用极点配置方法设计滑模控制器

近年来,滑模变结构系统在控制的许多方面得到了应用。恰当地选择开关超平面的参数,对极大地改善系统的性能是非常重要的。但是,过去的方法基本上是一种基于反复凑试的方法。本文提出了一种基于极点配置概念设计开关超平面参数的新方法,建立了一个适当选择开关超平面的系统化方法,简化了系统的设计。     

基于聚类的大样本支持向量机研究

针对大样本支持向量机内存开销大、训练速度慢的缺点,本文提出了基于聚类支持向量机,运用k-mean对样本聚类,压缩样本量,构造初始超平面,筛选出靠近超平面的支持粪和可能支持向量,重新构造决策超平面。实验表明,在保持泛化精度基本一致前提下,改进算法的训练速度明显提高。     

液压伺服系统变结构滑模超平面的一种新的设计方法

滑模超平面的设计是变结构控制系统设计中重要的一环。本文通过滑模系统闭环系统极点配置在左半扇形区域内来设计滑模超平面并对－液压伺服系统进行设计。仿真结果表明该方法是行之有效的。     

正控制总量受限下离散线性系统的能控域

本文研究正控制总量受限下一类多输入离散线性系统的能控域问题.利用线性系统的分解技巧、凸集的支撑超平面与阶的估计等方法,得到了能控域的结构性定理,还给出了状态属于能控域的显式判别条件,这些结果在工程控制与社会经济系统中是有实用价值的.     

利用组合序列改善随机模拟的效果

讨论了组合序列的ｋ维结构与周期问题，指出了在相当一般的条件下Δｋ总是分布于Ｒｋ中的一组具有相同法向量的超平面上，并通过一个多重积分的数字计算实例表明了在该条件下组合序列可明显地改善随机模拟的效果。     

改进的概率选择主动支持向量机算法

针对大多数主动学习支持向量机（ASVM）的主动学习策略只注重考察超平面附近的样本,忽略了有些距离超平面远但是支持向量的样本,而且没有考虑当前超平面是否接近实际的超平面。提出一种基于概率的主动支持向量机算法,采用一个置信因子来衡量当前的超平面接近实际的超平面的程度。实验结果都验证了该算法在分类精度与计算量方面都有了较大改进。     

一种基于类中心最大间隔的支持向量机

传统的支持向量机分类超平面对噪声和野值非常敏感．使用传统的支持向量机对含有噪声的数据分类时，所得到的超平面往往不是最优超平面．为了解决这个问题，本文以两个类中心距离最大为准则建立分类超平面，构造一个新的支持向量机,称作类中心最大间隔支持向量机．理论分析和仿真实验结果证明了该方法的正确性和有效性．     

一种基于Stripe算法的新型结构型神经网络

在非线性模式识别中，设计了一种利用Stripe算法调整辅助神经元的新方法，该方法利用相互平行的超平面将权值空问进行划分，使模型的非线性模式识别能力大大提高，用Matlab随机产生非线性分类问题，利用新模型得到较好的分类识别效果，证明了网络结构简单，非线性识别能力较高。     

超平面中心的RBF神经网络及其新方法

在传统的径向基神经网络框架的基础上,通过引入中心超平面的概念,提出了超平面中心的径向基函数神经网络。在此网络中以点到中心超平面的距离代替传统的径向基神经网络中点到点的距离,其优势在于中心超平面作为数据中心包含了更多原始数据之间的信息。以函数逼近和数据分类的实验为例,证明了超平面中心的径向基神经网络相对于传统的网络有一定的优势。     

Support Vector的几何解释及其在联想记忆中的应用

利用闭凸集上的投影解释support vector的几何意义，利用支持超平面讨论线性分类器的设计问题，对线性可分情形，Support vector由一类数据集合闭凸包在另一类数据集合闭凸包上投影的非零系数向量组成，SVM所决定的超平面位于两投影点关于各自数据集合支持超平面的中间，作为应用，文中给出一种设计理想联想记忆前馈神经网络的方法，它是FP算法的一般化。     

脊波变换及其在图像处理中的应用

Ridgelet是继小波变换(Wavelet)后提出的一种新型的多尺度分析方法。对于图像中的直线状和超平面的奇异性问题,Ridgelet变换体现了比Wavelet变换更好的处理效果。文中给出了Ridgelet变换的概念及其实现算法,将Ridgelet应用于图像去噪,并和小波去噪加以比较说明其优越性。     

