Spurious results of the component mode synthesis

The natural frequencies and mode shapes of a bentbeam in flexure-torsion oscillation are examined using the modal synthesis technique. The problem involves six boundary constraint conditions and the effect of ignoring a number of these conditions are examined. The choice of generalized coordinates offer other variations in the problem. In all, 50 different cases are studied on the digital computer. These are compared with the exact solution and also with experimental results obtained in the laboratory.     

A hybrid method of component mode synthesis

A method is described for representing a structural component by means of its vibration modes. The modes used to describe the component may have the connection points to the remainder of the structure free, or fixed, or some points free and some points fixed. The modes may either be calculated or experimentally measured. Statically determined deflection influence coefficients may be used to improve the accuracy of the representation.

The advantages claimed for the method derive from the generality of the conditions under which the component modes are calculated (or measured). Thus the boundary conditions may be selected to optimize accuracy or, in the event that the modes have already been obtained, the method permits the available data to be used. Examples are presented that illustrate use of the method, and the significance of the improvements derived from static calculations.     

Hybrid component mode synthesis based on test derived data

A method of component mode synthesis in which a majority of the dynamic characteristics of individual components are determined by modal testing is described. For the purpose of testing, the component can be supported on a suitable number of rigid supports. Alternatively, the test can be carried out in a free–free configuration, in which very flexible cables are used to hold the component. In either case, a number of practical difficulties arise. A new approach to modal testing in which the fixed–fixed modes of the component are obtained from a test configuration in which the component is held by supports of arbitrary characteristics is described. A complete theoretical formulation is developed. Computer simulation studies are carried out to determine the sensitivity of the calculated mode shapes and frequencies of the assembled structure to the various measurement errors in modal testing. Some results from these studies are presented.     

Validation,performance, convergence and application of free interface component mode synthesis

During the past three decades the component mode synthesis methods were extensively presented in the literature. Unfortunately, it seems that no convergence criterion was clearly defined in the literature. A free interface method is presented in this paper and six test cases are used to evaluate the number of retained substructure modes necessary to ensure convergence. The approach is validated numerically and experimentally. Also, an application of the proposed free interface method to optimization design is presented.     

A posteriori error analysis of component mode synthesis for the elliptic eigenvalue problem

We develop a posteriori error estimates for the error associated with model reduction of elliptic eigenvalue problems using component mode synthesis (CMS). The estimates reflect to what degree each CMS subspace influence the overall error in the reduced solution. This allows for automatic error control through adaptive algorithms that determine suitable dimensions of each CMS subspace.     

Reduction of system matrices of planar beam in ANCF by component mode synthesis method

A method of reducing the system matrices of a planar flexible beam described by an absolute nodal coordinate formulation (ANCF) is presented. In this method, we focus that the bending stiffness matrix expressed by adopting a continuum mechanics approach to the ANCF beam element is constant when the axial strain is not very large. This feature allows to apply the Craig–Bampton method to the equation of motion that is composed of the independent coordinates when the constraint forces are eliminated. Four numerical examples that compare the proposed method and the conventional ANCF are demonstrated to verify the performance and accuracy of the proposed method. From these examples, it is verified that the proposed method can describe the large deformation effects such as dynamic stiffening due to the centrifugal force, as well as the conventional ANCF does. The use of this method also reduces the computing time, while maintaining an acceptable degree of accuracy for the expression characteristics of the conventional ANCF when the modal truncation number is adequately employed. This reduction in CPU time particularly pronounced in the case of a large element number and small modal truncation number; the reduction can be verified not only in the case of small deformation but also in the case of a fair bit large deformation.     

The application of component mode synthesis for the dynamic analysis of complex structures using ADINA

This paper describes how a complex structure is divided into various kinds of substructures which are calculated independently using the ADINA program and, following this, the dynamic characteristics of the entire structure are recombined by the bi-consistent dynamic substructure method which is a free-interface method improved by the authors. Because both the displacement consistent of substructure interfaces and the interface force equilibrium are satisfied by the present method, all interface degrees of freedom in the synthesis equations can be eliminated. Several examples are presented to illustrate the effectiveness of the proposed method.     

