Clustering-Based Average State Observer Design for Large-Scale Network Systems

TL;DR

This paper introduces a clustering-based approach for designing low-order average state observers in large-scale network systems, enabling efficient estimation of cluster averages with minimal error.

Contribution

It presents a novel method combining clustering and aggregation to design computationally efficient average state observers for large-scale networks.

Findings

Significantly faster computation compared to traditional methods

Maintains acceptable estimation error levels

Applicable to large-scale network monitoring

Abstract

This paper addresses the aggregated monitoring problem for large-scale network systems with a few dedicated sensors. Full state estimation of such systems is often infeasible due to unobservability and/or computational infeasibility. Therefore, through clustering and aggregation, a tractable representation of a network system, called a projected network system, is obtained for designing a minimum-order average state observer. This observer estimates the average states of the clusters, which are identified with explicit consideration to the estimation error. Moreover, given the clustering, the proposed observer design algorithm exploits the structure of the estimation error dynamics to achieve computational tractability. Simulations show that the computation of the proposed algorithm is significantly faster than the usual $\mathcal{H}_2/\mathcal{H}_\infty$ observer design techniques. On the other hand, compromise on the estimation error characteristics is shown to be marginal.