• DocumentCode
    824278
  • Title

    The uncertainty threshold principle: Some fundamental limitations of optimal decision making under dynamic uncertainty

  • Author

    Athans, Michael ; Ku, R. ; Gershwin, S.

  • Author_Institution
    Massachusetts Institute of Technology, Cambridge, MA, USA
  • Volume
    22
  • Issue
    3
  • fYear
    1977
  • fDate
    6/1/1977 12:00:00 AM
  • Firstpage
    491
  • Lastpage
    495
  • Abstract
    This note shows that the optimal control of dynamic systems with uncertain parameters has certain limitations. In particular, by means of a simple scalar linear-quadratic optimal control example, it is shown that the infinite horizon solution does not exist if the parameter uncertainty exceeds a certain quantifiable threshold; we call this the uncertainty threshold principle. The philosophical and design implications of this result are discussed.
  • Keywords
    Linear systems, stochastic discrete-time; Optimal stochastic control; Stochastic optimal control; Uncertain systems; Additive noise; Control systems; Decision making; Econometrics; Optimal control; Riccati equations; Stochastic processes; Stochastic resonance; Stochastic systems; Uncertainty;
  • fLanguage
    English
  • Journal_Title
    Automatic Control, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9286
  • Type

    jour

  • DOI
    10.1109/TAC.1977.1101526
  • Filename
    1101526