Skip to main navigation Skip to search Skip to main content

An iterative algorithm based on model-reality differences for discrete-time nonlinear stochastic optimal control problems

  • Sie Long Kek
  • , Mohd Ismail Abd Aziz
  • , Kok Lay Teo
  • , Rohanin Ahmad
  • Universiti Tun Hussein Onn Malaysia
  • Universiti Teknologi Malaysia
  • Curtin University

Research output: Contribution to journal › Article › peer-review

3 Citations (Scopus)

Abstract

An iterative algorithm, which is called the integrated optimal control and parameter estimation algorithm, is developed for solving a discrete time nonlinear stochastic control problem. It is based on the integration of the principle of model-reality differences and Kalman filtering theory, where the dynamic integrated system optimization and parameter estimation algorithm are used interactively. In this approach, the weighted least-square output residual is included in the cost function by appropriately monitoring the weighted matrix. An improved linear quadratic Gaussian optimal control model, rather than the original optimal control problem, is solved. Subsequently, the model optimum is updated using the adjusted parameters induced by the differences between the real plant and the model used. These updated solutions converge to the true optimum, despite model-reality differences. For illustration, the optimal control of a nonlinear continuous stirred tank reactor problem is considered and solved by using the method proposed.

Original languageEnglish
Pages (from-to)109-125
Number of pages17
JournalNumerical Algebra, Control and Optimization
Volume3
Issue number1
DOIs
Publication statusPublished - Jan 2013
Externally publishedYes

Keywords

  • Adjusted parameters
  • Iterative algorithm
  • Model-based optimal control problem
  • Model-reality differences
  • Nonlinear stochastic optimal control

Fingerprint

Dive into the research topics of 'An iterative algorithm based on model-reality differences for discrete-time nonlinear stochastic optimal control problems'. Together they form a unique fingerprint.

Cite this