共查询到18条相似文献,搜索用时 59 毫秒
1.
文[1]研究了带未建模动态系统的频域辨识问题,辨识的结果为传递函数在单位圆上的有限个点估计及其误差界。本文指出文[1]定理的错误,并对其进行了更正,数字仿真表明,更正的误差界比文[1]中的误差界更精确。 相似文献
2.
不确定性系统的鲁棒辨识及其发展 总被引:1,自引:0,他引:1
对适合控制的辨识即鲁棒辨识的发展作了综述。首先给出鲁棒辨识的缘由、提法和一些基本概念;然后讨论了集员辨识的理论和发展,H∞辨识以及部分结论;最后论述了适合控制的辨识理论的进一步发展。 相似文献
3.
4.
5.
研究了一类基于动态神经网络的未知非线性多变量系统的鲁棒辨识问题,用Lyapunov稳定性理论获得了具有保护策略的鲁棒调权律,从理论上证明了被辨识的系统是鲁榛 ,辨识误差按建模误差和未建模动态收敛到一个稳定区域,该策略的特点是不需要离线学习又不需要对象的状态完全可测,仿真结果验证了提出的动态网鲁棒辨识策略的有效性。 相似文献
6.
7.
8.
一种鲁棒BP算法及其在非线性动态系统辨识中的应用 总被引:4,自引:0,他引:4
利用多层前馈神经网络的非线性建模特性,基于动态BP网络的串并联和并联模型,提出了一种高鲁棒性BP算法,与传统的BP算法相比,鲁棒BP算法有5个优点:(1)适合于非线性动态系统辨识,(2)辨识精度高;(3)不必内插所有训练样本;(4)具有高鲁棒性,能抵制过失误差和量测误差;(5)收敛速度得到了改进,因为错误差样本的影响得到了适度的抑制,把该算法用于非线性动态系统辨识,仿真结果表明此方法是有效的。 相似文献
9.
非线性系统神经网络辨识的鲁棒BP算法 总被引:2,自引:0,他引:2
讨论系统辨识神经网络算法的鲁棒性问题。通过构造新的动态鲁棒目标函数得到的RBP算法,能不断估计逼近精度,自动将品质好的样本置于强化学习域,并能有效地抵抗噪声干扰。实验结果表明,该算法具有鲁棒性强、收敛快、计算方便等特点。 相似文献
10.
11.
12.
The output tracking controller design problem is dealt with for a class of nonlinear strict-feedback form systems in the presence of nonlinear uncertainties, external disturbance, unmodelled dynamics and unknown time-varying virtual control coefficients. A new method based on signal compensation is proposed to design a linear time-invariant robust controller, which consists of a nominal controller and a robust compensator. It is shown that the closed-loop control system with a controller designed by the proposed method has robust asymptotical practical tracking property for any bounded initial conditions and robust tracking transient property if all initial states are zero. 相似文献
13.
本文提出了一种多变量鲁棒自适应前馈控制器。该控制器采用低阶模型来控制参数未知的高阶多变量系统时,在可测干扰和有界不可测干扰作用下能保证自适应系统稳定运行。当可测干扰与系统输出之间存在未建模动态时,它可以对可测干扰实行有效的动静态补偿。当只有有界干扰作用时,它能使有界干扰对系统产生的影响最小。本文还给出了采用所提出的控制器控制某钢厂间歇式余热锅炉的仿真结果。 相似文献
14.
本文提出了一种多变量鲁棒自校正解耦间接算法。它采用一种新的正则化和相对死区实现了鲁棒自校正解耦控制。这种方法可以使系统跟踪特性不受死区界和正则化因子的影响。本文还给出了算法的鲁棒性分析和仿真结果。 相似文献
15.
Araz Hashemi Ben Fitzpatrick Le Yi Wang 《International journal of systems science》2014,45(7):1563-1578
This paper investigates noise attenuation problems for systems with unmodelled dynamics and unknown noise characteristics. A unique methodology is introduced that employs signal estimation in one phase, followed by control design for noise rejection. The methodology enjoys certain advantages in its simple control design process, accommodation of unmodelled dynamics, and non-conservative noise rejection performance. Under mild information on unmodelled dynamics, we first derive robust performance bounds on noise attenuation with respect to unmodelled dynamics without noise estimation errors. Then more general results are presented for systems that are subject to both stochastic signal estimation errors and unmodelled dynamics. Examples are also presented to demonstrate our findings. 相似文献
16.
17.
This paper addresses the problem of network reconstruction from data. Previous work identified necessary and sufficient conditions for network reconstruction of LTI systems, assuming perfect measurements (no noise) and perfect system identification. This paper assumes that the conditions for network reconstruction have been met but here we additionally take into account noise and unmodelled dynamics (including nonlinearities). In order to identify the network structure that generated the data, we compute the smallest distances between the measured data and the data that would have been generated by particular network structures. We conclude with biologically inspired network reconstruction examples which include noise and nonlinearities. 相似文献
18.
本文分析具了L1误差的线性时不变系统的最不利情况的辨识问题,对系统假设的先验信息未知系统的脉冲响应函数控制稳定,并且假定实验数据具有噪声干扰的,本文提出了一般中心估计算法并分析了最不利情况的误差界限,并进一步研究了在某些特殊情况下,中心估计的简单求法及其性质,本文所给算法和辨识结果是面向鲁棒控制的。 相似文献