半参数广义线性模型惩罚似然估计的几个性质

本文在给出了半参数广义线性模型惩罚似然估计及惩罚准似然估计的基础上,得到了下面几个性质:(Ⅰ)极小惩罚似然估计及惩罚准似然估计等价于惩罚加权最小二乘估计;(Ⅱ)估计的收敛性及加速收敛法;(Ⅲ)估计的一种 Bayes 解释。     

一种梯度交替迭代设计测量矩阵的方法

在压缩采样中,测量矩阵应该和表达字典有尽可能小的相干性,随机测量矩阵一直被使用是因为其和任何表达字典都有较小的相干性。提出一种基于梯度迭代最小化方法,作为格拉斯曼框架设计的一种变体,通过优化一个初始的随机测量矩阵来得到相干性更小的测量矩阵。仿真结果表明所设计的测量矩阵具有更好的性能。     

基于幂图的属性约简

针对粗糙集理论中基于差别矩阵的属性约简方法存在的不足,提出一种基于幂图的属性约简算法.首先通过修改样本决策属性值将不相容决策表转化为简化的相容决策表;然后将样本对概念与幂图概念相结合,将基于修正差别矩阵的不相容决策表的属性约简转化为幂图的搜索问题;最后通过实例和实验验证了所提出算法的有效性.     

SiC晶须对Si3N4陶瓷基复合材料机械强度的影响

制备了含10vol％,20vol％及30vol％SiC晶须的Si_3N_4基复合材料,对其室温及高温机械强度的测试研究表明,SiC晶须对基体的高温机械性能有明显的增强作用。含20vol％SiC晶须的复合材料1200℃下强度达780MPa,比基体材料提高50％。通过SEM和TEM研究,讨论了SiC晶须增强的原因。     

基于容差关系的不完备信息系统的属性约简

粗糙集理论是一种处理不确定性知识的有效工具,属性约简是其核心内容之一,然而对于属性值有缺省的不完备信息系统,基于等价关系的经典粗糙集理论已经不再适用。由于容差关系下的不完备信息系统的属性约简的定义与经典粗糙集的属性约简定义相似,可以用容差关系对粗糙集理论进行扩充。文中通过定义容差关系下的可辨识矩阵,运用可辨识方法,得到了一种属性约简算法;接着分析了算法的不足之处,并且在此基础上提出了增加约简效率的改进型算法;最后通过一个数值例子,说明了该算法是合理的和有效的。     

H∞ filtering for discrete-time systems subject to stochastic missing measurements: a decomposition approach

This paper deals with the problem of H_∞ filtering for discrete-time systems with stochastic missing measurements. A new missing measurement model is developed by decomposing the interval of the missing rate into several segments. The probability of the missing rate in each subsegment is governed by its corresponding random variables. We aim to design a linear full-order filter such that the estimation error converges to zero exponentially in the mean square with a less conservatism while the disturbance rejection attenuation is constrained to a given level by means of an H_∞ performance index. Based on Lyapunov theory, the reliable filter parameters are characterised in terms of the feasibility of a set of linear matrix inequalities. Finally, a numerical example is provided to demonstrate the effectiveness and applicability of the proposed design approach.     

State estimation for complex systems with randomly occurring nonlinearities and randomly missing measurements

This paper is concerned with the state estimation problem for the complex networked systems with randomly occurring nonlinearities and randomly missing measurements. The nonlinearities are included to describe the phenomena of nonlinear disturbances which exist in the network and may occur in a probabilistic way. Considering the fact that probabilistic data missing may occur in the process of information transmission, we introduce the randomly data missing into the sensor measurements. The aim of this paper is to design a state estimator to estimate the true states of the considered complex network through the available output measurements. By using a Lyapunov functional and some stochastic analysis techniques, sufficient criteria are obtained in the form of linear matrix inequalities under which the estimation error dynamics is globally asymptotically stable in the mean square. Furthermore, the state estimator gain is also obtained. Finally, a numerical example is employed to illustrate the effectiveness of the proposed state estimation conditions.     

Stability analysis and controller synthesis for discrete uncertain singular fuzzy systems with distinct difference matrices in the rules

This paper mainly studies an extended discrete singular fuzzy model incorporating the multiple difference matrices in the rules and discusses its stability and design issues. By embracing additional algebraic constraint, traditional discrete Takagi-Sugeno (T-S) fuzzy model can be extended to a generalised discrete singular Takagi-Sugeno (GDST-S) model with individual difference matrices E_i in the locally singular models, where it can describe a larger class of physical or non-linear systems. Based on the linear matrix inequality (LMI) approach, we focus on deriving some explicit stability and design criteria expressed by the LMIs for the regarded system. Thus, the stability verification and controller synthesis can be performed by the current LMI tools. Finally, some illustrative examples are given to illustrate the effectiveness and validity of the proposed approach.     

Sliding mode control for Markov jump linear uncertain systems with partly unknown transition rates

In this paper, based on sliding mode control approach, the robust stabilisation problem for a class of continuous-time Markovian jump linear uncertain systems with partly unknown transition rates is investigated. The transition rate matrix under consideration covers completely known, boundary known and completely unknown elements. By making use of linear matrix inequalities technique, sufficient conditions are presented to derive the linear switching surface and guarantee the stochastic stability of sliding mode dynamics. Then a sliding mode control law is designed to drive the state trajectory of the closed-loop system to the specified linear switching surface in finite time in spite of the existing uncertainties and unknown transition rates. Finally, an example is given to verify the validity of the theoretical results.