Containment control for distributed networks subject to multiplicative and additive noises with stochastic approximation–type protocols

This paper is concerned with the stochastic containment control problem for distributed network with multiplicative and additive noises, in which the stochastic approximation technique is utilized. A major challenge, compared with the existing results with multiplicative and additive noises, is that the involving leaders are dynamically changing. It is shown that finite boundedness for dynamical leaders is of great importance for addressing the stochastic containment control issue. A suitable approach is proposed to handle multiplicative and additive noises based on the stochastic approximation, and some necessary/sufficient conditions are derived for mean square convergence with mild requirements. Furthermore, the convergence rates are studied for both dynamical leaders and static leaders scenarios. We conclude this paper by using two illustrative examples to verify the effectiveness of the theoretical results.     

具有乘性噪声的线性离散时间随机控制系统综述

具有乘性噪声的随机不确定系统的控制问题有着广泛的应用背景. 本文概述了具有乘性噪声的线性离散时间随机系统的稳定性分析、均方镇定、最优控制以及最优估计问题和相关结论. 同时, 本文研究了具有状态与控制乘性噪声的线性多变量离散时间系统的均方镇定和最优控制问题, 分析了这两个问题之间的联系, 并讨论了最优状态反馈控制器的设计算法.     

Partial controllability of stochastic linear systems

The controllability concepts for linear stochastic differential equations, driven by different kinds of noise processes, can be reduced to the partial controllability concepts for the same systems, driven by correlated white noises. Based on this fact, in this article, we study the conditions of exact and approximate controllability for linear stochastic control systems under various kinds of noise processes, including correlated white noises as well as coloured, wide band and shifted white noises. It is proved that such systems are never exactly controllable while their approximate controllability is equivalent to the approximate controllability of the associated linear deterministic systems at all past time moments.     

Iterative Learning Control for Discrete-time Stochastic Systems with Quantized Information

An iterative learning control (ILC) algorithm using quantized error information is given in this paper for both linear and nonlinear discrete-time systems with stochastic noises. A logarithmic quantizer is used to guarantee an adaptive improvement in tracking performance. A decreasing learning gain is introduced into the algorithm to suppress the effects of stochastic noises and quantization errors. The input sequence is proved to converge strictly to the optimal input under the given index. Illustrative simulations are given to verify the theoretical analysis.      

Optimal Control of Nonlinear Stochastic Systems under Constraints: An Approximate Determination Method

An approximate method is developed for determining the optimal control for nonlinear stochastic systems under mixed equality- and inequality-type constraints on the parameters of the system, control functions, and phase coordinate in the presence of random parameters and additive and multiplicative noises. The method is based on the reduction of the initial stochastic problem to a deterministic problem for the cumulants of a random process described by stochastic differential equations.     

Leader-following consensus for single-integrator multi-agent systems with multiplicative noises in directed topologies

This paper proposes a leader-following consensus control for continuous-time single-integrator multi-agent systems with multiplicative measurement noises under directed fixed and switching topologies. The consensus controller is developed by combining the graph theory and stochastic tools. The control input for each agent relies on its own state and its neighbours’ states corrupted by noises, the noises are considered proportional to the relative distance between agents, both of the noisy case and the noise-free case are studied, and conditions to achieve mean square convergence under noisy measurement and asymptotic convergence in absence of noises are derived. Finally, in order to prove the validity of the consensus control, some simulations were carried out.     

具有不确定Wiener噪声随机非线性系统的自适应逆最优控制

针对一类具有不确定Wiener噪声扰动和未知定常参数的随机非线性系统,采用随机微分方程描述系统,基于Backstepping算法,利用随机控制Lyapunov函数,研究了自适应逆最优控制问题的可解定理,系统地给出了全局依概率渐近稳定和自适应逆最优控制策略的设计方法.这种方法可同时获得控制律和自适应律,仿真结果表明该控制算法的有效性.     

带马尔科夫跳和乘积噪声的随机系统的最优控制

讨论了N个选手随机系统的最优控制问题. 设计了无限时间的带有马尔科夫跳和乘积噪声的随机系统的Pareto最优控制器. 应用推广的Lyapunov方法和解随机Riccati代数方程得到了系统的Pareto最优解, 证明了最优控制器是稳定的反馈控制器, 以及对应于最优控制器的反馈增益中的随机Riccati代数方程的解是最小解.