Methods that combine finite group theory with component mode synthesis in the analysis of repetitive structures

The static and dynamic calculations of symmetric mechanical structures are considered. This type of configuration is frequently found in rotary machines, turbomachines, parabolic antennas, spatial structures, etc. The objective of this study is to present two methods and a calculation code which takes this repetitivity into account in order to minimize the time and cost of predictive calculations. Both proposed methods are applicable to bidimensional and tridimensional problems with arbitrary loadings. They are based on the theory of linear representations of finite groups and a technique of component mode synthesis.     

Improvement of component mode synthesis model for vibration analysis of hard disk drives using attachment modes

I improved the component mode synthesis (CMS) model for free and forced-vibration analyses of hard disk drives using attachment modes. The convergence and the accuracy of the proposed CMS model was improved substantially by applying an attachment mode to a FDB shaft and a pivot shaft in the stationary part model. Different formulations were used for the FDBs and the pivot bearings because of their different damping properties. In the proposed formulation, additional general coordinates corresponding to the attachment modes of the FDB shaft are introduced into the system coordinates; on the other hand, the attachment modes of the pivot shaft moderate the stiffness and damping properties of the pivot bearings. To check the improvement of the convergence and the accuracy, I performed the free and forced-vibration analyses using the previous and proposed CMS models and a full finite element (FE) model. The convergence of the natural frequencies and the frequency response function (FRF) of the disk/spindle system were extremely improved. Moreover, the FRF of the head actuator better matched the full FE model than the previous CMS model when the same number of component modes are used.     

On the selection of the mode cut-off number in component mode reduction

Dynamic analysis of very large and complicated FE structures such as FE full-vehicle structures are mainly performed with synthesized, component mode-reduced sub-models with common interfaces. The theory of component mode reduction (CMR) is well known, but it is not a simple application in real-live analyses. One of its problems, discussed in this paper, has not appeared in literature or in commercial software releases and their reference guides yet, although, in advanced computer-aided engineering, most dynamic analysts are confronted with it. The problem is related to the mode cut-off number in CMR and its enormous influence on the components reduced representation and the response solution. The mode cut-off number is the number of retained mode shapes from the components in CMR and the frequency corresponding to the highest mode is called the cut-off frequency. Ultimately the response quantities in excited vibrations, predicted from the reduced order model, are only as good as the component modes and the system modes. A proper application of the CMR technique is one of the most important success factors in today’s analysis quality in the advanced product development process across many industries. Therefore, the mode cut-off number should be considered to be a measure of analysis quality. The paper illustrates the effect of the mode cut-off number in an example from automotive industry: CMR of a body-in-white component and system mode computation of the reduced system. A list of guide lines concludes the discussion and some proposals for a stable analysis process and optimal performance are also given.     

AVS帧内预测模式改进

目前视频编码标准帧内预测主要利用邻近的左侧块,上侧块和右上侧块的像素,未采用右侧块和下侧块像素,并不是因为右侧块和下侧块和当前块相关性不强,而是因为在解码端,右侧块和下侧块需要参考当前块数据才能解码显示。针对右侧块帧内预测模式为垂直模式,可以采用右侧块对当前块预测,提出一种新的预测模式左右均值模式。以AVS为例,对编码端和解码端做出相应修改。实验结果表明,在编解码复杂度增加很少的情况下,编码信噪比增加约0.02 dB,编码码率降低约1.22%。     

改进快速全局滑模控制方法

针对全局滑模控制的动态滑模面不能在有限时间内演化为线性滑模面的缺点, 提出一种改进的全局滑模控制方法, 其动态滑模面的衰减函数由3个指数函数项组成一阶可导函数, 并能在有限时间内衰减为零. 这样, 该方法不仅具有了全局滑模控制的全局鲁棒性, 而且能够使动态滑模面在有限时间内演化为线性滑模面, 从而加快了系统的响应速度. 利用该方法实现了二阶电力系统的混沌控制, 仿真结果验证了所提出方法的有效性.     

