两种模糊密度确定方法的实验比较

信息融合方法是减少决策不确定性的有效途径和热点问题.本文研究模糊积分信息融合方法中的重要问题-模糊密度的确定方法,对其中两种典型的确定方法进行了细致的比较研究.基于公开而有效的13个UCI标准数据集,进行了成员分类器选择实验、不同融合方法比较实验等,并采用了描述分析、秩次分析、探测性显著性分析,最后与文献[4]中最优单分类器、文献[53中Bagging,Boosting and random forests的最优融合结果进行了对比.结果显示,基于可能度的模糊积分方法优于基于可信度的模糊积分方法、优于文献[5]中最优融合结果;基于可信度的模糊积分方法与文献[5]中最优融合结果总体相当,优于简单平均融合方法,也优于文献[4]中最优单分类器.     

一般信息结构下的Robust控制器

本文在文献[1,2,3]的基础上,直接给出了一般信息结构(GIS)下,系统的一类Robust 控制器的结构、存在条件和设计方法,避免了用间接方法[4]的要增加输出向量维数和使观测 矩阵降秩的不足之处.推广了文献[1]和[2]的集中和块对角分散信息结构下的有关结论. 若所讨论的GIS是经济信息结构[5,6,7,8],则该GIS下的Robust控制器即为经济Robust控 制器.最后举例说明本文方法的应用.     

连续状态模糊控制系统的鲁棒稳定性分析

1　引言关于模糊系统的稳定性研究 ,目前已有许多相关的分析和设计方法 [1 ] .比如 ,文献 [2 ]中证明了凡用经典控制能稳定控制的系统必可用模糊控制达到相同的稳定性能 ,文献 [3]中有关 Takagi- Sugeno模糊系统的稳定性研究 ,文献 [4 ]中有关关系模型的模糊稳定性分析 ,文献 [     

区段定理和回路增益定理

本文提出了新的区段定理和回路增益定理.这两个定理包含和改进了文献[1]中的区段 定理、文献[2]中的钝性(passivity)定理、文献[3]中的一个重新阐述的钝性定理以及文献[1] 中的定理6、4、14.     

与"观测器型H∞控制器设计的LMI方法"一文商榷

1　引言基于 2 -Riccati方程的 DGKF方法和基于 LMI的方法 ,是目前 H∞ 控制器综合的主要方法 .与早期解法 (如“1 984年”方法 )相比 ,这两种方法极大地简化了 H∞ 控制器的结构以及控制器的设计 .但相对于不含外部扰动的确定性线性系统设计相比 ,H∞ 控制器的设计仍然显得烦琐 .文献 [1 ]基于观测器设计方法 ,试图简化 H∞ 控制器设计的思路是有意义的 .然而文献 [1 ]主要结果的证明有误 ,导致其基本结论不能成立 .以下讨论均采用文献 [1 ]的记号与假定 .2　问题及修改文献 [1 ]的主要结论定理 1不能成立 .实际上 ,仅有 (4 ) (5) (…     

双环网络G(N;1,s)的L形瓦仿真算法改进

利用仿真来研究双环网络L形瓦,在国内外文献中尚不多见,文献[1]提出了一种仿真算法,文献[2]使用数组对文献[1]的算法作了改进.其不足之处是使用了过多过长的中间数组,影响了仿真速度.针对这一不足,根据L形瓦的h边特性,使用了更少更短的数组改进了文献[2]的算法,实验结果表明,改进的算法极大地提高了仿真速度.此外还给出了L形瓦的h边分布特点,并作了进一步分析.     

线性多变量系统的分散Robust控制器

文献[1]利用固定模方法,最早给出了多通道的线性系统存在分散Robust控制器的充分 必要条件.本文利用文献[3]的理论,也获得了一个充分必要条件,检验这些条件只需计算一 些矩阵的秩,因此是比较容易的.同时,本文还给出了分散鲁棒控制器的设计方法及相应的 数值例子.     

