神经自适应控制：应对未知MIMO系统与传感器故障的不确定目标跟踪

92 浏览量更新于2024-08-26 收藏 1.04MB PDF 举报

"这篇研究论文探讨了在传感器故障情况下，如何使用间接神经自适应控制方法来处理未知的多输入多输出(MIMO)系统，以跟踪不确定目标。文章着重解决了在系统动态不确定和传感器出现故障的复杂环境下，如何设计控制器使系统能够紧密跟踪目标运动轨迹的问题。" 这篇发表在《Automatica》77(2017)103–111期的研究论文，由Song Yongduan、Zhang Beibei和Zhao Kai三位作者共同完成，来自中国重庆大学自动化学院。论文主要关注的是一个具有挑战性的控制问题：在未知动态特性和传感器故障的条件下，如何设计控制器使未知的MIMO系统能够精确地跟踪不确定轨迹的目标。作者提出了一种间接的神经网络自适应控制策略，即使在传感器发生故障的情况下，也能确保不确定非线性系统的动态行为能紧密跟踪移动目标。他们建立了一个分析模型，该模型允许估计和补偿由于系统不确定性及传感器故障导致的动态变化。这种方法的关键在于，它能够处理系统模型的未知部分，同时还能应对传感器可能发生的故障情况。论文中引入了两个重要的概念：均匀最终有界性(UUB)和屏障李雅普诺夫函数(BLF)。UUB表示控制系统能够在有限时间内达到并保持在一个边界内，即使在存在不确定性的情况下也能保证系统的稳定性。而BLF则是一种特殊的李雅普诺夫函数，用于确保系统的安全性，防止系统状态进入某些禁止区域，例如当传感器出现故障时。通过使用神经网络作为非线性函数的逼近器，该策略能够在线学习和校正系统模型的未知部分，从而提高跟踪性能。同时，为了应对传感器故障，设计的控制器还包含了故障诊断和隔离机制，以减少故障对系统跟踪能力的影响。这篇研究论文为解决复杂环境中未知MIMO系统的跟踪控制问题提供了新的思路，特别是在面对不确定性动态和传感器故障时，提出的间接神经自适应控制策略有望在实际工程应用中发挥重要作用，提高系统在恶劣条件下的稳定性和可靠性。

Automatica 77 (2017) 103–111

Contents lists available at ScienceDirect

Automatica

journal homepage: www.elsevier.com/locate/automatica

Brief paper

Indirect neuroadaptive control of unknown MIMO systems tracking

uncertain target under sensor failures

✩

Yongduan Song

, Beibei Zhang, Kai Zhao

School of Automation, Chongqing University, Chongqing 400044, China

a r t i c l e i n f o

Article history:

Received 17 December 2015

Received in revised form

13 September 2016

Accepted 7 November 2016

Keywords:

Indirect neuroadaptive tracking control

Uncertain target

Sensor failures

Uniformly ultimately bounded (UUB)

Barrier Lyapunov Function (BLF)

a b s t r a c t

The problem of tracking a moving target with unknown trajectory is interesting and nontrivial. The

underlying problem becomes even more challenging if uncertain dynamics and sensor failures are

involved. This work presents an indirect adaptive neural network control strategy capable of making

uncertain multi-input multi-output nonlinear systems track a moving target with uncertain trajectory

closely despite sensor faults. An analytical model is proposed to allow the estimated (predicted) target

trajectory to be linked mathematically with the actual disguised (polluted) target trajectory, thus

facilitating the control design and stability analysis. A barrier Lyapunov function based design technique

is employed to ensure that the inputs to the neural network remain within the compact set such that

the neural network unit maintains its learning/approximating functionality during the entire process

of system operation. The proposed control scheme guarantees the boundedness of all the closed-loop

signals and the uniformly ultimately bounded stable tracking. Numerical simulation results also confirm

the effectiveness of the proposed neuroadaptive tracking control method.

