中立联想记忆神经网络鲁棒稳定性新准则

针对一类具有离散时滞和参数范数有界的不确定性中立联想记忆神经网络的全局渐近鲁棒稳定性问题进行了研究.通过应用范数理论和矩阵不等式分析方法,并构造合适的Lyapunov-Krasovskii泛函,推导出了与时滞无关的新稳定性判定准则,用于保证神经网络的平衡点是全局渐近鲁棒稳定的.该准则中包含的未知参数少、计算复杂度低,易于验证.仿真算例验证了新判定准则的有效性.     

一类具有Markov跳跃参数的不确定混合线性时滞系统鲁棒稳定性研究

讨论了一类具有Markov跳跃参数的不确定混合线性时滞系统的鲁棒稳定性问题．分别给出了非匹配条件下不确定部分范数上界已知时使混合线性系统以概率1渐近稳定的充分条件,和匹配条件下不确定部分范数上界未知时同样可以实现混合系统以概率1渐近稳定的鲁棒自适应控制设计方案．文章研究结果表明,此控制方案对混合线性时滞系统的不确定部分是有效的．     

利用H∞范数的最优异步自学习控制频域方法

利用H∞范数对异步自学习控制系统进行了深入的研究.基于用H∞语言重新改写的问 题提法,给出了四个引理与一个定理的严格证明,得到了渐近稳定时控制器存在的一个充分条 件,并给出了计算实例.     

方差约束了不确定线性离散时间系统的鲁棒H_∞控制(英文)

本文考虑具有H∞范数及方差约束的不确定线性离散时间系统的鲁棒控制器设计问题，即设计鲁律控制器，使闭环系统对所有可允的参数扰动保持渐近稳定，同时传递函数满足预先给定的H∞范数约束，且各状态的稳态方差值不大干预先给定值．结果表明，一种有效的代数方法可用来使不确定线性随机离散时间系统满足给定的H∞范数约束及方差约束．     

方差约束下不确定性离散时间系统的鲁棒性H控制

本文考虑具有Ｈ∞范数及方差约束的不确定线性离散时间系统的鲁棒控制设计问题，即设计鲁棒控制器，使闭环系统对所有可允的参数扰动保持渐近稳定，同时传递函数满足邓布告我给定的Ｈ∞范数约束，且各状态的稳态方差值不大于预先给定值。     

基于H∞理论的关联大系统分散鲁棒控制

基于H∞理论提出了关联大系统分散鲁棒控制的一种新的设计方法.得到的控制器不但 使闭环系统渐近稳定,而且能使闭环系统从扰动输入到被控输出的传递矩阵的H∞范数小于 某个给定的扰动变小常数,并对系统参数的不确定性具有鲁棒性.     

基于线性矩阵不等式的参数不确定性关联模糊大系统的分散鲁棒镇定

讨论了参数不确定性关联模糊大系统的分散鲁棒镇定问题,所考虑的参数不确定性满足范数有界条件.基于李雅普诺夫稳定性理论及大系统分散控制理论,采用分散化PDC(parallel distributed compensation)控制器,给出了保证该关联模糊大系统闭环渐近稳定的LMI形式的充分条件,通过MATLAB软件中的LMI工具箱可求解出这些LMI中的控制器参数.仿真例子说明了所提方法的有效性.     

具有时滞和时变系数的离散多智能体系统的一致性

研究一类具有时滞和时变系数的离散多智能体系统的一致性问题. 首先, 通过构造合适的控制协议, 并以第一个智能体的位移作为参考状态, 将原系统的一致性问题转化为误差系统中零解的渐近稳定性问题; 然后, 运用矩阵范数理论研究误差系统零解的渐近稳定性, 导出使多智能体系统实现一致的充分条件; 最后, 通过数值模拟验证了该判据的正确性和有效性.

     

一类具有连续分布时滞的抛物型系统的无记忆滑模控制器设计

研究了一类具有连续分布时滞的抛物型系统的滑模控制问题.首先,通过构造辅助函数与使用矩阵范数不等式设计了无记忆功能的滑模控制器;其次,给出了滑模运动方程指数渐近稳定的充分条件;最后给出了从任意初始位置出发的轨线到达滑动模态区的时间估计.仿真结果说明了本文方法的有效性.     

不确定离散时滞中立神经网络鲁棒稳定性分析

针对一类具有离散时滞和参数范数有界的不确定性中立神经网络的全局渐近鲁棒稳定性问题,通过应用范数和矩阵不等式分析方法,构造合适的Lyapunov-Krasovskii泛函,得到了新的与时滞无关的稳定性充分条件。该条件能够保证离散时滞中立神经网络在平衡点全局渐近鲁棒稳定。与现有文献中大多数LMI形式的稳定性准则不同,该稳定性判定准则中未知参数少且计算复杂度低,易于计算验证。最后,一个仿真算例验证了结论的有效性。     

Analysis for Cohen-Grossberg neural networks with multiple delays

The stability analysis of Cohen-Grossberg neural networks with multiple delays is given. An approach combining the Lyapunov functional with the linear matrix inequality （LMI） is taken to obtain the sufficient conditions for the globally asymptotic stability of equilibrium point. By using the properties of matrix norm, a practical corollary is derived. All results are established without assuming the differentiability and monotonicity of activation functions. The simulation samples have proved the effectiveness of the conclusions.     

