线性离散时间系统<italic style="font-style: italic">H</italic><sub>∞</sub>控制的极小极大<italic style="font-style: italic">Q</italic>-学习设计

李新兴; 奚乐乐; 查文中; 彭志红

doi:10.1631/FITEE.2000446

Your Location：

Home >

Browse articles >

线性离散时间系统H_∞控制的极小极大Q-学习设计

常规文章 | Updated：2022-03-28

- 线性离散时间系统H_∞控制的极小极大Q-学习设计
  Enhanced Publication
- Minimax Q-learning design for H_∞ control of linear discrete-time systems
- 信息与电子工程前沿（英文） 2022年23卷第3期页码：438-451
- Affiliations：
  
  1.Information Science Academy, China Electronics Technology Group Corporation, Beijing100086, China
  2.School of Automation, Beijing Institute of Technology, Beijing100081, China
  3.Peng Cheng Laboratory, Shenzhen518052, China
- Author bio：
  
  E-mail: lixinxing_1006@163.com;
  E-mail: xilele.bit@gmail.com;
  ‡Corresponding author
  E-mail: peng@bit.edu.cn
- Funds：
  
  National Natural Science Foundation of China(U1613225)
- DOI：10.1631/FITEE.2000446
  中图分类号： TP13
- 收稿：2020-08-31，
  
  录用：2021-01-10，
  
  网络出版：2022-02-04，
  
  纸质出版：2022-03-0
- Accepted：
Scan QR Code
李新兴, 奚乐乐, 查文中, 等. 线性离散时间系统H_∞控制的极小极大Q-学习设计[J]. 信息与电子工程前沿（英文）, 2022,23(3):438-451.

Xinxing LI, Lele XI, Wenzhong ZHA, et al. Minimax Q-learning design for H_∞ control of linear discrete-time systems[J]. Frontiers of Information Technology & Electronic Engineering, 2022, 23(3): 438-451.
李新兴, 奚乐乐, 查文中, 等. 线性离散时间系统H_∞控制的极小极大Q-学习设计[J]. 信息与电子工程前沿（英文）, 2022,23(3):438-451. DOI： 10.1631/FITEE.2000446.

Xinxing LI, Lele XI, Wenzhong ZHA, et al. Minimax Q-learning design for H_∞ control of linear discrete-time systems[J]. Frontiers of Information Technology & Electronic Engineering, 2022, 23(3): 438-451. DOI： 10.1631/FITEE.2000446.

摘要

∞

控制是一种消除系统扰动的有效方式，但是由于需要求解非线性哈密顿—雅克比—伊萨克斯方程，

∞

控制器往往很难得到，即便对于线性系统。本文考虑了线性离散时间系统的

∞

控制器设计问题。为求解涉及的博弈代数黎卡提方程，在离线策略算法基础上提出一种新型无模型极小极大

-学习算法，并证明离线策略迭代算法是求解博弈代数黎卡提方程的牛顿法。提出的极小极大

-学习算法采用离轨策略强化学习技术，利用行为策略产生的系统状态数据，可实现对最优控制器和最佳干扰策略的在线学习。不同于当前

-学习算法，本文提出一种基于梯度的策略提高方法。证明在一定持续激励条件下，对于初始可行的控制策略并结合合适学习率，提出的极小极大

-学习算法可收敛到鞍点策略。此外，算法收敛所需的持续激励条件可通过选择包含一定噪声激励的合适行为策略实现，且不会引起任何激励噪声偏差。将提出的极小极大

-学习算法用于受负载扰动的电力系统

∞

负载频率控制器设计，仿真结果表明，最终得到的

∞

负载频率控制器具有良好抗干扰性能。

Abstract

The

control method is an effective approach for attenuating the effect of disturbances on practical systems

but it is difficult to obtain the

controller due to the nonlinear Hamilton–Jacobi–Isaacs equation

even for linear systems. This study deals with the design of an

controller for linear discrete-time systems. To solve the related game algebraic Riccati equation (GARE)

a novel model-free minimax

-learning method is developed

on the basis of an offline policy iteration algorithm

which is shown to be Newton’s method for solving the GARE. The proposed minimax

-learning method

which employs off-policy reinforcement learning

learns the optimal control policies for the controller and the disturbance online

using only the state samples generated by the implemented behavior policies. Different from existing

-learning methods

a novel gradient-based policy improvement scheme is proposed. We prove that the minimax

-learning method converges to the saddle solution under initially admissible control policies and an appropriate positive learning rate

provided that certain persistence of excitation (PE) conditions are satisfied. In addition

the PE conditions can be easily met by choosing appropriate behavior policies containing certain excitation noises

without causing any excitation noise bias. In the simulation study

we apply the proposed minimax

-learning method to design an

load-frequency controller for an electrical power system generator that suffers from load disturbance

and the simulation results indicate that the obtained

load-frequency controller has good disturbance rejection performance.

