A GENERAL APPROACH TO THE CHARACTERIZATION OF CHANGES IN SYSTEM BEHAVIOUR BY OUTPUT FEEDBACK

The purpose of this paper is to formulate in a general and unified manner the change of system behaviour produced by output feedback and then to clarify from a qualitative point of view its general structure and meaning.

In this paper feedback is defined as a system transformation on the class of general linear systems represented by input-output relations. As a measure indicating the degree of how the behaviour of a system S?X×Y changes under feedback transformation three classes of systems are introduced for S and their system theoretic meaning is considered in a general setting. Furthermore, a variety problem of several structured systems under feedback transformation is investigated by characterizing these classes in terms of basic system concepts.     

An integrated robust probing motion planning and control scheme: A tube-based MPC approach

‘This paper introduces the integration of a probing scheme into a robust MPC-based robot motion planning and control algorithm. The proposed solution tackles the output-feedback tube-based MPC problem using the partially-closed loop strategy to incorporate future measurements in a computationally efficient manner. This combination will provide not only a robust controller but also avoids overly conservative planning which is a drawback of the original implementation of the output-feedback tube-based MPC. The proposed solution is composed of two controllers: (i) a nominal MPC controller with probing feature to plan a globally convergent trajectory in conjunction with active localization, and (ii) an ancillary MPC controller to stabilize the robot motion around the planned trajectory. The performance and real-time implementation of the proposed planning and control algorithms have been verified through both extensive numerical simulations and experiments with a mobile robot.     

异构车载自组网中两级资源调度机制研究

WiMAX网络和WAVE网络中资源调度方式不同,但又需要在移动网关节点处进行互通。为此,提出一种基于反馈的两级资源调度机制。在WiMAX网络中采用基于权重的轮询调度,针对WAVE网络提出一种基于反馈的EDCA接入方式,使节点各业务的发送速率契合WiMAX 网络中的调度规则。仿真结果表明,相比同类方法,该机制使系统的平均传输时延降低至20 ms,网关节点处的排队队长缩短了50%。     

面向用户体验的Web软件系统优化控制方法

在环境动态变化和系统状态无法预知的情况下,如何同时满足功能性和非功能性需求是一个极具挑战的问题。传统的方法是利用请求调度策略对We b服务请求进行优化,以解决应用的超负荷问题,但这种方法并不能完全满足QoS复杂性和灵活性的需要。基于系统管理员设置的自适应方法在很大程度上依赖于管理员的主观判断和实践经验,很多情况下人为的预估结果与实际效果存在较大差距。针对上述问题,提出了一个基于用户体验的动态QoS优化方法。该方法使用一个用户体验计算模块动态计算系统响应时间的期望值,一个外层反馈控制器动态调整虚拟资源的个数,并结合一个内层反馈控制器来动态调整各个虚拟资源处理的负载比例关系,从而实现了系统的动态QoS优化。通过与简单规则调节动态QoS的方法进行对比实验,验证了所提方法的有效性。     

复合混沌-人工鱼群混合算法的改进及性能研究

针对人工鱼群算法在寻优过程中接近最优点时收敛速度下降而难以得到精确解,优化复杂问题时易陷入局部极值的缺点,提出了一种复合混沌搜索技术与改进人工鱼群算法相结合的混合算法。该算法采用更具遍历性的组合映射产生复合混沌局部搜索方法,来避免人工鱼长时间陷入局部极值区域,从而更加精确地达到全局最优点;同时,对人工鱼引入反馈-吞食行为进行改进,改进的人工鱼群算法降低了优化后期的复杂度,并提高了优化精度,保证了收敛效率。实验结果表明,在相同参数条件下,该混合算法的收敛速度、优化精度和全局寻优能力均优于基本人工鱼群算法,实例验证了算法的有效性。     

自然循环锅炉主蒸汽压力双值动态矩阵控制器

针对自然循环锅炉主蒸汽压力对象的大时滞和非线性的特性,将双值动态矩阵控制算法应用于锅炉燃烧系统蒸汽压力回路的控制中。考虑双值动态矩阵控制算法在模型失配和干扰影响下稳态精度低的问题,引入预测前馈补偿措施。基于分布式控制系统和可编程逻辑控制器编制程序实现了双值动态矩阵控制器,所设计的双值动态矩阵控制器模块具有可变采样和双向无扰动切换等功能。实验运行结果表明,双值动态矩阵控制计算量能满足快速采样频率的实时性要求,鲁棒稳定性、稳态精度、设定值跟踪速度、抗干扰能力和动态过渡过程均满足控制要求。     

基于力反馈的手部康复训练装置研究进展

携带力反馈功能的装置进行训练已经成为手功能康复训练的重要手段.详细介绍了典型的用于手部康复训练的力反馈装置,通过对其优缺点进行比较,发现手部康复训练力反馈装置存在安全性要求高、重量较大、功能比较欠缺和实时性差的问题,并提出了相应的解决思路.最后展望了手部康复训练力反馈装置的发展趋势.     

基于压电式加速度传感器的船用振动测量仪设计

振动测量在提高船舶动力设备的使用寿命和可靠性方面已经成为船舶工程测试的一项重要课题。为了诊断船舶动力设备的机械故障并监测其运行时振动状态,设计了一种基于压电式加速度传感器,采用差分运算电路与电压并联负反馈电路实现信号的稳定转换,由巴特沃斯滤波电路等进行信号调理的船用振动测量仪。该船用振动测量仪输出的电压信号稳定,满足机械设备振动测量的精度要求,且与市面上大多数的振动分析仪匹配,已在工程中得到了大量应用。     

面向长周期的交通状态反馈诱导策略

受检测器和诱导信息板显示形式的限制,诱导信息更新周期通常较长。为此,根据备选路径交通状态而非行驶时间制定诱导策略,利用交通状态相对稳定的特点,进一步提出状态反馈诱导策略;为减少长周期与时变交通需求间的矛盾及其他误差,提出了状态与时间反馈策略相结合的混合策略。仿真实验表明,面对长周期时,状态反馈策略可以满足稳定性和准确性的双重要求,混合策略则进一步提高了诱导的准确性。     

Stabilisation and control synthesis of switching systems subject to actuator saturation

This article presents sufficient conditions for the stabilisation of switching discrete-time linear systems subject to actuator saturations. These conditions are obtained by using successively state and output feedback control laws. The obtained results are formulated in terms of linear matrix inequalities (LMIs). The saturating and non-saturating controllers are synthesised for both cases in this work. Three sets of LMIs are presented for output feedback case. Numerical examples are used to illustrate these techniques by using a linear optimisation problem subject to LMI constraints.