及实验采集数据的分析，对该方法的有效性进行了验证。其结果说明，系统的吸引子不变量可准确反映出其非线性运动的状态和程度，是实际复杂系统非线性特性识别的有效特征参数。 (4)对转子故障非线性特征差异较大的一阶临界转速域和高频转速域进行吸引子不变量信息熵的量化提取，，得出其熵带和熵值分量谱，作为区分四种典型故障的依据。同时，研究了利用熵值分量谱图反映故障严重程度的方法，为采用非线性理论进行转子系统的故障模式辨识提供了一种新型图像依据。研究表明，机械系统非线性动力特征的量化提取作为机械故障诊断的新趋势，具有很大的研究空间和研究价值，同时也为智能故障诊断，即数据驱动技术在故障诊断中的应用奠定了基础。
[Abstract]:Rotating machinery is the most widely used type of mechanical equipment in the industrial sector, its core component is rotor-bearing system. Using vibration signals to monitor, analyze and diagnose the rotor-bearing system in real time is an important measure to ensure the stable and efficient operation of the rotating machinery. However, with the increase of rotating speed and the emergence of new materials and new structures, the traditional linear theory can not meet the needs of modern rotor system fault analysis. Nonlinear theory and methods play an increasingly important role in rotor fault diagnosis. However, the existing results still have the defect that it is difficult to systematically sum up the nonlinear features that can be used in fault identification, and it is difficult to give accurate quantitative evaluation results for the signal, which reduces the accuracy of fault identification to a certain extent. Based on this, the dynamic model of typical rotor faults is studied in this paper. The problem of quantification extraction of nonlinear state characteristics of rotor vibration signals is studied by using chaos theory and information entropy method. The main contents and conclusions of the study are as follows: 1) through the establishment and calculation of the mathematical model of the system under different slowly varying parameters, the dynamic behavior analysis of typical fault rotor based on chaos theory is completed, and the dynamic response characteristics of four kinds of faults in different rotational speed domain are summarized. The differential feature is consistent in the process of system fault slow change, so it can be used as the characteristic information needed for rotor fault mode identification, which proves the rationality of the method proposed in this paper. 2) according to the requirement of nonlinear vibration signal processing of rotor fault, an improved algorithm of local projection filter is designed. The adaptive selection of parameters such as neighborhood radius and noise subspace dimension is added to the original algorithm. In order to fully retain the chaotic motion signal of the system under the strong noise background. The algorithm is applied to the noise reduction of the fault signal collected by the rotor test-bed, and the better results are obtained. In view of the actual complex and unmodeled system, by calculating the correlation dimension and the maximum Lyapunov exponent of rotor speed variation under different faults, the dynamic behavior of the rotor is reflected. The validity of the method is verified by analyzing the simulation sequence of the fault rotor model and the data collected from the experiment. The results show that the attractor invariant of the system can accurately reflect the state and degree of its nonlinear motion and is an effective characteristic parameter for the identification of nonlinear characteristics of practical complex systems. In this paper, the information entropy of the first order critical speed domain and the high frequency speed domain of rotor fault is extracted by quantifying the information entropy of the attractor invariant, and the entropy band and entropy component spectrum are obtained, which can be used as the basis for distinguishing the four typical faults. At the same time, the method of using entropy component spectrum to reflect the degree of fault severity is studied, which provides a new image basis for fault mode identification of rotor system based on nonlinear theory. As a new trend of mechanical fault diagnosis, the quantitative extraction of nonlinear dynamic features of mechanical system has great research space and value, and it is also a kind of intelligent fault diagnosis. That is, the application of data-driven technology in fault diagnosis laid the foundation.
【学位授予单位】：兰州理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH165.3

【参考文献】