Performance and design of consensus on matrix-weighted and time scaled graphs

TL;DR

This paper analyzes how multi-time scale consensus dynamics on matrix-weighted graphs affect network robustness to noise, proposing optimization methods for design parameters and demonstrating an application in multi-vehicle formation control.

Contribution

It introduces a novel analysis of the $\\mathcal{H}_2$-norm in multi-time scale consensus systems on matrix-weighted graphs and develops optimization algorithms for their design.

Findings

Separation of weighting and scaling effects on the $\\mathcal{H}_2$ norm.

Optimization algorithms effectively minimize network response to noise.

Application demonstrated in multi-vehicle formation control.

Abstract

In this paper, we consider the $\mathcal{H}_2$-norm of networked systems with multi-time scale consensus dynamics and vector-valued agent states. This allows us to explore how measurement and process noise affect consensus on matrix-weighted graphs by examining edge-state consensus. In particular, we highlight an interesting case where the influences of the weighting and scaling on the $\mathcal{H}_2$ norm can be separated in the design problem. We then consider optimization algorithms for updating the time scale parameters and matrix weights in order to minimize network response to injected noise. Finally, we present an application to formation control for multi-vehicle systems.