1. Introduction

Most practical control problems can be formulated as trajectory

tracking which includes both regulation and stabilization as two

special cases (i.e., regulation corresponds to the case that the

desired trajectory is constant, while stabilization corresponds to

the desired trajectory being strictly zero, respectively). For this

reason tracking control problem has been widely investigated

under various scenarios (Liu, Yu, & Zhong, 2014; Wang, Su, & Hong,

2004; Wu, Wu, Luo, & Guan, 2012; Zhou, Shi, Tian, & Wang, 2015)

and (Krstic, Kanelakopoulos, & Kokotovic, 1995; Li, Tong, & Li,

2015; Meng, Yang, Jagannathan, & Sun, 2014; Na, Ren, & Zheng,

2013; Wen & Ren, 2011). However, the vast majority of the existing

tracking control methods are based on the assumption that the

desired target trajectory is known a priori and the sensoring

✩

This work was supported in part by the Major State Basic Research Development

Program 973 (No. 2012CB215202, No. 2014CB249200) and the National Natural

Science Foundation of China (No. 61134001). The material in this paper was not

presented at any conference. This paper was recommended for publication in

revised form by Associate Editor Shuzhi Sam Ge under the direction of Editor

Miroslav Krstic.

E-mail addresses: ydsong@cqu.edu.cn (Y. Song), beibei@cqu.edu.cn (B. Zhang),

zhaokai@cqu.edu.cn (K. Zhao).

Fax: +86 023 65103001.

system works healthily during the entire system operation (Deng,

Li, & Wu, 2008; Dong, Zhao, Chen, & Farrell, 2012; Ge & Wang, 2004;

Ge & Zhang, 2003; Li, Tong, & Li, 2016; Sanner & Slotine, 1992;

Spooner & Passino, 1999; Zhang, Ge, & Hang, 2000).

It should be noted that, in many practical applications, the

desired trajectory to be tracked might not be available precisely.

For instance, in missile interception, the target to be intercepted

might be disguised on purpose or the target data might be

unconsciously damaged or unexpectedly polluted, rendering

the desired trajectory unavailable for guidance and control of

the interceptor (Alfriend, Vadali, Gurfil, How, & Breger, 2009).

Likewise, in industrial robotic applications (Akella, 2005; Loria,

Dasdemir, & Jarquin, 2016; Park & Han, 2011), for a given task the

desired trace for the end effector might not be available precisely

in advance, so that only the estimated trajectory can be used for

control design.

Another scenario causing imprecise trajectory information on

the target is that the target searching equipment for acquiring the

desired trajectory for the target is defected. Furthermore, sensoring

devices might malfunction during system operation. Therefore it

is of theoretical and practical importance to investigate tracking

control problem of nonlinear dynamic systems under unknown

desired target and sensor failures. To our best knowledge, very

little effort has been made in literature in addressing this

technically important issue (Bai, Arcak, & Wen, 2009; Wang,

Huang, Wen, & Fan, 2014; Yu & Xia, 2012).

http://dx.doi.org/10.1016/j.automatica.2016.11.034

下载后可阅读完整内容，剩余8页未读，立即下载

weixin_38654382

粉丝: 1
资源: 932

神经自适应控制：应对未知MIMO系统与传感器故障的不确定目标跟踪

自适应模糊神经控制器提升MIMO系统复杂性下的控制性能

MIMO最小相位系统执行器故障的自适应容错控制策略

神经网络在MIMO非仿射系统自适应控制中的应用

基于广义模糊双曲模型的不确定MIMO非线性系统的自适应跟踪控制

一类不确定MIMO非线性系统的鲁棒自适应神经控制

MIMO不确定非线性系统的自适应模糊输出跟踪控制。

具有时变时滞和未建模动力学的不确定MIMO非线性系统的自适应跟踪控制

具有非模型动力学和输出约束的MIMO不确定非线性系统的自适应神经控制

基于神经网络MIMO非仿射系统自适应控制 (2010年)

时变延迟与死区输入的不确定MIMO系统自适应控制策略

最新资源