On finite-time stability and stabilization of nonlinear port-controlled Hamiltonian systems

The finite-time stability and stabilization of nonlinear port-controlled Hamiltonian(PCH)systems are investigated in this paper,and a number of new results are proposed.Firstly,by exploiting the Hamiltonian structural properties,a proper form of the Hamiltonian function is obtained,based on which a finite-time stability criterion is then presented for a class of Hamiltonian systems.Secondly,using the obtained stability criterion and the so-called"energy shaping plus damping injection"technique,the continuous finite-time stabilization problem is studied for the PCH system,and several global stabilization results are provided.Finally,the continuous robust finite-time stabilization of the PCH system with external disturbances is investigated,and two results on designing global robust finite-time stabilizers are obtained.Study of several examples with numerical simulations shows that the control design approach developed in this paper works very well.     

A Robust Controller for A 3‐DOF Flexible Robot with a Time Variant Payload

This article describes a new control scheme designed for a three degree of freedom (3‐DOF) flexible robot. The control scheme consists of two multi variable control loops. The inner loop is the motor's position control system, while the outer loop controls the robot tip's position, thus canceling vibrations which are originated by the structural flexibility of the manipulator during movement. As it will be shown, the outer control loop is robust to payload variations. The outer loop performance is based on a perfect cancelation of the inner loop dynamics. The effects of not achieving such perfect cancelation are also studied, and rules for designing a robust controller in this case are developed. Simulations assuming different payloads have been carried out with successful results for trajectory tracking. Trajectory tracking with a variable payload is also achieved.     

基于特征参数归一化的鲁棒语音识别方法综述

目前,自动语音识别系统往往会因为环境中复杂因素的影响,造成训练环境和测试环境存在不匹配现象,使得识别系统性能大幅度下降,极大地限制了语音识别技术的应用范围。近年来,很多鲁棒语音识别技术成功地被提出,这些技术的目标都是相同的,主要是提高系统的鲁棒性,进而提高识别率。其中,基于特征的归一化技术简单而有效,常常被作为鲁棒语音识别的首选方法,它主要是通过对特征向量的统计属性、累积密度函数或功率谱的归一化来补偿环境不匹配产生的影响。该文主要对目前主流的归一化方法进行介绍,其中包括倒谱矩归一化方法、直方图均衡化方法以及调频谱归一化方法等。     

Robust fuzzy stabilization of dithered chaotic systems using island-based random optimization algorithm

Applying dither to highly nonlinear systems may suppress chaotic phenomena, but dynamic performance, such as convergence rate and disturbance attenuation, is usually not guaranteed. This paper presents a dithered H_∞ robust fuzzy control scheme to stabilize chaotic systems that ensures disturbance attenuation bounds. In the proposed scheme, Takagi-Sugeno (T-S) fuzzy linear models are used to describe the relaxed models of the dithered chaotic system, and fuzzy controllers are designed based on an extension to the concept of parallel distributed compensation (PDC). Sufficient condition for the existence of the H_∞ robust fuzzy controllers is presented in terms of a novel linear matrix inequalities (LMI) form which takes full consideration of modeling error and disturbances, but cannot be solved by the standard procedures. In order to solve the LMI problem and to identify the chaotic systems as T-S fuzzy modes, we propose a compound optimization strategy called the island-based random-walk algorithm (IRA). The algorithm is composed of a set of communicating random-walk optimization procedures concatenated with the down-hill simplex method. The design procedure and validity of the proposed scheme is demonstrated via numerical simulation of the dithered fuzzy control of a chaotic system.     

不确定随机时滞系统的鲁棒H∞保性能控制

讨论了一类具有时变时滞的不确定It类型随机系统的鲁棒H∞保性能控制问题. 运用Lyapunov-Krasovskii泛函方法, 设计使得闭环系统鲁棒随机指数均方稳定, 且具有给定H∞干扰抑制度 γ 的状态反馈保性能控制器. 控制器存在的充分条件以线性矩阵不等式(LMIs)的形式表示. 进一步, 通过求解具有LMIs约束的凸优化问题, 给出了不确定随机时滞系统的最优保性能控制器的设计方法. 最后通过一个数值例子验证了所提方法的有效性.     

An asynchronous recurrent linear threshold network approach to solving the traveling salesman problem

In this paper, an approach to solving the classical Traveling Salesman Problem (TSP) using a recurrent network of linear threshold (LT) neurons is proposed. It maps the classical TSP onto a single-layered recurrent neural network by embedding the constraints of the problem directly into the dynamics of the network. The proposed method differs from the classical Hopfield network in the update of state dynamics as well as the use of network activation function. Furthermore, parameter settings for the proposed network are obtained using a genetic algorithm, which ensure a stable convergence of the network for different problems. Simulation results illustrate that the proposed network performs better than the classical Hopfield network for optimization.     

具有圆盘极点约束的不确定切换系统的鲁棒控制

本文研究了一类具有圆盘极点约束的不确定线性切换系统的鲁棒镇定问题．对此类系统设计了状态反馈控制器,使得闭环系统的所有极点均位于一个给定的圆盘中,并得到了一个基于LMI的充分条件．通过求解线性矩阵不等式,可以构造一个具有较小反馈增益参数的状态反馈控制器．最后的仿真算例证明了本文结果的有效性。     

手功能康复机器人的鲁棒控制研究

以手功能康复机器人为研究对象,对被控对象建立数学模型,推导鲁棒控制器,基于simMechanics建立仿真模型,实现手功能康复机器人的鲁棒控制系统。考虑到手功能康复机器人是帮助偏瘫患者进行康复训练并最大限度地诱发患肢的主动作用,控制器中加入偏瘫患者的主动作用函数。控制系统在MATLAB／Simulink的环境下仿真,结果表明所设计的鲁棒控制器能适应不同程度的偏瘫患者,具有较佳的动态特性和很强的鲁棒性能。     

一类不确定采样系统的鲁棒D控制研究

本文对一类范数有界不确定线性采样系统,提出了存在状态反馈控制律,使得闭环系统的所有极点均位于一给定的圆盘中且具有最优鲁棒H∞性能的一个充分条件。结合控制律反馈增益参数极小化的要求,建立一个具有线性矩阵不等式约束的凸优化问题,通过该问题的解决,可以构造一个具有较小反馈增益参数和给定要求的控制律,算例结果进一步表明,该控制器具有更好的扰动抑制性能。