Paper Info
Title
Continuously Generalized Model Predictive Control
Abstract
We introduce a generalized framework for model predictive control (MPC) based on the repetitive utilization of finite horizon open loop optimal control (OLOC) with a current-state-dependent terminal constraint set and cost function (and control law). We employ continuously indexed terminal constraint sets, cost functions and control laws, and we allow for their (joint with the predicted state and control sequences) online optimization. The proposed parametrization of these terminal ingredients renders the related online optimization computationally feasible, and it facilitates a relaxation of the standard stabilizing MPC conditions. The developed framework outperforms conventional MPC in terms of structural properties, and it also enhances applicability and computability of our previously proposed discretely generalized MPC.
Year
Venue
Field
2016
2016 IEEE 55TH CONFERENCE ON DECISION AND CONTROL (CDC)
Mathematical optimization,Optimal control,Parametrization,Control theory,Computer science,Model predictive control,Computability,Online optimization,Finite horizon,Open-loop controller,Numerical stability
DocType
ISSN
Citations 
Conference
0743-1546
0
PageRank 
References 
Authors
0.34
0
3
Name
Order
Citations
PageRank
Sasa V. Rakovic100.34
William S. Levine2135.15
Behçet Açikmese34115.88