Analysis for Cohen-Grossberg neural networks with multiple delays

The stability analysis of Cohen-Grossberg neural networks with multiple delays is given. An approach combining the Lyapunov functional with the linear matrix inequality (LMI) is taken to obtain the sufficient conditions for the globally asymptotic stability of equilibrium point. By using the properties of matrix norm, a practical corollary is derived. All results are established without assuming the differentiability and monotonicity of activation functions. The simulation samples have proved the effectiveness of the conclusions.     

Dynamical analysis of uncertain neural networks with multiple time delays

This paper investigates the robust stability problem for dynamical neural networks in the presence of time delays and norm-bounded parameter uncertainties with respect to the class of non-decreasing, non-linear activation functions. By employing the Lyapunov stability and homeomorphism mapping theorems together, a new delay-independent sufficient condition is obtained for the existence, uniqueness and global asymptotic stability of the equilibrium point for the delayed uncertain neural networks. The condition obtained for robust stability establishes a matrix–norm relationship between the network parameters of the neural system, which can be easily verified by using properties of the class of the positive definite matrices. Some constructive numerical examples are presented to show the applicability of the obtained result and its advantages over the previously published corresponding literature results.     

Asymptotic stability of digitally redesigned control systems

This paper analyses an asymptotic stability of a digitally redesigned control system when the states of the analogue and the digital control systems are approximately matched at every sampling point. The digital redesign is a simple method of converting a given analogue controller to an equivalent digital controller in the sense of state-matching. The concerned state-matching technique is to minimise the norm distance between the discretised closed-loop system matrix of linear analogue control system and that of linear digital control system. It is shown that (i) there exists an upper bound of the norm distance to guarantee the asymptotic stability of the digitally redesigned control system and (ii) the trajectories of the linear analogue and the linear digital control systems coincide at every sampling point if the norm distance is zero. Also, a robustness result is provided in the case that nonlinear perturbations occur in the analogue and the digital control systems. Moreover, design conditions for the developed stability analysis are proposed in terms of linear matrix inequalities.     

二阶神经网络的全局指数稳定性分析

当神经网络应用于最优化计算时,理想的情形是只有一个全局渐近稳定的平衡点,并且以指数速度趋近于平衡点,从而减少神经网络所需计算时间,二阶神经网络较一般神经网络具有更快的收敛速度,对于二阶连续型Hopfield神经网络,用Lyapunov方法讨论平衡点的全局指数稳定性,给出了平衡点全局指数稳定的几个判别准则,作为特例,获得了连续型Hopfield神经网络全局指数稳定的新判据。     

A note on stability of analog neural networks with time delays

This note presents a generalized sufficient condition which guarantees stability of analog neural networks with time delays. The condition is derived using a Lyapunov functional and the stability criterion is stated as: the equilibrium of analog neural networks with delays is globally asymptotically stable if the product of the norm of connection matrix and the maximum neuronal gain is less than one.     

具免疫时滞的HBV传染病模型的稳定性和Hopf分支

考虑了溶解性和非溶解性机制下的一类具有免疫时滞的HBV感染动力学模型。分析了无感染平衡点及感染无免疫平衡点的全局稳定性,讨论了感染免疫平衡点的局部稳定性和Hopf分支的存在条件。数值模拟结果表明：当易感细胞生成率的取值使得基本再生数满足平衡存在条件且低于某一临界值时,时滞对平衡点的稳定性没有影响;当大于该临界值时,随着时滞增大,平衡点不稳定,出现一系列Hopf分支,最终表现为周期波动模式。     

Stability regions of nonlinear autonomous dynamical systems

A topological and dynamical characterization of the stability boundaries for a fairly large class of nonlinear autonomous dynamic systems is presented. The stability boundary of a stable equilibrium point is shown to consist of the stable manifolds of all the equilibrium points (and/or closed orbits) on the stability boundary. Several necessary and sufficient conditions are derived to determine whether a given equilibrium point (or closed orbit) is on the stability boundary. A method for finding the stability region on the basis of these results is proposed. The method, when feasible, will find the exact stability region, rather than a subset of it as in the Lyapunov theory approach. Several examples are given to illustrate the theoretical prediction     

Close tracking of equilibrium paths

A method to control a generic system of nonlinear ordinary differential equations between equilibrium states is analyzed. The objective is to ensure that the system's state space trajectory closely tracks an equilibrium path. The control law is obtained via time parameterization of the corresponding equilibrium control path. Conditions which guarantee that the system's state space trajectory closely tracks the equilibrium path are proved using two approaches. One approach uses the mean value theorem, and the other uses the slowly time‐varying systems theory. Importantly, both methods provide relationships between the control rate norm and the tracking error norm. These allow computation of upper bounds on the control rate norm which guarantee a desired upper bound on the tracking error norm. They also enable computation of upper bounds on the tracking error norm for a given upper bound on the control rate norm. Examples illustrate the theoretical results.     

非对称互联神经网络平衡点的稳定性分析(Ⅱ)

在人工神经网络的应用方面,无论是联想记忆,还是神经优化算法,都毫不例外地利用了网络系统的稳定吸引子的性质。因此,都支持神经网络应用的基础理论而言,研究其动力学行为,尤其是稳定性的判断问题,是十分有意义的。