一种改进的经验模态分解方法

针对经验模态分解中存在的端点效应及模态混叠现象,提出一种新的改进方法。利用镜像延拓方法对信号两端数据进行延拓后,结合余弦窗函数以解决端点效应对分解结果的影响,再利用高频谐波法结合掩膜信号法抑制EMD分解过程中存在的模态混叠。通过实验对比验证了该方法的有效可行性。     

Improved principal component monitoring of large-scale processes

In this work, the integration of ARMA filters into the multivariate statistical process control (MSPC) framework is presented to improve the monitoring of large-scale industrial processes. As demonstrated in the paper, such filters can remove auto-correlation from the monitored variables to avoid the production of false alarms. This is exemplified by application studies to a synthetic example from the literature and to the Tennessee Eastman benchmark process.     

构件模型的容错体系结构和策略研究

为了提高分布应用的健壮性,通常需要开发人员编写相应的容错代码。现有的CORBA构件模型通过定义构件的端口特征,以组装的方式实现代码的二进制级复用,它使用户能够快速开发和部署分布应用。在此基础上,如何在构件模型下快速灵活地建立容错应用成为一个令人关注的话题。通过设计构件模型下的容错体系结构,提供了快速灵活开发容错应用的机制,并提出了解决两种失效类型的容错策略和算法。     

带有模式选择的改进的方向性菱形搜索算法*

针对H.264/AVC标准的多模式运动估计算法编码模式复杂、计算量大的不足,提出了一种带有模式选择的方向性菱形搜索算法。基于宏块最佳编码模式的统计分析,按运动特征的不同进行分类,将运动估计的直接搜索过程改进为先进行编码模式判断,再利用改进的方向性菱形搜索算法进行块匹配搜索。实验结果表明,该算法有效地减少了搜索点数,与DS的快速搜索算法相比,编码时间减少了32%。     

Improved principal component monitoring using the local approach

This paper shows that current multivariate statistical monitoring technology may not detect incipient changes in the variable covariance structure nor changes in the geometry of the underlying variable decomposition. To overcome these deficiencies, the local approach is incorporated into the multivariate statistical monitoring framework to define two new univariate statistics for fault detection. Fault isolation is achieved by constructing a fault diagnosis chart which reveals changes in the covariance structure resulting from the presence of a fault. A theoretical analysis is presented and the proposed monitoring approach is exemplified using application studies involving recorded data from two complex industrial processes.     

改进OCB认证功能的工作模式

针对具有认证加密功能的OCB模式在单个密钥下处理大容量的数据时容易产生碰撞,降低其认证功能这一缺陷,提出对OCB认证功能校验和的生成方式的改进方案,降低了在碰撞发生时丢失认证功能的概率,在不改变原方案中速度快、效率高等优势的前提下,提高了OCB模式认证功能在碰撞发生时的安全性。     

Improved dual mode GMDH with automatic switch

This paper proposes a stable dual mode GMDH. In each layer, two types of functions are prepared and compared with each other. The output of the model is given by the superior function, which is automatically selected according to the criterion proposed in this paper. Two real social systems arc applied to check the effectiveness of this method. A modelling interval longer than the period of the real system suppresses large predictionsl errors.     

改进的广义主分量分析及在人脸识别中的应用

广义主分量分析是一种利用图像矩阵直接计算的二维主分量分析,较传统主分量分析提高了特征抽取速度及识别率.通过对广义主分量分析中的产生矩阵进行分析,并重新定义,在类间散布矩阵定义的基础上引入了径向基函数,通过调整径向基函数的系数得到更有利于分类的特征信息,获得较高的识别率.在Yale,ORL两个人脸数据库上的实验结果表明了改进方法的有效性和鲁棒性.