一种基于密码的层次访问控制方案及其分析

论文在研究现有访问控制方法的基础上,详细分析了文献[1]和文献[2]中利用密码技术实现层次访问控制的机制,指出了它们的优缺点。更重要的是,提出了一种新的简单而有效的利用密码技术实现层次访问控制的机制,通过与文献[1]和文献[2]中机制的比较,方案在存储和计算复杂度等方面都具有优越性。     

相空间坐标受限时的最优控制问题

对于相空间坐标受限的最优控制问题,P.B.加姆克列利德泽等得到了控制最优性的必要条件(即熟知的极大值原理)和衔接点处的“跳跃条件”.但是文献[1]中所用的数学工具不是初等的,推证过程也较繁长,不易为工程技术界所理解.本文在应用文献[1]中某些分析的基础上,采用文献[2,3]中的方法处理这一问题,作法较为简单,而且除了得到[1]中的结论外,还可得到一些新的结果,例如,文中所得到的某些充分条件以及对于线性系统“小范围”最优性的必要充分条件等.     

文后参考文献著录规则

1.专著[序号]主要责任者.题名:其他题名信息[文献类型标志].其他责任者.版本项.出版地:出版者,出版年:引文页码[引用日期].获取和访问路径.示例:[1]霍斯尼.谷物科学与工艺学原理[M].李庆龙,译.第2版.北京:中国食品出版社,1989:15-20.[2]赵耀东.新时代的工业工程师[M/OL].台北:天下文化出版社.1998[1998-09-26].http://www.ie.nthu.edu.tw/info/ie.newie.htm.2.期刊、报纸中的析出文献[序号]析出文献主要责任者.析出文献题名[文献类型标志].连续出版物题名:其他题名信息,年,卷(期):页码[引用日期].获取和访问路径.示例:[1]张旭,张通和,易钟…     

Augmented System Approach to Measurement-Based Robust Tracking

This paper considers the robust tracking problem using an augmented system approach. Our results extend previous full-state results to observer-based feedback. The system matrix is assumed to contain time-invariant uncertainty. Two cases are considered. In the first, the tracked output is available for feedback and a controller can be obtained by considering a robust stabilization problem. The resulting control law can be applied to polynomial reference signals of known order and results in zero steady-state error. In the second case, where the tracked output is not available for feedback, our approach is restricted to step references, and the steady-state error is non-zero, although bounded by a prescribed value. To obtain the controller, a robust H problem must be solved rather than a robust stabilization problem.     

Robust asymptotic model matching and its application to output synchronization of heterogeneous multi-agent systems

Output synchronization of heterogeneous multi-agent systems has been one of the most interesting cooperative control problems. This paper first gives a brief survey of the research on the problem from which we see that the problem can be solved in a two-step manner with the aid of a properly designed local reference for each agent: (i) a controller is designed for each agent to achieve the trajectory regulation of the agent output to its associated reference; (ii) network collaboration is added to achieve consensus among references. In the presence of system uncertainties, the robust trajectory regulation problem in (i) can be solved by an internal model design. In this paper, we formulate a novel robust asymptotic model matching problem which is less conservative than trajectory regulation and can be solved by a static controller not relying on an internal model. Moreover, network collaboration is designed in (ii) within the so-called output communication setting such that consensus among references occurs concurrently with robust asymptotic model matching. As a result, output synchronization of heterogeneous multi-agent systems is achieved with a novel approach.     

Robust path-based spectral clustering

Spectral clustering and path-based clustering are two recently developed clustering approaches that have delivered impressive results in a number of challenging clustering tasks. However, they are not robust enough against noise and outliers in the data. In this paper, based on M-estimation from robust statistics, we develop a robust path-based spectral clustering method by defining a robust path-based similarity measure for spectral clustering under both unsupervised and semi-supervised settings. Our proposed method is significantly more robust than spectral clustering and path-based clustering. We have performed experiments based on both synthetic and real-world data, comparing our method with some other methods. In particular, color images from the Berkeley segmentation data set and benchmark are used in the image segmentation experiments. Experimental results show that our method consistently outperforms other methods due to its higher robustness.     

