Abstract
An overall evaluation of the computational efficiency of two decentralized optimal control approaches, namely, the non-singular perturbation and the hierarchical control approaches, is performed in order to provide the grounds for a quantitative comparison between them. The linear-quadratic regulator problem is adopted as the vehicle to conduct the comparative study. Major conclusions drawn from this comparison are the computational superiority, wide applicability and potential flexibility of the hierarchical control approach over the non-singular perturbation approach.

This publication has 7 references indexed in Scilit: