A hybrid wrapper–filter approach to detect the source(s) of out-of-control signals in multivariate manufacturing process

With modern data-acquisition equipment and on-line computers used during production, it is now common to monitor several correlated quality characteristics simultaneously in multivariate processes. Multivariate control charts (MCC) are important tools for monitoring multivariate processes. One difficulty encountered with multivariate control charts is the identification of the variable or group of variables that cause an out-of-control signal. Expert knowledge either in combination with wrapper-based supervised classifier or a pre-filter with wrapper are the standard approaches to detect the sources of out-of-control signal. However gathering expert knowledge in source identification is costly and may introduce human error. Individual univariate control charts (UCC) and decomposition of T²$T^2$ statistics are also used in many cases simultaneously to identify the sources, but these either ignore the correlations between the sources or may take more time with the increase of dimensions. The aim of this paper is to develop a source identification approach that does not need any expert-knowledge and can detect out-of-control signal in less computational complexity. We propose, a hybrid wrapper–filter based source identification approach that hybridizes a Mutual Information (MI) based Maximum Relevance (MR) filter ranking heuristic with an Artificial Neural Network (ANN) based wrapper. The Artificial Neural Network Input Gain Measurement Approximation (ANNIGMA) has been combined with MR (MR-ANNIGMA) to utilize the knowledge about the intrinsic pattern of the quality characteristics computed by the filter for directing the wrapper search process. To compute optimal ANNIGMA score, we also propose a Global MR-ANNIGMA using non-functional relationship between variables which is independent of the derivative of the objective function and has a potential to overcome the local optimization problem of ANN training. The novelty of the proposed approaches is that they combine the advantages of both filter and wrapper approaches and do not require any expert knowledge about the sources of the out-of-control signals. Heuristic score based subset generation process also reduces the search space into polynomial growth which in turns reduces computational time. The proposed approaches were tested by exhaustive experiments using both simulated and real manufacturing data and compared to existing methods including independent filter, wrapper and Multivariate EWMA (MEWMA) methods. The results indicate that the proposed approaches can identify the sources of out-of-control signals more accurately than existing approaches.     

Dismutational and Global‐Minimum Isomers of Heavier 1,4‐Dimetallatetrasilabenzenes of Group 14

Aromatic species with heavier Group 14 elements show remarkable differences in terms of stability, structure, and reactivity. Herein we report our experimental and theoretical investigations regarding isomers of germanium‐ and tin‐containing benzene analogues E₂Si₄R₆ (E=Ge, Sn). The germanium‐substituted dismutational isomer with a tricyclic six‐membered scaffold is isolable, but unlike the homonuclear Si₆ analogue slowly rearranges even at room temperature to give the propellane‐type global minimum isomer. In case of E=Sn the dismutational isomer may be an intermediate on the pathway to the propellane‐type species obtained, but cannot be detected even at low temperature. Unprecedentedly large chemical shift anisotropies in the ²⁹Si NMR spectra that increase from the Si₆ species through Ge₂Si₄ to Sn₂Si₄ are rationalized by progressively larger paramagnetic‐term contributions to the chemical shift tensor as a result of diminishing HOMO–LUMO gaps, which are also reflected in the absorption spectra, as well as by appearance and symmetry of these frontier orbitals.     

基于改进确定采样型滤波的疲劳裂纹扩展预测

针对确定性采样滤波在进行状态估计预测时随维数增加时出现计算量增加且精度不高的问题,提出一种确定采样型滤波的算法并将其应用到疲劳裂纹扩展预测当中去。首先,阐述了确定采样型滤波器的基本原理;其次,从多维数值积分的角度出发,引入完全对称积分公式,根据不变容积法则选取容积节点作为基础采样点,计算积分节点、节点个数和权重;最后,将改进后的确定采样型滤波器应用到构件疲劳裂纹损伤扩展中去,并与无迹卡尔曼滤波算法、容积卡尔曼滤波算法进行比较,提升了裂纹扩展预测的精度,仿真分析验证了该方法的可行性和有效性。     

无人水下航行器推进器故障定位研究

无人水下航行器(UUV)在复杂多变的海洋环境中作业时,推进器的故障可阻挠使命的执行,严重时还可能造成UUV的丢失与损毁。为增强UUV的安全保障能力,针对UUV多台推进器的故障定位问题,采用基于几何相关性分析的方法,设计了UUV同一平面多推进器的故障定位算法,并提出了故障定位约束条件,通过仿真实验验证了所提算法的有效性。     

Fuel cell cathode with Nafion and platinum: Calculating its overall performance with consideration of gas-and vapor-exchange processes in the gas-diffusion layer

