基于有限单元热网络法的航空发动机主轴轴承热分析

航空发动机主轴轴承温度场的精准预测对保证发动机润滑系统及整机稳定运转至关重要。针对传统热网络法精度低的问题,提出一种基于有限单元思想的有限单元热网络法。将轴承结构通过网格划分为有限个单元体,每个单元体设置一个温度节点代表该单元体的集总参数,节点间构建热阻关系并形成热网络,通过求解以温度为未知量的大型稀疏矩阵线性方程组来获得滚子轴承横截面的温度分布。该方法可以与轴承局部产热理论相结合,实现轴承局部产热的精准加载,完成温度场的精细化分析。通过与试验结果对比,本文方法对轴承外环表面中心线温度的预测结果与试验结果误差不超过13%。     

基于最小二乘法的磁力计误差补偿与校准

磁力计在行人自主导航系统中用于提供绝对航行信息,而航向的精度直接决定着系统定位的精度。针对目前磁力计标定方法中存在磁干扰等问题,提出了一种基于最小二乘法和"8"字校准法对磁力计零偏进行实时校准的方法。该方法先利用最小二乘法修正磁力计零偏,再根据磁力计的模值大小及方差作为磁干扰的判断条件,用"8"字旋转方式进行实时校准。通过仿真和实验验证,新的方法有效地解决了磁力计误差补偿问题,证明了改进误差补偿算法的可行性和有效性。     

Fundamental (f) oscillations in a magnetically coupled solar interior-atmosphere system – An analytical approach

Solar fundamental (f) acoustic mode oscillations are investigated analytically in a magnetohydrodynamic (MHD) model. The model consists of three layers in planar geometry, representing the solar interior, the magnetic atmosphere, and a transitional layer sandwiched between them. Since we focus on the fundamental mode here, we assume the plasma is incompressible. A horizontal, canopy-like, magnetic field is introduced to the atmosphere, in which degenerated slow MHD waves can exist. The global (f-mode) oscillations can couple to local atmospheric Alfvén waves, resulting, e.g., in a frequency shift of the oscillations. The dispersion relation of the global oscillation mode is derived, and is solved analytically for the thin-transitional layer approximation and for the weak-field approximation. Analytical formulae are also provided for the frequency shifts due to the presence of a thin transitional layer and a weak atmospheric magnetic field. The analytical results generally indicate that, compared to the fundamental value ($\omega =\sqrt{\mathit{gk}}$), the mode frequency is reduced by the presence of an atmosphere by a few per cent. A thin transitional layer reduces the eigen-frequencies further by about an additional hundred microhertz. Finally, a weak atmospheric magnetic field can slightly, by a few percent, increase the frequency of the eigen-mode. Stronger magnetic fields, however, can increase the f-mode frequency by even up to ten per cent, which cannot be seen in observed data. The presence of a magnetic atmosphere in the three-layer model also introduces non-permitted propagation windows in the frequency spectrum; here, f-mode oscillations cannot exist with certain values of the harmonic degree. The eigen-frequencies can be sensitive to the background physical parameters, such as an atmospheric density scale-height or the rate of the plasma density drop at the photosphere. Such information, if ever observed with high-resolution instrumentation and inverted, could help to gain further insight into solar magnetic structures by means of solar magneto-seismology, and could provide further insight into the role of magnetism in solar oscillations.     

Kink mode m = 1 in magnetic tube with discontinuous magnetic field

In this paper, we study conditions of realization and stability of kink modes with azimuthal wave numbers $m=\pm 1$ in cylindrical plasma flex with twisted magnetic field and homogeneous current along its axis. We assume permanent axial magnetic field both inside and outside the flex, surrounded by currentless plasma. Azimuthal magnetic field decreases inversely proportional to the distance from the boundary beyond the flex. We derived dispersion equations for stable and unstable modes in approximation of “thin” plasma flex. The analysis of equations has been provided for the case of discontinuous axial magnetic field on flex’s boundary. Conditions of propagation of wave modes have been defined. It was shown, that unstable modes can be implemented in certain interval of longitudinal wavenumbers. Results can be applied for the interpretation of solar magnetic tubes behavior, using measurements, provided by spacecrafts.     