关键词

Keywords

references

Al-Tamimi A , Lewis FL , Abu-Khalaf M , 2007 . Model-free Q -learning designs for linear discrete-time zero-sum games with application to H-infinity control . Automatica , 43 ( 3 ): 473 - 481 . doi: 10.1016/j.automatica.2006.09.019 http://doi.org/10.1016/j.automatica.2006.09.019

Başar T , Bernhard P , 1995 . H <math id="M583"><mrow><msup><mrow/><mi>∞</mi></msup></mrow></math> -Optimal Control and Related Minimax Design Problems ( 2 nd Ed.). Springer , Boston, USA .

Doyle JC , Glover K , Khargonekar PP , et al. , 1989 . State-space solutions to standard <math id="M584"><mrow><msub><mi>H</mi><mn>2</mn></msub></mrow></math> and <math id="M585"><mrow><msub><mi>H</mi><mi>∞</mi></msub></mrow></math> control problems . IEEE Trans Autom Contr , 34 ( 8 ): 831 - 847 . doi: 10.1109/9.29425 http://doi.org/10.1109/9.29425

Hansen TD , Miltersen PB , Zwick U , 2003 . Strategy iteration is strongly polynomial for 2-player turn-based stochastic games with a constant discount factor . J ACM , 60 ( 1 ): Article 1 . doi: 10.1145/2432622.2432623 http://doi.org/10.1145/2432622.2432623

He HB , Zhong XN , 2018 . Learning without external reward . IEEE Comput Intell Mag , 13 ( 3 ): 48 - 54 . doi: 10.1109/MCI.2018.2840727 http://doi.org/10.1109/MCI.2018.2840727

Ioannou PA , Fidan B , 2006 . Adaptive Control Tutorial . SIAM , Philadelphia, USA .

Kiumarsi B , Lewis FL , Jiang ZP , 2017 . H <math id="M586"><mrow><msub><mrow/><mi>∞</mi></msub></mrow></math> control of linear discrete-time systems: off-policy reinforcement learning . Automatica , 78 : 144 - 152 . doi: 10.1016/j.automatica.2016.12.009 http://doi.org/10.1016/j.automatica.2016.12.009

Kiumarsi B , Vamvoudakis KG , Modares H , et al. , 2018 . Optimal and autonomous control using reinforcement learning: a survey . IEEE Trans Neur Netw Learn Syst , 29 ( 6 ): 2042 - 2062 . doi: 10.1109/TNNLS.2017.2773458 http://doi.org/10.1109/TNNLS.2017.2773458

Li HR , Zhang QC , Zhao DB , 2020 . Deep reinforcement learning-based automatic exploration for navigation in unknown environment . IEEE Trans Neur Netw Learn Syst , 31 ( 6 ): 2064 - 2076 . doi: 10.1109/TNNLS.2019.2927869 http://doi.org/10.1109/TNNLS.2019.2927869

Li XX , Peng ZH , Jiao L , et al. , 2019 . Online adaptive Q-learning method for fully cooperative linear quadratic dynamic games . Inform Sci , 62 : 222201 . doi: 10.1007/s11432-018-9865-9 http://doi.org/10.1007/s11432-018-9865-9

Littman ML , 2001 . Value-function reinforcement learning in Markov games . Cogn Syst Res , 2 ( 1 ): 55 - 66 . doi: 10.1016/S1389-0417(01)00015-8 http://doi.org/10.1016/S1389-0417(01)00015-8

Luo B , Wu HN , Huang TW , 2015 . Off-policy reinforcement learning for <math id="M587"><mrow><msub><mi>H</mi><mi>∞</mi></msub></mrow></math> control design . IEEE Trans Cybern , 45 ( 1 ): 65 - 76 . doi: 10.1109/TCYB.2014.2319577 http://doi.org/10.1109/TCYB.2014.2319577

Luo B , Yang Y , Liu DR , 2018 . Adaptive Q -learning for data-based optimal output regulation with experience replay . IEEE Trans Cybern , 48 ( 12 ): 3337 - 3348 . doi: 10.1109/TCYB.2018.2821369 http://doi.org/10.1109/TCYB.2018.2821369

Luo B , Yang Y , Liu DR , 2021 . Policy iteration Q -learning for data-based two-player zero-sum game of linear discrete-time systems . IEEE Trans Cybern , 51 ( 7 ): 3630 - 3640 . doi: 10.1109/TCYB.2020.2970969 http://doi.org/10.1109/TCYB.2020.2970969

Mehraeen S , Dierks T , Jagannathan S , et al. , 2013 . Zero-sum two-player game theoretic formulation of affine nonlinear discrete-time systems using neural networks . IEEE Trans Cybern , 43 ( 6 ): 1641 - 1655 . doi: 10.1109/TSMCB.2012.2227253 http://doi.org/10.1109/TSMCB.2012.2227253

Modares H , Lewis FL , Jiang ZP , 2015 . <math id="M588"><mrow><msub><mi>H</mi><mi>∞</mi></msub></mrow></math> tracking control of completely unknown continuous-time systems via off-policy reinforcement learning . IEEE Trans Neur Netw Learn Syst , 26 ( 10 ): 2550 - 2562 . doi: 10.1109/TNNLS.2015.2441749 http://doi.org/10.1109/TNNLS.2015.2441749