The overall performance of cathodes with solid polymer electrolytes is calculated for three different cases in which gas-and vapor-exchange processes in the gas-diffusion layer are considered. First of all, specific results are presented for calculating the overall performance of oxygen and air cathodes under the assumption that external diffusion restrictions are completely absent. Then with these same conditions (i.e., all the parameters characterizing the active layer of the cathode remain the same), results of the overall performance are given for the idealized case in which gas exchange in the gas-diffusion layer causes the appearance of limiting currents, but in this case vapor exchange is completely ignored. Finally, an analysis is carried out to determine how vapor-exchange processes in the gas-diffusion layer alter overall cathode performance and what measures should be taken to reduce the significance of external diffusion restrictions.     

A time-domain homogenization technique for lamination stacks in dual finite element formulations

A time-domain homogenization technique is developed to take the eddy currents in lamination stacks into account with dual 3-D magnetodynamic b$\mathbf{b}$- and h$\mathbf{h}$-conform finite element formulations. The lamination stack is considered as a source region carrying predefined current density and magnetic flux density distributions describing the eddy currents and skin effect in each lamination. These distributions are related and are approximated with sub-basis functions. The stacked laminations are then converted into continuums with which terms are associated for considering the eddy current loops produced by parallel fluxes, through the homogenization of the sub-basis function contributions.     

产品质量危机事件原因挖掘模型研究

为有效解决因产品质量危机事件发生原因揭示不准确而导致应对措施选择与实施不当问题,根据产品质量危机影响因素发掘结果,提取出了产品质量危机事件可能原因,采用故障树法构建了质量危机事件潜在原因关系模型;将故障树法和贝叶斯网络相结合,基于专家调查与模糊集理论,推理出底层原因先验概率算法,并将其先验概率输入贝叶斯网络模型,获得底层原因后验概率及关键重要度,将后验概率及关键重要度共同作为质量危机事件关键原因诊断依据,构建了质量危机事件原因挖掘模型;通过实际案例验证了关键原因挖掘结果的准确性,为企业挖掘质量危机事件发生原因提供了理论依据。     

Fault identification in nonlinear hybrid systems

The problem of fault identification in hybrid systems is investigated. It is assumed that the hybrid systems under consideration consist of a finite automaton, the set of nonlinear differential equations, and so-called mode activator that coordinates the action of these two parts. To solve the fault identification problem, sliding mode observers are used. The suggested approach for constructing sliding mode observers is based on the reduced order model of the original system. This allows to reduce complexity of sliding mode observers and relax the limitations imposed on the original system. Examples illustrate details of the solution.     

基于近红外TDLAS变压器油中溶解气体在线检测装置

变压器绝缘油以链烷烃（C_nH_2n+2）为主要化学成分,在变压器长期运行过程中因电弧、放电、过热、受潮等原因导致化学键逐步发生断裂,产生与故障有关的故障判别气体（CH₄,C₂H₂,C₂H₄,C₂H₆,CO和CO₂）,因此变压器绝缘油中会溶解多组分气体,故需要一种多组分气体的在线检测装置,以保证变压器的正常运行。针对电力行业装配需求,研制基于可调谐激光吸收光谱法（TDLAS）多组分气体的在线检测装置。针对6种故障特征气体的近红外吸收波段,分别选取1 580,1 654,1 626和1 530 nm四个近红外激光器,使用分时扫描的时分多路技术,实现对多组分气体的分时快速顺序检测并采用波长调制技术,消除背景气体的交叉干扰。主要检测气体为绝缘油化学键断裂所产生的烃类化合物（CH₄,C₂H₂,C₂H₄和C₂H₆）和碳氧化合物（CO和CO₂）。在线检测,与变压器油气象色谱测量方法进行对比实验,并对其进行工况稳定性测试。实验结果表明：乙炔浓度测量范围为0.5~1 000 μL·L-1,范围小于5 μL·L-1时最大测量误差小于0.8,5~1 000 μL·L-1时最大误差在6 μL·L-1以下;甲烷、乙烷、乙烯的浓度测量范围为0.5~1 000 μL·L-1,最大测量误差小于6 μL·L-1;碳氧化合物（CO和CO₂）测量范围分别为25～5 000,25～15 000 μL·L-1,最大测量误差分别在2与20 μL·L-1以下。所设计的近红外TDLAS多组分气体检测装置能够用于变压器油中溶解气体的在线检测,测量的气体浓度满足在线检测要求,能够稳定运行且适应恶劣工况条件,为检测变压器油中溶解气体在线测量提供了有效的实践经验。