Distributed Optimal Control Using an Efficient Method for Decomposition

Please enable javascript in your browser.

Aryankia, Kiarash | 2015

582 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 47673 (05)
University: Sharif University of Technology
Department: Electrical Engineering
Advisor(s): Farhadi, Alireza
Abstract:
This work is concerned with new methods for decomposing large scale systems into disjoint neighborhoods with specific size determined a priori by the communication load of the Jacobi based distributed optimization technique. This iterative optimization technique has the following features: it handles hard control input and state constraints; and for convex optimization problem, in the limit, its solution converges to the optimal solution. The proposed decomposition techniques cluster subsystems based on the strength of control inputs interaction. Three techniques are proposed. The first technique is an extended version of REGA. The second is an extended version of Participation Matrix(PM).The last one is a new technique which is based on the response of the system. It is illustrated via extensive computer simulation that the third technique forming the best results
Keywords:
Large Scale System ; Decomposition ; Distributed Control ; Interaction ; Distributed Predictive Control (DPC)