一种新型陀螺的力矩器非圆性误差补偿方法

为提高磁悬浮控制敏感陀螺（MSCSG）对陀螺载体姿态的敏感精度,基于其洛伦兹力磁轴承（LFMB）的设计结构,提出了一种力矩器非圆性误差补偿方法。首先,针对一种新型双球形包络面转子MSCSG,介绍了MSCSG的结构特点与陀螺载体姿态角速度敏感原理,并分别建立了MSCSG力矩器半径误差模型、转子偏转干扰力矩模型与陀螺载体姿态角速度敏感误差模型。其次,通过实验测量了力矩器的圆度,通过MATLAB进行数据拟合得到了力矩器的非圆特性,采用勒让德多项式级数对力矩器非圆性进行了描述,并有效补偿了因力矩器非圆性误差导致的姿态角速度敏感误差。最后,对误差补偿效果进行了仿真验证,结果表明该补偿方法使陀螺载体姿态角速度敏感误差降低了83.5%。此外,本文方法还可以解决LFMB陀螺的相关共性问题。      

基于感测线圈和KPCA的电机轴承故障检测

在电机定子中嵌入感测线圈并结合核主元分析(KPCA)进行电机轴承的故障检测。8只线圈分别嵌入电机定子的前端和后端,通过分时复用形成6组反应电机状态的差动信号。针对6组感测信号与电机状态间的非线性特征,采用KPCA进行多变量分析。着重介绍分析了KPCA故障检测的算法、监测指标和步骤。最后,用机车变压器油泵电机为对象开展了试验研究,验证了所提电机轴承故障监测方法的有效性和正确性。     

径向磁气组合轴承系统的动态性能

为探究磁轴承失效时能否减轻转子跌落带来的损害,将人字槽径向动压气体轴承引入磁轴承转子系统,研究磁气组合轴承对系统动态性能的影响和动压气体轴承的支承特性。采用有限差分法和小扰动法求解气膜厚度方程和雷诺方程,研究动压气体轴承的静动态特性。对磁轴承电磁力进行了分析,采用磁轴承不完全微分PID（比例-积分-微分）控制策略,对系统进行了理论模态分析和试验模态分析,完成了系统高速旋转试验,测试了动压气体轴承在不同偏心率和转速下的承载力。结果表明,引入动压气体轴承可提高系统的动态性能,在磁轴承失效造成高速转子跌落瞬间,偏心率趋近于1,两个径向动压气体轴承能够产生较大承载力,减轻转子跌落造成的损害。      

基于ARCN模型的轴承故障诊断

提出注意力循环机制与胶囊网络融合的注意力循环胶囊网络（ARCN）的诊断模型。提取时序特征信息构建初级胶囊;自适应融合路由机制、注意力循环机制构建数字胶囊特征;基于西储大学轴承实验数据,验证了ARCN模型的准确率、鲁棒性、稳定性、收敛误差,其准确率相比Caps模型识别准确率提高1.2%、收敛误差达到0.2。基于实验仿真平台,采集正常、内环故障、外环故障和滚动体故障的振动信号,并通过小波基变换获取的时频图构建ARCN模型的数据集。仿真实验结果表明:ARCN模型下,每类故障被误诊的概率不超过总样本的1%。      

基于EMD熵特征融合的滚动轴承故障诊断方法

研究了滚动轴承故障诊断单一故障信号的局限性和故障特征的非线性,从信息融合的理论出发,利用非线性动力学参数熵作为特征,提出了基于经验模态分解(EMD)熵特征融合的方法来解决滚动轴承故障诊断问题.首先将原始信号进行EMD,利用EMD的自适应多分辨率的特点计算EMD得到的固有模态函数(IMF)信号的多种熵值,然后采用核主元分析(KPCA)对提取的状态特征进行信息融合,从而得到互补的特征,最后将提取的融合特征通过支持向量机(SVM)进行故障诊断.滚动轴承故障诊断实验表明:该方法结合了EMD、信息熵理论和KPCA强大的非线性处理能力的特点,可以进行滚动轴承故障诊断.