鲁棒对角优势及在多变量系统鲁棒设计中应用

本文基于多变量系统奈氏阵列设计方法和鲁棒对角优势保证系统鲁棒稳定的结论,提出一种多变量系统鲁棒设计方法,该方法设计的鲁棒预补偿器使广义对象在一定摄动范围内严格符合鲁棒对角优势定义,因而系统一定是鲁棒稳定的,该方法具有保守性小,设计的控制器简单,易于工程实现等优点,用该方法对一参数不确定性工业对象进行了鲁棒系统设计,结果令人满意。     

Robust and perfect tracking of discrete-time systems

We study in this paper the problem of robust and perfect tracking for discrete-time linear multivariable systems. By robust and perfect tracking, we mean the ability of a controller to track a given reference signal with arbitrarily small settling time in the face of external disturbances and initial conditions. A set of necessary and sufficient conditions under which the proposed problem is solvable is obtained and, under these conditions, constructive algorithms are given that yield the required solutions. As is general to discrete-time systems, the solvability conditions to the above problem are quite restrictive. To relax these conditions, we propose an almost perfect tracking scheme, which is capable of tracking references precisely after certain initial steps.     

Robust pole location with experimental validation for three-phase grid-connected converters

This paper provides design and experimental validation of robust current controllers for three-phase grid-connected converters. The main objectives here are: (i) to show that a simple polytopic model can be used for designing robust controllers for predominately inductive grids; (ii) to help in the choice of the control design parameter, based on a trade-off between an upper bound of the transient settling times and the control gain sizes. Linear matrix inequality based conditions are used to design the robust control gains with lower numerical complexity than similar conditions on literature. It is shown that small values the radius of pole location lead to better bounds for the transient responses, at the price of higher control gains. Good tracking of references for the grid currents is also illustrated in practice, allowing the closed-loop system to inject active and reactive power into the grid. Simulation and experimental results prove that the system connected to the grid can provide three-phase currents complying with requirements of an important international standard.     

一种新的鲁棒子空间建模方法

针对传统子空间建模技术中存在的两个难点问题,即对训练数据中的噪音或局外点非常敏感和基于批处理方式的大尺度高维样本模型学习计算非常费时,提出了一种新的鲁棒子空间建模方法.该方法先利用基于双平方函数的鲁棒估计,基于梯度下降的学习规则和M-估计器来同时学习和估计线性模型的初始参数,自动分级检测出初始训练样本集中的样本级局外点和样本中的信号级局外点;然后利用鲁棒的增量学习来更新参数,获得可靠的子空间模型.实验证明,这种新的鲁棒子空间建模方法能有效处理不同类型的噪音数据,在学习亮度子空间模型时能有效解决亮度明显变化、遮挡、噪音污染等敏感问题,并且具有较快的学习速度.     

Robust exponential smoothing of multivariate time series

Multivariate time series may contain outliers of different types. In the presence of such outliers, applying standard multivariate time series techniques becomes unreliable. A robust version of multivariate exponential smoothing is proposed. The method is affine equivariant, and involves the selection of a smoothing parameter matrix by minimizing a robust loss function. It is shown that the robust method results in much better forecasts than the classic approach in the presence of outliers, and performs similarly when the data contain no outliers. Moreover, the robust procedure yields an estimator of the smoothing parameter less subject to downward bias. As a byproduct, a cleaned version of the time series is obtained, as is illustrated by means of a real data example.     

不确定时变系统的鲁棒学习控制算法

研究不确定性时变系统在有限时间区间上重复作业和在无限时间区间上周期作业的跟踪控制问题. 基于Lyapunov-like方法, 给出了形式简单的鲁棒迭代学习控制和鲁棒重复控制两种算法. 两种学习算法均可弥补单一控制算法的缺陷, 鲁棒控制部分被用来保证闭环系统中所有变量的有界性, 学习控制部分可有效消除系统跟踪误差, 改善系统的跟踪性能. 仿真结果验证了两种学习算法的有效性.