• Citer

Scalability and Performance Improvement of Distributed Sequential Command Governor Strategies via Graph Colorability Theory

par Casavola, Alessandro;Garone, Emanuele ;Tedesco, Francesco
Référence (Cape Town, South Africa), IFAC Wold Congress, page (9400-9405)
Publication Publié, 2014

Publication dans des actes

• Détails
• Contenu

Résumé :

Graph colorability theory is shown here to be instrumental for the improvement of scalability and performance of recently introduced distributed non-cooperative sequential Command Governor (CG) strategies. Such properties were limited in earlier distributed CG schemes because the structure of constraints was not taken into account in their implementation. Here, by exploiting the idea that agents that are not jointly involved in any coupling constraint can simultaneously update their control actions, this is done by grouping all agents in the network into particular subsets (Turns) and allowing all agents belonging to a single turn to update simultaneously their commands while agents in other turns are instructed to keep applying their current commands. Graph colorability concepts and results are used to determine the minimum number of independent turns and distribute exclusively agents into them. A Turn-Based distributed CG strategy is therefore proposed and its main properties analyzed. A final example is also presented to illustrate the effectiveness of the proposed strategy by comparisons.