全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

带饱和执行器的非线性离散时滞系统的最优控制

DOI: 10.3969/j.issn.1005-3026.2014.04.002, PP. 461-464

Keywords: 近似动态规划,启发式动态规划,值函数,神经网络,最优控制

Full-Text   Cite this paper   Add to My Lib

Abstract:

主要针对带有饱和执行器的时滞非线性离散时间系统更加一般的形式,通过启发式动态规划(HDP)算法求解无限时间最优控制策略问题,并在值函数中引入折扣因子.首先通过迭代HDP算法给出值函数序列和相应的控制序列,并给出了收敛性证明,即值函数序列收敛到值函数的最优值,以及控制序列收敛到最优控制;其次为了实现HDP算法,引入3个神经网络:模型网络、评判网络、控制作用网络.模型网络用来近似系统模型,评判网络用来近似值函数,控制作用网络用来近似控制;最后通过一个仿真例子说明上述方法的可行性.

References

[1]  Lewis F L,Syrmos V L.Optiaml control[M].2nd ed.Hoboken:Wiley,1995.
[2]  Murray J J,Cox C J,Lendaris G G,et al.Adaptive dynamic programming[J].IEEE Transactions on Systems,Man and Cybernetics:C,2002,32(1):140/153.
[3]  Tamimi A,Lewis F L,AbuKhalaf M.Discretetime nonlinear HJB solution using approximate dynamic programming:convergence proof[J].IEEE Transactions on Systems,Man and Cybernetics:B,2008,38(4):943/949.
[4]  Werbos P J.Approximate dynamic programming for realtime control and neural modeling[M].New York:Van Nostrand Reinhold,1992.
[5]  Sussmann H J,Sontag E D,Yang Y D.A general result on the stabilization of linear systems using bounded controls [J].IEEE Transactions on Automatic Control,1994,39(12):2411/2425.
[6]  Saberi A,Lin Z L,Teel A R.Control of linear systems with saturating actuators[J].IEEE Transaction on Automatic Control,1996,41(3):368/378.
[7]  Luo Y H,Zhang H G.Approximate optimal control for a class of nonlinear discretetime systems with saturating actuators[J].Progress in Natural Science,2008,18:1023/1029.
[8]  Wei Q L,Zhang H G,Liu D R,et al.An optimal control scheme for a class of discretetime nonlinear systems with time delays using adaptive dynamic programming[J].Acta Automatica Sinica,2010,36(1):121/129.
[9]  Song R Z,Zhang H G,Luo Y H,et al.Optimal control laws for timedelays systems with saturating actuators based on heuristic dynamic programming[J].Neurocomputing,2010,73:3020/3027.
[10]  Wang F Y,Jing N,Liu D R,et al.Adaptive dynamic programming for finitehorizon optimal control of discretetime nonlinear systems with εerror bound[J].IEEE Transaction on Neural Networks,2011,22(1):24/36.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133