Comparison of the Time-Delay Margin of a Distributed and Centralized Observer

The benefit of a distributed observer concept for large-scale linear plants is shown by taking time-delays into account. It is asserted that a centralized observer suffers from delays in the measurement input, while a distributed structure allows to avoid them. In contrast, the network of the distributed observer induces communication delays among observer nodes. The delay margins for both observer concepts are estimated on a numerical example and compared using an eigenvalue-based frequency domain approach and an LMI based time-domain approach.