基于扩展卡尔曼滤波的磁悬浮轴承支承参数辨识

发布时间：2019-03-27 16:51

【摘要】：磁悬浮轴承作为新型支承部件，具有传统轴承无法比拟的优势，现已应用到诸多领域中。由于其刚度、阻尼参数具有在线可调性，磁悬浮轴承支承特性对系统的动态特性具有更为重要的影响，因此，磁悬浮轴承支承参数的识别是磁悬浮技术研究的重要组成。而多数辨识方法是建立在频域计算过程上的，因此，本文进行了基于时域计算过程的扩展卡尔曼滤波算法辨识支承参数的研究，分别针对不平衡激励、冲击激励下的磁悬浮轴承-转子系统进行分析，识别出磁悬浮轴承刚度与阻尼参数。通过分析利用扩展卡尔曼滤波算法识别非线性系统的状态参数的原理，结合本实验所使用的转子系统结构，将磁悬浮轴承刚度阻尼系数定为系统的状态参数，确定利用扩展卡尔曼滤波算法辨识磁悬浮轴承支承参数的过程，编制Matlab程序，分别在仿真与实验中，获取转子稳态运转时不平衡质量激励下的转子位移响应，，实现对磁悬浮轴承支承参数的辨识。同时将辨识结果与边界元辨识方法进行对比分析。将系统的激励方式改为瞬态激励，修改扩展卡尔曼滤波算法中系统输入项，分别运用Samcef仿真及搭建的基于冲击激励的磁悬浮轴承转子刚度阻尼测试与辨识试验平台进行实验，通过采集信号及数据处理获得了系统在冲击激励下的轴承处位移响应，并分别通过扩展卡尔曼滤波和传递矩阵方法辨识了磁悬浮轴承的刚度阻尼。
[Abstract]:Magnetic levitation bearing (AMB), as a new supporting component, has many advantages compared with traditional bearing, and has been applied in many fields. Because its stiffness and damping parameters are on-line adjustable, the bearing characteristics of magnetic levitation bearing have more important influence on the dynamic characteristics of the system. Therefore, the identification of bearing parameters of magnetic levitation bearing is an important part of magnetic levitation technology research. Most of the identification methods are based on the frequency domain calculation process. Therefore, the extended Kalman filter algorithm based on the time domain calculation process is studied to identify the support parameters, respectively, aiming at the unbalanced excitation. The magnetic suspension bearing-rotor system under impact excitation is analyzed and the stiffness and damping parameters of the magnetic suspension bearing are identified. By analyzing the principle of identifying the state parameters of the nonlinear system by using the extended Kalman filter algorithm and combining with the rotor system structure used in this experiment, the stiffness and damping coefficient of the magnetic levitation bearing is determined as the state parameter of the system. The process of using extended Kalman filter algorithm to identify the bearing parameters of magnetic levitation bearing is determined, and the Matlab program is compiled to obtain the rotor displacement response under unbalanced mass excitation in steady-state operation in simulation and experiment, respectively. The identification of bearing parameters of magnetic levitation bearing is realized. At the same time, the identification results are compared with the boundary element identification method. The excitation mode of the system is changed to transient excitation, and the system input in the extended Kalman filter algorithm is modified. The test platform of stiffness and damping test and identification of magnetic suspension bearing rotor based on impact excitation is simulated by Samcef and the experiment platform is built to test the rotor stiffness and damping of maglev bearing. The displacement response of the system under impact excitation is obtained by collecting the signal and processing the data. The stiffness and damping of the magnetic bearing are identified by the extended Kalman filter and the transfer matrix method respectively.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TH133.3

【参考文献】