Prokhorov DV , Wunsch DC , 1997 . Adaptive critic designs . IEEE Trans Neur Netw , 8 ( 5 ): 997 - 1007 . doi: 10.1109/72.623201 http://doi.org/10.1109/72.623201

Rizvi SAA , Lin ZL , 2018 . Output feedback Q-learning for discrete-time linear zero-sum games with application to the H-infinity control . Automatica , 95 : 213 - 221 . doi: 10.1016/j.automatica.2018.05.027 http://doi.org/10.1016/j.automatica.2018.05.027

Sakamoto N , van der Schaft AJ , 2008 . Analytical approximation methods for the stabilizing solution of the Hamilton–Jacobi equation . IEEE Trans Autom Contr , 53 ( 10 ): 2335 - 2350 . doi: 10.1109/TAC.2008.2006113 http://doi.org/10.1109/TAC.2008.2006113

Sutton RS , Barto AG , 1998 . Reinforcement Learning: an Introduction . MIT Press , Cambridge, USA .

Valadbeigi AP , Sedigh AK , Lewis FL , 2020 . <math id="M589"><mrow><msub><mi>H</mi><mi>∞</mi></msub></mrow></math> static output-feedback control design for discrete-time systems using reinforcement learning . IEEE Trans Neur Netw Learn Syst , 31 ( 2 ): 396 - 406 . doi: 10.1109/TNNLS.2019.2901889 http://doi.org/10.1109/TNNLS.2019.2901889

Vamvoudakis KG , Modares H , Kiumarsi B , et al. , 2017 . Game theory-based control system algorithms with real-time reinforcement learning: how to solve multiplayer games online . IEEE Contr Syst Mag , 37 ( 1 ): 33 - 52 . doi: 10.1109/MCS.2016.2621461 http://doi.org/10.1109/MCS.2016.2621461

Watkins CJCH , Dayan P , 1992 . Q -learning . Mach Learn , 8 ( 3 ): 279 - 292 . doi: 10.1007/BF00992698 http://doi.org/10.1007/BF00992698

Wei QL , Lewis FL , Sun QY , et al. , 2017 . Discrete-time deterministic Q -learning: a novel convergence analysis . IEEE Trans Cybern , 47 ( 5 ): 1224 - 1237 . doi: 10.1109/TCYB.2016.2542923 http://doi.org/10.1109/TCYB.2016.2542923

Wei YF , Wang ZY , Guo D , et al. , 2019 . Deep Q-learning based computation offloading strategy for mobile edge computing . Comput Mater Contin , 59 ( 1 ): 89 - 104 . doi: 10.32604/cmc.2019.04836 http://doi.org/10.32604/cmc.2019.04836

Yan HS , Zhang JJ , Sun QM , 2019 . MTN optimal control of SISO nonlinear time-varying discrete-time systems for tracking by output feedback . Intell Autom Soft Comput , 25 ( 3 ): 487 - 507 .

Zhang HG , Qin CB , Jiang B , et al. , 2014 . Online adaptive policy learning algorithm for <math id="M590"><mrow><msub><mi>H</mi><mi>∞</mi></msub></mrow></math> state feedback control of unknown affine nonlinear discrete-time systems . IEEE Trans Cybern , 44 ( 12 ): 2706 - 2718 . doi: 10.1109/TCYB.2014.2313915 http://doi.org/10.1109/TCYB.2014.2313915

Zhong XN , He HB , Wang D , et al. , 2018 . Model-free adaptive control for unknown nonlinear zero-sum differential game . IEEE Trans Cybern , 48 ( 5 ): 1633 - 1646 . doi: 10.1109/TCYB.2017.2712617 http://doi.org/10.1109/TCYB.2017.2712617

Zhu YH , Zhao DB , Li XJ , 2017 . Iterative adaptive dynamic programming for solving unknown nonlinear zero-sum game based on online data . IEEE Trans Neur Netw Learn Syst , 28 ( 3 ): 714 - 725 . doi: 10.1109/TNNLS.2016.2561300 http://doi.org/10.1109/TNNLS.2016.2561300

浏览量

113

Downloads

CSCD

文章被引用时，请邮件提醒。

Submit

工具集

关联资源

Optimal synchronization control for multi-agent systems with input saturation: a nonzero-sum game

Jiu fusion artificial intelligence (JFA): a two-stage reinforcement learning model with hierarchical neural networks and human knowledge for Tibetan Jiu chess

Efficient learning of robust multigait quadruped locomotion for minimizing the cost of transport

Reinforcement learning based privacy-preserving consensus tracking control of nonstrict-feedback discrete-time multi-agent systems

Domain adaptation in reinforcement learning: a comprehensive and systematic study

线性离散时间系统H∞控制的极小极大Q-学习设计

Minimax Q-learning design for H∞ control of linear discrete-time systems

DOI：10.1631/FITEE.2000446

摘要

Abstract

关键词

Keywords

references

线性离散时间系统H_∞控制的极小极大Q-学习设计

Minimax Q-learning design for H_∞ control of linear discrete-time systems