A Distributed Model Predictive Controller for Linear Systems with Automatic Partitioning

Rosiane R. Rocha and Luís Cláudio Oliveira-Lopes

Keywords

Distributed control, Model based control, Decomposition methods

Abstract

This paper introduces a distributed model predictive control strategy based on a new plant decomposition. The new decomposition is proposed directly based on sets with elements of the state (A, B, C) matrices, representing interactions between controlled outputs, states and inputs. The proposed strategy was applied in a highly ill-conditioned system and the properties of the controller were evaluated. Furthermore, the performance of the developed controllers was compared to one from centralized MPC (Model Predictive Control) structure. The results indicate that the proposed framework provides an acceptable performance when compared to a centralized MPC within a distributed framework. Furthermore, the new methodology extends previous results by dealing with processes with full B matrix.

Important Links:

Go Back