Loading...

Computational load reduction in model predictive control of nonlinear systems via decomposition

Adelipour, S ; Sharif University of Technology | 2018

504 Viewed
  1. Type of Document: Article
  2. DOI: 10.1109/ICCIAutom.2017.8258681
  3. Publisher: Institute of Electrical and Electronics Engineers Inc , 2018
  4. Abstract:
  5. The aim of this study is to reduce the computational load in model predictive control of multi-input nonlinear systems. First, the nonlinear system which has a high number of states and inputs is decomposed into several subsystems by solving a linear integer programming problem offline. Then, the model of each subsystem is revised by considering the effect of coupling and interactions of other subsystems. Next, the robust model predictive technique based on linear matrix inequalities is employed to compute control signal for each subsystem. An industrial chemical reaction example is used to illustrate the effectiveness of the proposed method. © 2017 IEEE
  6. Keywords:
  7. Computational load ; Linear matrix inequality ; Multi-input nonlinear systems ; Integer programming ; Linear matrix inequalities ; Nonlinear systems ; Robustness (control systems) ; Computational loads ; Control signal ; Linear integer programming ; Multi-input nonlinear ; Number of state ; Offline ; Robust modeling ; System decomposition ; Model predictive control
  8. Source: 5th International Conference on Control, Instrumentation, and Automation, ICCIA 2017, 21 November 2017 through 23 November 2017 ; Volume 2018-January , 2018 , Pages 216-221 ; 9781538621349 (ISBN)
  9. URL: https://ieeexplore.ieee.org/document/8258681