Loading...

Recursive Feasibility of Distributed Model Predictive Control: Application to Automated Irrigation Networks

Eassapour Haftkhani, Sajjad | 2016

877 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 48576 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Farhadi, Alireza
  7. Abstract:
  8. The following thesis is concerned with the control of automated irrigation channel using model predictive control in order to eliminate upstream transient error propagation and amplification phenomenon. It is illustrated in the literature that one way of dealing with this phenomenon is to solve a constrained optimization problem at each receding horizon. However, this problem becomes infeasible for automated irrigation channels. Several methods in the literature have been proposed that exploit terminal constraints to guarantee the feasibility of this problem. Finding these constraints, however, is extremely time-consuming as well as impractical for large-scale automated irrigation channels. Hence, in order to guarantee the feasibility, a novel method for estimating those terminal constraints is proposed and applied for both centralized and distributed model predictive control; and then, the explicit solution, which significantly reduces computational complexity, for both centralized and distributed model predictive controls are derived. Subsequently; the performance of model predictive controls which exploit the proposed method are evaluated by doing several computer simulations on an automated irrigation channel
  9. Keywords:
  10. Distributed Predictive Control (DPC) ; Automated Irrigation Networks ; Feasibility Study ; Computer Simulation ; Recursive Feasibility

 Digital Object List

 Bookmark